I wasted $40k on a fantastic startup idea | Hacker News Reader

I wasted $40k on a fantastic startup idea

tjcx.me

851 pointsswyx385 comments

That was a fun read. I wish the author mentioned how much he was trying to sell the service for. It could have been $59 a month or $599 a month and with doctors you could potentially expect the same answer.

I'm not a psychologist but some of the author's quoted text came off extremely demeaning in written form. If the author happens to read this, did you really say those things directly to them?

For example, Susan (psychologist) was quoted as saying:

> "Oh sure! I mean, I think in many cases I'll just prescribe what I normally do, since I'm comfortable with it. But you know it's possible that sometimes I'll prescribe something different, based on your metastudies."

To which you replied:

> "And that isn't worth something? Prescribing better treatments?"

Imagine walking into the office of someone who spent the last ~10 years at school and then potentially 20 years practicing their craft as a successful psychologist and then you waltz in and tell them what they prescribe is wrong and your automated treatment plan is better.

It's actually a very good sales question, I don't find it demeaning at all.

If you're on a sales call selling a product that increases user retention and someone says "no we don't need that", you would often reply with "So you have perfect user retention then?" to probe them and re-open the conversation.

It could come off as standoffish but when used correctly it's very effective because it gets the person on the other end to open up more and you try to get to the bottom of their objections.

>If you're on a sales call selling a product that increases user retention and someone says "no we don't need that", you would often reply with "So you have perfect user retention then?" to probe them and re-open the conversation.

Assuming that I didn't initiate the call, if I tell some sales punk that I don't need their product and they come back at me with "So you have perfect user retention then?" my answer is going to be "fuck you" followed by ending the call.

Arrogance might work in used car sales but it's not a panacea for closing the deal.

I'm sure that happens occasionally as well.

I get it, sales people can be annoying but it really only hurts your business (in this hypothetical case) if you have a user retention problems and are actively fighting against people trying to help you solve that problem with a mutually beneficial business agreement.

They will call the next person on their list, I'm sure it won't matter much.

Incoming spam is unwelcome in general, and I would not assume that this people are trying to help me.

>I get it, sales people can be annoying but it really only hurts your business (in this hypothetical case) if you have a user retention problems and are actively fighting against people trying to help you solve that problem with a mutually beneficial business agreement.

Oh please. Not every product that's applicable to a given business would necessarily be beneficial to that business.

I get it, sales people have to drink the kool aid, but some humility is needed. Your product isn't right for everyone, and the sooner you understand that the sooner you can improve the quality of your lead generation.

If you have a vendetta against sales people you can just say that. We are talking about a completely hypothetical situation here, I don't see how you're now implying that the lead generation isn't good enough.

On HN, people hate cold emails. In real life, I've found that most people will respond or ignore. Like a tiny minority will act like you killed their mother, but that's life.

I know you know this, if you're in sales, but I, like many other engineers who read this forum was overly cautious when I first started speaking to people because I anticipated that 99/100 would be upset at having to talk to me.

The truth was that 99/100 were willing to speak to me and listening to HN and Reddit set me back farther than I expected until I unlearned that lesson.

So I'm saying this for the benefit of all those other engineers like me.

Makes sense, you're typically not cold calling or emailing engineers which is presumably the majority of hacker news users

I guess that it is a selection bias.

People hating spam will complain about it, people not hating it will not bother with commenting.

And HN may have higher share of people more affected by spam (surveys send to emails scrapped from github and so on) and more likely to be able to find needed services - what makes beneficial cold calling even more rare.

And I guess that my reaction is unusual. I moved to another phone service after previous one cold called me offering a loan on bank service operating under their brand - and that was a sole reason.

I don't have a vendetta. I used to work in sales. My first job out of school was cold-calling and I saw first hand what types of people tend to rise to the top in that environment.

I quit that job as soon as I could because I was sick of manipulating people or pressuring them into buying something they didn't want. At the end of the day that's all sales is. If your customer wants your product then by definition you don't have to sell them.

That makes a lot of sense then honestly. "Boiler rooms" are definitely not good environments.

>If your customer wants your product then by definition you don't have to sell them.

That's a misnomer, what if they don't know you exist? That's what sales and marketing are for, to inform people that could benefit from your product that it exists, essentially. Will their business implode if they don't buy your product? No, almost certainly not. But, they might derive a massive cost savings, time savings, increased employee satisfaction, or other efficiencies by using whatever you're selling and your job in sales is to get them to listen which is the hard part.

>That's a misnomer, what if they don't know you exist? That's what sales and marketing are for, to inform people that could benefit from your product that it exists, essentially.

This is a fiction that salespeople tell themselves to help them sleep at night. I use twitter not because someone from Twitter Sales sold it to me but because I sought it out of my own volition.

Only bad products need sales. Good products can thrive on marketing alone because people want good products.

We're talking about the context of a cold call here. It's extraordinary unlikely for something that is genuinely beneficial to me to come to my attention via a cold call. Maybe your product is actually good, but there's a wave of pushy salesmen trying to sell snake oil via these types of channels. At best, you've got a very tough job to make a sale to a business under those conditions. I don't know about the typical business, but that kind of talk isn't going to keep me on the line.

HN guidelines: “Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. Assume good faith.”.

I would presume that sentence is a quickly typed paraphrase. If you presume they are a competent salesperson, you can also presume that they say it less antagonistically in real life. Edit: Or perhaps they have found that antagonism is the most profitable solution for the business to turn around a “no” answer.

>HN guidelines

The comment to which I'm responding explicitly argues that the phrase, verbatim, is a good sales tactic.

People have argued that the tone and context in which it is being said will influence effectiveness. Verbatim has nothing to do with it but you keep ignoring that point.

you referred to the caller as a sales punk before he uttered the phrase, exposing your attitude. Just because you are biased against sales people and also have a thin skin when it comes to the slightest challenge in a question does not mean that the product would not be beneficial and worth it to you if you had a more open mind, nor does it mean that the sales technique on average is a failure.

i.e. when I read your comment it struck me as much more revealing about you than about the salesman, his technique, or his product.

I was ready to agree with you because question in the OP wasn't so bad, but "So you have perfect user retention then?" -- seriously? Yeah, that's being demeaning. Maybe being demeaning is a good sales technique, I dunno, but that's definitely being a jerk.

But "Are you sure increasing your user retention isn't worth something to you?" or something like that maybe.

It really depends on your tonality when you say it. In either case, saying "Are you sure increasing your user retention isn't worth something to you?" is just going to be met with another "No, we don't need it" from the other end.

"So you have perfect user retention then?" is a better question because you know for a fact that they can't have 100% user retention and they know that as well so it forces further dialogue.

i guess this is why I'm not a salesperson. If it works it works, but I'm having trouble accepting that it's not demeaning, which was the original contention. Because it's not really a question at all, it's a sarcastic question. Maybe being demeaning gets sales, sure.

Agree with you on that.

Think of it more as being a pattern interrupt. They are saying no to you over and over, so by asking the question, "So you have perfect user retention then?" they can either say "No" instinctively which opens your sales pitch because they've now told you they have a user retention problem, or they stop and think about the question in which case you've successfully interrupted the pattern of no's that preceded the question and can further the discussion.

There are definitely better examples I could come up with, but now I'm stuck with this one because it's what I quickly typed out earlier.

Sales - like marketing - definitely does have an element of psych to it.

Anecdotally, it feels as though it leans on a lot on the targets level of politeness and decorum to want to continue the conversation. I favour polite but abrupt and seemingly heartless conversation enders for this reason.

"Hello sir, would you like to help a child keep eating for $2 a month?" "No, thank you!"

Back to this specific case, I guess the trick is to deliver your question almost rhetorically with both parties knowing that it leaves an answer that is common knowledge. Kinda like safe small talk about the weather.

Keeps the conversation going buying time for another hook to be deployed.

> Because it's not really a question at all, it's a sarcastic question

It's not sarcastic, it's serious and that's the point.

Obviously they don't have 100% retention, so this question might open them up to talk about their retention instead of saying "no, it's fine like it is".

It doesn't matter how good your retention currently is, if the product can boost it even by a couple of percent it would probably pay for itself many times over.

I think you could make the argument that saying "no, we don't need it" before even trying to understand the value prop is just as demeaning.

You seem to have an idiosyncratic definition of “demeaning”.

Arg I HATE that angle ! It's like that sales calls.. "Do you like money ?".. if you say no thank you I"m busy... they like... oh so you don't like free money ?

beep beep beep

The conversations are surely paraphrased and exaggerated—just look at the style of the rest of the article.

The position that doctors should be trying new things to improve their care sounds good but in practice most doctors are strongly biased towards the status quo and usually inaction is preferred to a slightly unknown action, even if that action has better expectancy.

I dunno. I think tech people have a tendency to assume they know for certain that they have a better solution, and their words/tone can reflect that in a way that can come off as very insulting to people who have been working in the space for literal decades.

Not everyone, and not all the time. But many people, and often enough that it's a stereotype. So I think it's worth considering, particularly when you're looking at a customer base who (by and large) really aren't used to being condescended to.

Even if you and I personally aren't offended in the slightest by what OP said on sales calls, it's possible a psychologist in the fourth decade of her career might take "Are you sure that's right?" differently than we would.

He struck me as completely oblivious to what was likely to have been a complete lack of interest.

His approach was never going to work, as doctors do not spend their time evaluating drugs in the way that he imagined.

It's a great example of "I'm going to make a pretty chart, and sell to.... $PROFESSION!"

It may have been a great product like you and others have said, he hasn't the faintest idea what physicians actually do day-to-day. He had apparently spent $40k over nearly a year before he talked to the first physician.

That's because his first plan was to sell to consumers. Only when that didn't work out he switched to physicians.

Like someone else said, he got greedy. Even if he had stuck to the WebMD model, he would have made a fortune. People like to Google symptoms before seeing a doctor, and although unethical, pharma companies would have gone head over heels to market in those spots.

I don’t think that was the point. The point was better healthcare doesn’t necessarily translate to more revenue. Healthcare is weird like that - you get paid a flat fee for visits. There might be an argument to be made that better prescriptions = happier patients = more retention, but it’s a stretch. If your practice is already booked full what’s the point?

This biz was clearly made for consumers but yea ads are tough - need a lot of eyeballs.

Yeah I agree that’s the point of the article. I think what I wrote would still be a reason even if it is rationalised differently.

The argument that “patients won’t know the difference so I can just do whatever” must break down at some point (hopefully before malpractice) but I think an argument of “I’ll just keep doing what I did before, it’s worked fine so far” doesn’t encourage worsening treatment or paying for more experimentation.

As somebody said here, first line drug is prescribed since it’s know to work based on experience. Any experimentation puts too much unwanted responsibility on the doctor.

Better healthcare outcomes are of interest to health insurance companies and employers, not the healthcare providers. A reduction in overall claims over a patient's lifetime and better employee health are quantifiable financial results.

Healthcare providers hope for better outcomes I'm sure, but it's true there's not much financial incentive in most cases. Your comment sort of suggests the model for such a service should be on a per-use model, covered by an insurance company when a doctor uses it for their patient. That's a trickier model than subscriptions, but it may be viable.

Psychiatrists are simply experimenting on each person they prescribe medications to. The status quo is indoctrination, and in this case, the psychiatrist wasn't even willing to use actual research based data to improve the treatment of their patients - even referencing that they won't make anymore money because the patient won't come back more often or refer more patients because of it. This is abhorrent unprofessional behaviour - but it's likely the attitude of 90%+ of the field.

Edit to add: if you think a practitioner putting their own revenue/profits above providing quality care, then something's wrong with you.

I've seen this happen and the entire process of adding new drugs and waiting a week or more between visits seems cruel.

Lots of psychoactive medications take days or weeks to build up in the body to the point of effectiveness. They're finding out neat things about esketamine and psilocybin, but in general a large enough dose of a mood-altering drug to take immediate effect is a bad or at least very risky thing.

In medicine "quality care" is a binary attribute. Something either meets the quality threshold or it doesn't. And as the author found out, physicians do not spend their time researching medications. It's not their job, they're not trained in it, and most of them wouldn't be particularly good at it.

So I'm not sure why you would expect a psychiatrist to do something she's not trained to do to (maybe) increase an already acceptable metric some arbitrarily small amount higher.

Right, and both doctors and patients don't want to run another experiment. If they are comfortable with a certain drug, that's a good enough indication its a good first step to getting better.

If you are to proud to visit stackoverflow or Google to search for best practices you're not a very good developer. Sounds like the same rule should apply to a doctor or psychiatrist.

Do they have such fragile egos that they can't have someone showing them the "new Google for doctors" without feeling offended?

(Sorry if that came out a bit edgy. But this hole story irks me. It's frustrating when value can't be delivered because of cash flow issues.)

Agree

> Do they have such fragile egos that they can't have someone showing them the "new Google for doctors" without feeling offended?

It's very hard to understand their thinking without being yourself a patient or doctor.

TLDR: You need to spend 1K hours as a doctor, or shadowing, or something similar.

When you are responsible for people with psychosis your perspective changes. Can't play nosql games with people's lives.

One has to be accepting that medical professionals are not superhuman and don't have magic bullets for chronic illness. But I don't know where you are coming from with such a pompous statement about being "responsible" for psychosis. The minimum "responsibility" that I can imagine being taken for a patient who has a crisis would be to move a regular appointment up after (or before) they land in the ER, and in the healthcare system I'm familiar with, it doesn't appear to happen. Are you used to something different? Or maybe you have a different idea of responsibility?

It's nothing special, just a job. But people are perplexed and they can't imagine why doctors are like that. I said you have to do the hours to understand, nothing special.

Doctors do have curated platforms where they can lookup information, and also have diagnosis tools / checklists in them.

> you waltz in and tell them what they prescribe is wrong and your automated treatment plan is better.

That doesn't really capture the conversation though. Susan specifically said (as you quoted) that it was possible she would use the recommendations of the app. If she took a recommendation that means she agreed it was a better treatment. His question was not whether his algorithm was better than her default prescriptions - because they both agreed that was the case at least some of the time - it was whether it was better enough to be financially worthwhile.

Better didn't factor in for the doctor at any cost even free.

This tech could be used to replace the function of this doctor making them less useful.

You’d be surprised how many doctors neglect the state of the art in medicine... That’s also why second opinions are a thing. Medicine is a science and hence, an ever changing field.

I probably would be surprised if that's true considering that in most states in the US, doctors are required to complete between 20 and 50 hours of structured CME (continued medical education) annually as a prerequisite to relicensure, and every one I've spoken to took it pretty seriously.

They are not unaware of the new stuff. But it doesn’t enter their day to day until it’s reached critical mass or if they’ve identified a couple suitable patients to recommend it to. Eg. Advil’s not working for a patient’s headache so, tell patient to try Tylenol because I learned about it during CMEs. Advil remains default for a long while. In many cases, until this doctors kid replaces him in the practice. Their generation learned in med school that Tylenol should be default.

Are you a physician, work in medical education, or have some other sort of broader source of information than that of a patient? I've primarily had experience as a patient, but it certainly wasn't what you describe. Obviously anybody limited to a patient's perspective wouldn't have anything more than anecdotal data.

Though back in the day, I did support for a CD-ROM based CME product... It was simple software so most of the people who called were in their 70s and 80s and not very comfortable with technology. They all seemed pretty concerned with learning the new stuff.

I worked along side physicians for several years. I'm still in healthcare but more corporate. I have interacted with physicians as a clinician, business partner, and patient. They love to learn that's no doubt. But the OP content was about changing their prescription protocols based on some hacked together ML that has no traditional clout. That was never going to fly; not without some [insert specialty] doctor's endorsement or some medical journal nods.

When it comes to the business of healthcare, doctors are generally open to things that will make their jobs easier (they are money motivated creatures too). But, they don't necessarily like learning new things themselves. As a whole, they are rather resistant to change. It took a long time to incorporate any technology bedside (e-charts, etc).

As a patient, I/family has been referred to other doctors specifically because they were younger or knew their education differed in a meaningful way. The old guy knew his ways were outdated and we would benefit from the new ways. So he was aware of the new way, but wasn't practicing it.

A second opinion is a mainly a thing because different people have different approaches. And they can be both correct.

Based on the response he got, it was the right question, actually. People aren't Internet-style insecure in real-life, especially those who have high social cred (like doctors). Even accounting for the humorous exaggeration, the kind of question asked from a professional doctor is less likely to cause them to be offended and more likely for them to just tell you why not. They're not going to be "How dare you question my decades of experience?!". They'll answer like they did in the OP.

In The Mom Test, he suggests getting right to the core of a customer's pain points. This is just corollary to that.

Although do note that doctors are just humans, with all the same flaws. There's just a bias in the kinds of humans that tend obtain the title.

Yeah, I do clinical placements in hospitals as part of volunteer work, and doctors run the range from open and receptive to questions to incredibly arrogant and dismissive.

The fact that they agreed to talk to the OP probably means they aren't going to be immediately dismissive, but I'd hesitate to assign any supranormal human traits to doctors.

He had not the slightest idea of how doctors prescribe drugs.

The typical doctor has minimal training in evaluating medicines - that is not their job.

They defer to so-called opinion-leaders, who are the experts on particular diseases.

These people are the targets of drug companies' marketing - think scientific conferences in 5 star hotels in exotic locations.

The cost of influencing them would be millions.

So,the author was barking up the wrong tree.

That's not to say that he didn't have something, but had no idea how to market it.

Some doctors use expert systems. They select symptoms and computer spits out possible list of treatments and then doctor picks one. If it doesn't work asks to come back and tries the next one. It's kind of like a human in today's self driving cars. Especially when it comes to mental health and anti-depressants. Essentially tests on production.

This isn't even close to how doctors prescribe medications. You don't prescribe meds without having a working diagnosis. Once you have that, you use the knowledge gained in med school and residency to pick the first line drug. If there are contraindications due to comorbidities (which there often are), you have to figure out what other meds you can use. You can consult online resources (e.g. UpToDate) to look up second, third, fourth line meds as well as advice on specific complicated scenarios.

Trial and error with prescription drugs without a diagnosis as you suggest is malpractice. Maybe you're specifically referring to psychiatry? That specialty is uniquely difficult since our understanding of psychiatric diseases is still murky. But even within psychiatry there are best practice guidelines on how to manage and treat different diseases.

This is what I saw my doctors were doing. Also I saw cardiologist comparing my diagnosis using Google images.

Having software to show differential diagnoses, or using Google images because you know to search for, are not the smoking guns you think they are.

Yeah, I believe your experience, and I think it highlights the issues we have with communication in medicine. While it may look like your doctor is just blindly Googling something, I would imagine they're probably using it as more of a reference source (at least that's what I often do). I regularly use radiopaedia.org just to look up a quick fact or find alternative examples of a diagnosis I'm working with.

It's like Googling coding questions and reading a StackOverflow thread. Obviously no programmer is solely relying on StackOverflow to do their job as no physician is solely relying on Google, UpToDate, or any other resource. They're simply quick references.

> Obviously no programmer is solely relying on StackOverflow to do their job

I've encountered a few people who were doing something very close to this. I really hope that doesn't happen in medicine too.

That's down to the (obvious) fact that job performance and working ethics is not equal but distributed among practitioners.

I think this is especially visible in software "engineering" with people joining the craft after a few weeks of boot camp. (think engineering vs programming)

However, we put doctors through an especially rigorous and long training and certification process to minimize the amount of unqualified practitioners.

Doctors are supposed to have a higher entry barrier than software developers. Does not mean that all of them are brilliant, of course.

You might be right for single-casual illnesses like a broken leg, CoVid-19, Tuberculosis, and others. But it's a whole other thing when it comes to more complex illnesses.

Having been for 18 months through different types of psychiatrists and clinics, I came out quite surprised in how "trial and error" this whole system is.

I'm writing this from Switzerland, where we have an (arguably) high quality health care system. But the amount of "OK, that didn't work, lets try this other drug". Or, now, 18 months later, "Oh well, we did the list once through. But who tells me that the MD prescribing the first drug did a correct analysis? Lets start the list from top again." Or, for a friend of mine, his girlfriend found a working cure like that: "Oh, this brochure describes your symptoms so clearly, and it's completely different from what you've been treated for these last 20 years. Let's try it!"

Spoiler alert for my case with the list: the top of the list was not better the second time around.

If you think this is cynical, well, I would like at least _one_ of these drugs to work. If you think the MDs are all useless: well, at some times I was glad they were there.

So, well, I think having a little less than random system might be helping. Let the MDs watch if it makes sense, enter the correct diagnosis, and catch the stupid errors data entry people can make. But i'd give it a try...

The closest thing to an expert system I've seen is clinical practice guidelines, which sometimes includes decision trees of indications/contraindications for administering certain medications for common and time-sensitive medical events, like cardiac arrest or exacerbation of breathing difficulty with COPD.

Hey remember when the opioid companies paid the clinic management(?) software companies to push opioids to people?

https://www.washingtonpost.com/nation/2020/01/28/opioid-kick...

Part of regulatory capture and the industrial complex, why medicines like MDMA and psychedelics have been illegal for decades due to them not being patentable - and being competition that most recent research shows is far more effective and without the "side" effects of many big pharma's drugs.

Far more effective? Source?

For a user named ketamine_ are you really questioning the well known efficacy of few-shot therapies for treatment resistant depression and PTSD?

Psychedelic-Assisted Psychotherapy: A Paradigm Shift in Psychiatric Research and Development: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6041963/

Reviewing the Potential of Psychedelics for the Treatment of PTSD: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7311646/

etc.

> For a user named ketamine_ are you really questioning the well known efficacy of few-shot therapies for treatment resistant depression and PTSD?

Maybe they want to convince their friends?

Not all experts are cheerleaders.

I'm questioning your enthusiasm because I have experience and have read the research.

It is still early days. If one can end depression with an oral medication that isn't scheduled one is better off.

This is so removed from reality to the point that it’s hilarious. Doctors evil. Doctors bad. Doctors corrupt. Doctors rich. That psychiatrist in the article sure must have been bribed to prescribe those 30-year-old drugs, right? There’s this thing called evidence-based medicine, go educate yourself.

I mean, there's a reason pharmaceutical companies in the US spend $20 billion a year marketing to physicians and it's not because it doesn't work. Doctors in the end are human and as capable of being influenced and biased and taking shortcuts as any other person. That doesn't mean they're evil, just human.

A big chunk of that $20B is on high cost ads to the general public, which doesn’t occur in many other parts of the world. It’s a backwards system when a patient is told to ask their doctor for a prescription to a medication by the company that manufactures said medication.

Yes, one of the few culture shocks when I lived in America was pharmaceutical advertising to the public. Another was constant political advertising on TV.

It is also a source of culture shock for Americans like me who visit friends and relatives that watch television. I mean, I watch netflix, but haven't ever had a cable subscription, and while I have a digital antena, I pretty much never use it. Visiting a home where the TV is on constantly is shocking, with all the ads and the viewpoints presented. It really shows how much the media fracturing is helping drive the political divide in the country too. I can't really imagine having those messages constantly pounded into my head.

No, that $20 billion is ONLY for doctor marketing. There's an additional $10 billion on top of that spent on patient advertising/marketing.

edit: There's a whole industry around providing pharma companies better tools to influence doctors. I believe the industry name for this part of pharma companies is Medical Affairs so feel free to google the tooling being offered.

Nobody has to be evil or bad to read the white papers presented by their vendors rather than doing independent research. A doctor's job is not to figure out the best possible treatment for each patient, as nice as that sounds. It's to improve the life enough of enough people given the time available. If four drugs are approximately equivalent don't blame your doctor if you get the one that's only 90% as effective in your particular case.

> Imagine walking into the office of someone who spent the last ~10 years at school and then potentially 20 years practicing their craft as a successful psychologist and then you waltz in and tell them what they prescribe is wrong and your automated treatment plan is better.

As I understand it, this is exactly what new pharmaceutical sales reps are asked to do?

It's not hard to get a doctor's attention when he knows you will provide him with golf days and holidays, if he plays his cards right.

At least in the UK, sandwiches are the most that pharma reps can use to bribe doctors with... Which is not to say they aren't effective (you'll get butts in chairs at least, no guarantee they will pay any attention to you though).

Damned bribery and corruption act. I could really do with some free holidays and lavish parties. Now it's all branded mugs and that's your lot. Someone offered us free beer and we had to refuse. Oh, the humanity!

No, the sales rep is selling the outcome of the knowledge of the medical experts within the company.

In the same way the author was selling the outcome of the knowledge of their algorithm.

The author was selling the outcome of the knowledge of many expert medical researchers, as summarised by the author's algorithm.

As opposed to the sales rep and the medical experts within the sales rep's company, neither the author nor his algorithm have any obvious direct incentive to recommend one medication over another.

His writing style seems similar to Hunter Thompson’s - I wouldn’t read into it too deeply, exaggeration is the backbone. Personally I enjoy it.

As for the actual content, there’s a massive difference between customers dying to use your product and them telling you it’d be “neat”. People don’t buy “neat” products. This is why you talk money to them as soon as possible. No real surprise it didn’t work out for him, the incentives just aren’t there.

He could have prevented all this by reading the Mom Test - oh well, experience is the best teacher anyways.

And you back it up with proof but you hear I don't care about better treatment I care about prescribing what I feel works best and what pays best. Science be damned.

There in a nutshell is the problem with healthcare. Doctors care about different outcomes. Doctors have this image that they care about your best health outcome but they would always trade an extra dollar over any patient outcome as long as they are legally within guidelines.

As a developer. I went to school for a period of time. I have had a 20 year career of sucessful development jobs/projects.

If someone walked in without any experience but showed me a better way to develop a project through understandable datapoints I would listen and not ignore them because I was somehow all knowing. I may even buy. Why are Doctors different?

Note: psychologists cannot prescribe anything here (Eastern Europe), psychiatrists can.

I do not see any issues telling doctors that they are prescribing the wrong medication because you may know it better than them. I personally learnt a lot about pharmacology due to my illness, and I know what kind of medications are should not be supposed to be prescribed which the doctor working at the ER at that moment may not. In any case, you could just point out the reasons, I believe.

For example: highly lipophilic beta-blockers are an issue in my case as I get anxious over its CNS side-effects that others may take it for performance anxiety, see: propranolol. The doctor prescribing me propranolol would be in the wrong. My own psychiatrist had no clue that lipophilic beta-blockers can cause all sorts of issues (CNS side-effects, such as brain fog) that lead to anxiety and even panic attacks for me. In my case, something like atenolol or even nebivolol would have worked better. Some doctors know this, some do not. I do not reasonably expect them to know everything though. It is sort of a detail that is not known because it is commonly prescribed to the elderly who do not report those CNS side-effects because they attribute them to their old age.

There were cases of metoprolol causing hallucinations[1]. The old person attributed it to having some kind of a super power of seeing the dead because their heart stopped for a few seconds and they were "dead". I am not sure if this is the article I posted, but there was such a case.

People who research their own specific illness may know better than some doctors, really.

Sorry if I was a tad off-topic.

[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3295654/

> I do not see any issues telling doctors that they are prescribing the wrong medication because you may know it better than them.

If you're not comfortable with a medication's potential side effects I do think it's worth bringing it up (I'm not a doctor, just an occasional patient) but it really comes down to how it's phrased.

There's a very big difference between:

"Actually propranolol won't work for me because after Googling around I discovered one of the side effects makes me feel anxious, can we try atenolol or nebivolol based on what I read online?"

vs.

"I know Googling for medical advice is sketchy but I read one of the side effects of propranolol is brain fog and this really concerns me because my job remains that I stay sharp. I trust your diagnosis and I know just because a side effect is listed I'm not guaranteed to have it, but are there any other beta blockers that we can start with that don't have this side effect and will still work?

There's a lot of very carefully selected phrases in the 2nd way of saying the same statement.

- Admission that Googling for medical issues isn't usually a good idea on its own to take actionable advice on

- You bring up a valid concern about one of the side effects

- Saying "diagnosis" vs "advice" is respectful of their decision ("advice" and "opinion" are a bit demeaning because it discredits their decision)

- You call out that you know not all side effects will come true (you come off as someone who isn't just blindly Googling stuff, this raises your credibility)

- You name drop "beta blockers" which shows you've done a decent amount of research

- You let them take the reigns and offer alternative medication, giving them a chance to maybe pick the ones you've researched

- You close things out as being optimistic by saying "and will still work" rather than questioning their decision before they give it by saying "think it will work"

If I said the 2nd wording to a doctor and they denied my request without even discussing it then I'd leave and look for a 2nd opinion. Asking people questions without demeaning them is a great way to get an honest look at how someone operates.

You're offering reasonable practical advice, I think. But imagine if this were about consulting an auto mechanic. It's like "You can't just question them, you have to suck up!" It reminds me of a small Apple developer's open letter to Apple a while back, clearly very carefully worded, full of protestations about what a big longtime fan they were of Apple.

The difference in both cases is the gatekeeping.

This article was posted before several years ago. The whole premise is bumptious - "I can copy data out of a bunch of papers [which I am in no position to screen for quality or relevance], run a canned 'gold standard' analysis in R [the idea that there is one true way to generate valid data is ridiculous], and then go tell the professionals what they are doing wrong." He even brags that his meta-analysis for depression had more papers than the published one, as if this was a valid metric. The Cipriani meta-analysis he cites was publised in February 2018. His meta-analysis was done in July 2018, and had 324 more papers - what explains this difference, other than obviously sloppy methodology. A proper meta-analysis is a lot of work, researchers spend years on one meta-analysis. The whole concept is ill conceived, and the author is too caught up in themselves to even realise why.

Meta-analyses are a good idea, but the mere presence of a meta-analysis does not denote a useful undertaking. The literature is polluted with thousands of meta-analyses. As far as I can see this is mainly because there is software available which lets almost anyone do it, and once someone else has done a meta-analysis it is much easier to do another one because they have already found all the papers for you. The publication rate of meta-analyses far outstrips the publication rate of all papers, and shows some unusual geographic variation (Fig 2) [1].

[1] https://systematicreviewsjournal.biomedcentral.com/articles/...

Statistically speaking, isn't it sound to throw all the papers into the mulcher and see what comes out the other end? We do use the term "outliers" a lot in statistics, do we not? I understand that the quality might not be up to snuff for some, but won't the law of averages take care of that?

Have you ever used a mulcher to chop up some yard waste, only to accidentally put in some dog shit, and then the whole thing stinks to high heaven?

In all seriousness, with meta-analyses it's still "garbage in, garbage out". It only takes one or a few egregiously bad studies to throw off your results if that study has a large sample size but something fundamentally wrong with its methodology or implementation.

This assumes all papers are of equal quality, peer-review and accuracy of results. Which we know they are not. Some studies should have more weight than others. Which has been mentioned in a previous comment; there is no 'right' answer, just a variety of ways to allocate different weights to papers based on various metrics.

You misinterpret the law of large numbers. What the law says is that if you have a large amount of samples, and assuming there's no pervasive bias in the samples, then any large enough sample (and often that's much smaller than you think - the classic example being election voters, with a group of only a few thousand representative voters being enough to predict the outcome of an election over a large country with millions of voters) will look identical to any other... that is, over a large enough sample, in the case of this article, the conclusion of many papers should converge to the same answer, with outliers being marked out as likely "bad" papers.

The only assumption you may reject here is that there's no systematic bias in the papers. Perhaps there is... or perhaps most papers are just very unreliable, in which case there should also be no convergence... but if you find convergence, there's a good chance the result is "real".

But the crucial bit here is the "large" in "large numbers". I expect that even for quite popular drugs the number of studies are maybe in the hundreds, which depending on statistics could well be quite a way from large enough. In particular if a significant fraction are crap studies.

I've dealt with enough types of data that I feel super skeptical that you can just dump numbers from hundreds of studies into some data store programmatically, do statistical calculations, and get valid results. It's very difficult to believe that there aren't a ton of variations in how the data is gathered, filtered, and presented that need to be accounted for before any comparisons can be made. I'm not going to trust the law of averages to negate the effect of completely out of whack data when peoples' health is on the line.

You get some numbers, they look good - fine, but at best it’s grounds for a proper study, at worst wildly misleading. You can easily fool yourself with statistics, and other people too.

For a good read about studies with solid statistics and bogus results, see [0].

[0] https://slatestarcodex.com/2014/04/28/the-control-group-is-o...

You mean the Law of Large Numbers (LLN), not the Law of Averages, right? Both the Weak LLN and the Strong LLN presume all samples are independent and identically distributed. If we make a hierarchical model on the data of each paper, we can bind all the data into a single distribution, but assuming that each of these studies is independent is a _long_ shot. WLLN and SLLN _only_ apply to, roughly, sampling from the same process. Its scope is more applicable to things like sensor readings.

The Law of Large Numbers is an actual math theorem. The Law of Averages is a non-technical name for various informal reasoning strategies, some fallacious (like the gamblers fallacy), but mostly just types of estimation that are justified by more formal probability theory.

More generally, see "concentration of measure".

https://en.wikipedia.org/wiki/Concentration_of_measure

With all the negative pushback this is getting, it’s making me think he was onto something. The exact same criticisms would apply to Airbnb, for example. “They have not the slightest idea how the hotel industry works. This is a very professional industry with a lot of legal hurdles...”

Well, Airbnb and Uber aren’t the best examples, are they? Their growth and “success” is fueled by either operating in a legal gray zone, or defying the local regulations all together. Many people all over the world think their lives were made much worse since Airbnb is negatively affecting the long-term rental market.

Point is, the effect of the company on the society can’t just be measured by market cap.

Back to the original article, the author was using statistical analysis to provide medical advice. Now, it’s incredibly easy to arrive to false conclusions with statistics. That’s why there’s regulations, peer reviews etc. What if the “Egyptian contractors” screwed the data up. Was the founder qualified to spot an issue?

I think they’re ideal examples. Market cap is pretty much everything. It affects the world more than morals do.

HN has drifted further and further from reality, which has been very strange to watch. The classic example was someone dismissing Dropbox when they first launched, but now it’s turned into dismissing billion dollar companies after they’ve clearly won.

Only the parent doesn't dismiss "billion dollar companies", they dismiss comparing them to shut down, non winning, companies like in TFA.

Ah yes, one failure = the idea is horrible. Another classic trope.

Well, the idea was rejected by patients, advertising revenue, and doctors...

I also note the "weasel word" idea. This wasn't just an idea, but an implementation.

The same thing might make sense as a value-added feature in a more comprehensive health service (so the "idea" might be good when put in that use).

But as an idea for a service based entirely on it, it failed hard. What exactly twist do you have in mind to save it? Or are you just saying "we'll never be sure" with more words?

If you study a lot of history, you start to notice that old ideas are bad until they’re suddenly very good. Cannons sucked for a long time, till Napoleon showed they weren’t so bad.

I think posting haughty words is a lot easier than trying to make something work.

>If you study a lot of history, you start to notice that old ideas are bad until they’re suddenly very good.

Also the opposite: dumb ideas are tried again and again to no avail ever.

Adding as example of dumb ideas, people constantly trying to patent perpetual motion machines.

Nitpic on Napoleon, cannons were used centuries before him in sieges. People knew they were good all along.

You’re constantly steering the conversation somewhere else, aren’t you?

The meta-analysis idea wasn’t terrible. It’s just that there’re many assumptions in a statistical sense, the founder might not be the right person to implement it and he might have targeted the wrong market. Some people are under the impression that everything can be solved just by build an app. However, some fields are much more complicated than your gig economy food delivery.

Also known as “having a conversation.” I’m not sure why you don’t see the bad faith in your words, but I have no interest in talking more. Goodnight.

Wasn't the someone who dismissed Dropbox when they first launched Steve Jobs?

How is that a "classic example" of drifting away from reality? "I don't think this startup business that doesn't seem likely to succeed will succeed" is a comment that looks funny in hindsight, that's all

Arguably one reason those two businesses were successful in areas with entrenched players and business practices was because they handle the money. If AirBnB was asking either travelers or hosts to pay $X to be on a recommendation site, probably very few people would. There's always a cheaper competitor when you're selling information. Because you book through AirBnB, for a service which is relatively expensive, they can skim off quite a lot of money in an opaque way.

> Their growth and “success” is fueled by either operating in a legal gray zone, or defying the local regulations all together.

We just got an impossible vaccine in under a year, I'm happy for all of medicine to spend a bit of time in a "legal gray zone" to see what might happen.

> Many people all over the world think their lives were made much worse since Airbnb is negatively affecting the long-term rental market.

Absolutely, just like the Luddites (Although they would be well off Luddites) their world is worse. But humanity has been made far better.

Aren't you falling for the survivorship bias trap? Sure, people have said that about Airbnb. But I think there are loads of that startups that failed because they didn't understand the industry they were in, or because of legal hurdles.

I think there's probably a lot more to be learned about business success by studying business failures. There's certainly a load more source material.

Saying that there was a nuanced difference in what these two companies tried to accomplish would be a gross understatement.

Just because people think an idea is bad is doesn't mean it's a billion dollar startup idea. Indeed, most ideas people think are bad are actually bad - it's only the few outliers that are actually successful. Even then I think there is a ton of mythologizing around this idea that the founders were able to see something nobody else did, usually to make the founders look like some sort of diamond-in-the-rough geniuses, when in reality what they built was just a natural evolution of tech that existed at the time (successful founders usually just execute better and faster than others).

I mean, despite all stuff I've heard on HN about how a lot of big VCs passed on AirBnB, when I first heard of it it seemed like a very natural evolution from sites like Couch Surfing and VRBO that had existed for years.

At least the last part seems to be right, doesn't it? As a user I love Airbnb, way cheaper and better than hotels, epsecially with a family.

It was indeed a fun read. As a pharmacist I had a similar 'idea' years ago and got two other pharmacists excited. We were sick of seeing prescribers not follow 'Evidence based medicine' and thought maybe something that took Cochrane Meta Analyses and UpToDate info in a nice interface would be amazing.

> I started babbling about network meta-analyses, statistical power, and p-values, but he cut me off.

> "Yeah okay that's great but nobody cares about this math crap. You need doctors."

So true. As an insider in healthcare I would probably have disregarded this for idealism.

It also sonds a lot like more than one tech / data scientist I met. Applying statistics and tch to a domain they have zero experience in. Kind of giving calculators to 1st grades and hope they know whether or not the results are correct.

Just because someone spends 30 years learning something does not mean they learned the right thing.

Just because you didn't have 30 years of learning doesn't mean you know better.

> I'll just prescribe what I normally do, since I'm comfortable with it

This is actually something that drives me absolutely nuts about doctors in the UK (I presume they are the same elsewhere) - inertia.

It's like doctors leave medical school with "best practices" about what they should prescribe - like they are glorified, human decision trees - and then across their 40-year career, they never read a paper, never read any new guidance, and general never change.

Inertia seems to be a particular problem in the NHS, where doctors have a set list they are willing to prescribe. Why? Because it's what they've prescribed previously, so they are "comfortable" with it. You can see there is a bit of a "chicken and egg problem" with other medications.

There is also the 'NHS approved medication' list that they can only prescribe from, which restricts whats available to you, and it takes years for something newish to make it onto that list (for reasons...)

If you want better options, the only choice in the UK is private healthcare.

But yeah, I do still agree with you - GPs don't seem to even be up-to speed on the latest info with regard to a patients condition - surely their industry requires on-going professional learning like the rest of us ?

None of this is true.

It's not technically true, but there is a lot of truth in it.

There are lists of recommended treatments, and there are many restrictions on prescription of medications, expensive ones in particular.

At a national level, NICE[0] (or SMC[1] in Scotland) decide what medicines the NHS will pay for, and for what conditions. They determine what medications are licensed for different treatments. They also set guidance on what treatments should be used for different conditions, what the 1st line, 2nd line, 3rd line treatments should be, and what treatments they think should not be used.

At a regional level, there are "formulary" groups, which take the national guidance, and make some tweaks - for example, for monetary reasons, they often increase the restrictions on expensive medications, making them harder to obtain. An example of those restrictions might be that the patient has to have been suffering for longer, the impact of the condition has to be more severe, or the patient must have tried several other (cheaper) medications first.

For expensive medications, it's often not as simple as a consultant saying "I want my patient to have this" - at a regional or hospital level, there are quotas/limits on how many patients per year will be treated with sich expensive medications, and there are comittees that meet regularly to decide who is worthy.

Now, if we put aside those expensive medications for moment: yes, technically consultants are free to use their experience, knowledge and judgement, and prescribe what they see fit (even unlicensed drugs) - but in reality, within the NHS it is uncommon to prescribe outwith the regional formulary guidance, or/and outwith licensed medication uses. There are a number of reasons for this:

1. Consultants take on individual liability when they prescribe unlicensed medicines, and individual hospitals fear legal risk too

2. Consultants will have their balls broken by their department head when they prescribe expensive medications - dept heads have budget responsibility, and in turn they will have their balls broken by the board

3. Consultants regularly meet with others from their department to discuss difficult/unusual cases, and have to justify their decisions to the panel

4. Perhaps it's the workload, but (IME) NHS consultants often have outdated knowledge. I swear some have never so much as read a journal article since they left medical school 30 years previous :( Again IME, private consultants seem to be much more confident to think out of the box with treatments; perhaps it's the commercial competition?

[0] https://www.nice.org.uk [1] https://www.scottishmedicines.org.uk

At it's core "modern medicine" is not actually a science. We've added a veneer of science, there's science at the edges, but at it's most basic form it remains an art from inception in the 19th century to today.

* Pharmaceuticals: science, mostly, but beware of "pseucutcals" like supplements and herbal remedies. Also beware of new ideas that cut into profits.

* Cutting-edge surgery: science

* Mental health: art, with a dose of science from big pharma

* General practice: mostly the art of laying on hands, distilled experience, a bit of research with your sales rep and whatever you took with you from medical school.

If this feels harsh, remember that, despite solid scientific evidence, it took over two decades for the conventional medical wisdom to move stomach ulcers from "caused by stress you need to relax" to "caused by helicobacter pylori we can treat it with a convenient antibiotic." Inertia is a helluva drug.

I'm kind of concerned that people read this as satire but miss the most important part, namely the absence of concern about the safety and validity of the results. You know, the part where he stuck some statistical software in front of a database populated by a "motley crew" of contractors and wanted doctors to use it as a shortcut for making patient care decisions. The part where he implicitly compares the HTML spit out by his system to peer-reviewed work by professional researchers. The part where he is proud of "beating" a "record" for least discriminating meta-analysis.

Reading this story and talking about his marketing and product development process feels like watching Lovecraft Country and then then only talking about the time travel physics of it. There's something real and awful here, hopefully presented in a fictionalized or highly exaggerated form. The people in my social circles who mistrust tech and despise startup culture -- this is exactly how they see us.

But... this is peer-reviewed work by professional researchers[1]. He just tabulated them / made them searchable.

[1] Given the reproducibility crisis, I'm not sure we should trust these much either, but that's a discussion for another time.

Is his meta-analysis or its methodology peer-reviewed work by professional researchers? Has any suitably qualified professional reviewed his output?

No, he did, by his own words, a network meta-analysis[1] on the results which combined the results of many studies into a single table. However this was not a mere tabulation or summary of existing results but a fairly involved statistical aggregation of results. In scientific literature this would merit it's own publication and peer-review.

[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5386629/

They're already searchable. He's claiming to extract usable information from them in a way that can be instantly used by doctors, so that they don't have to read the studies or find an up-to-date summary of the literature by a skilled professional.

He even called what his software did a meta-analysis. My understanding is limited, but a meta-analysis is difficult and laborious to do correctly. You have to design the standards you use to include and exclude studies and examine each study to decide whether including it will improve or harm the quality of your results. For example, maybe a method to evaluate outcomes that used to be common has been discredited. Maybe a study was done under a principal researcher who has been caught fabricating results in other studies. Maybe a study claims to be double-blind, but when you read it carefully, it turns out that it isn't. Maybe a study is well-designed in every way, but it was designed for a slightly different purpose, so the data can't be used the way you want to.

Factors like these result in the exclusion of a lot of studies from a meta-analysis, after painstaking examination by researchers who have the expertise to design and run the studies they're reading. It's scientifically difficult and important work. The author is bragging about not doing this work:

> On July 2, 2018, GlacierMD powered the world's largest depression meta-analysis, using data from 846 trials, beating Cipriani's previous record of 522.

You are on point. For common diseases there are guidelines manually created and periodically updated by 40 doctors. Doctors should just use the last updated protocol for easy decisions.

Even though for headache this is the recommended first line medication:

Ibuprofen 400 mg, ASA 1000 mg, naproxen sodium 500–550 mg, acetaminophen 1000 mg

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4541429/#!po=0....

That's for migraines, not general headaches.

For both in case it’s acute...table 2 and 3 starts the same :) (I was lucky though, I didn’t check)

Yeah ... I was actually thinking he may have dodged a bullet because if he got any further with this he could have gotten on the end of a law suit or getting shutdown and / or fined by the FDA if things went the wrong way.

I'm kind of concerned that our gold standard in medical practice is "this needs to work for everybody". We should really be more intent on doing personalized medicine.

The reason it takes so long to become a physician is because they have to tailor treatments to specific patients. Guidelines are there to help them make decisions.

This article's thesis seems to be that medical professionals are not incentized to provide the best interventions, and as a result, wouldn't pay for this service. However, what the author fails to mention is the competitors in this space that are successful, such as Up To Date, which provides really high-quality research trial data: https://www.uptodate.com/home

Rather than building a product that informs medical professionals about effective interventions, I wonder if the creator would have had more success if he deeply explored what sources of information these medical professionals pay for now - do they pay for anything at all, such as UpToDate, and don't want to pay this because it's an additional expense? If the creator found which sources people are using, the creator could sell this database as a feature for these partners and widely disseminate this data through partner channels rather than creating a competing source of information. It seems to be a case of this being a good instance of a B2B2C model, where selling this service to other businesses that sell directly to medical professionals could be more viable than trying to sell directly to them.

Alternatively, if the creator wanted to sell to patients, rather than medical professionals, the blueprint here is all of the consumer reports companies, such as Wirecutter, which is one of the New York Times's most popular services. Here, again, a "Wirecutter for medical interventions" could be quite successful, and you could sell this service to media companies that provide consumer reports as a service that would bolster these companies.

It's bad the creator wasn't able to find traction, as getting more medical data into the hands of consumers could have a huge postive impact over time.

A hospital system is incentivized in some ways to get people discharged as soon as possible and ideally with a good outcome. A service like UpToDate is one of the tools that may facilitate that by providing valuable clinical pearls to facilitate decision making. So it's not surprising that UpToDate is largely paid for by large institutions and academic medical centers, and not as much by individual medical providers.

Yes, that's an interesting point. It seems like an individual medical provider is incentivized for repeated care where a large facility is incentivized for a fast outcome.

I don't think he had the correct buyer either. While medical professionals may be the user, for this type of service you need to be selling into hospitals or health systems. They have the incentive for their providers, collectively, to improve the quality of care.

I doubt UpToDate makes their bones off individual subscriptions. The real money to keep a company afloat is from b2b enterprise contracts.

I thought exactly the same when they were talking about trying to convince doctors to buy it - surely they don’t misunderstand their market that much though?

Yes, exactly - it's though big enterprise contracts that generate revenue, but they are hard to close as a startup. Selling this service to companies that already have those contracts is a faster way of creating value for those customers.

There are also services in non-us countries like "How to treat" https://www.ausdoc.com.au/howtotreat Which are relatively popular and used by doctors. Although it's already edited rather than raw data.

https://opensourcemed.com

It's a crudely built version of UpToDate from 2018 as it was on April 2018. Useful for 98% of the population still.

Edit: definitely works better on mobile, and the search needs to be fixed.. this isn't my website but a resource that I've been passed down/told about by medical students.

Wow, that's really interesting. Did UpToDate publish their entire content library as of April 2018 under an open access license?

There seems to be a ton of good data here that you could use to build some type of WebMD competitor.

Competition analysis should be number one on the "Before You Write A Line Of Code" list. As the author has found, the value proposition is also important, but looking at competitors offerings will often inform that too.

Seems like taking a moment to think about monetization before beginning would have saved some trouble.

When you run a service that provides information you either make money through ads or direct subscriptions. And if it's B, you better have an airtight value proposition and an intimate understanding of your competition.

I remember reading this the last time and it was posted and I still think the core failure is that the author didn't actually recognize the issue he was solving. He thought the problem was choosing the most effective medicine when the real problem was decision fatigue looking at endless shelves of things that all seem to do the same thing. Those two problems sort of look the same, but the latter cannot be resolved by selling the tool to doctors in their offices.

The tool should have been designed (IMO) as a consumer tool, either a kiosk at CVS/Walgreens/pharmacies to assist with OTC med selection or possibly as a website with ads/referrals. I would absolutely choose a pharmacy over another as a result of them having something to help through that process, especially when I have a headache.

I'm grateful for this comment because it put into words the thing I couldn't.

I'm reading the epiphany part of this post, to quote:

You have a mind-shattering headache. You're standing in the aisle of your local CVS, massaging your temples while scanning the shelves for something—anything—to make make the pain stop.

What do you reach for? Tylenol? Advil? Aleve?

Most people, I imagine, grab whatever's cheapest, or closest, or whatever they always use. But if you're scrupulous enough to ask Google for the best painkiller, here's how your friendly neighborhood tech behemoth would answer:

[Screenshot of Google Search Results]

Oh thanks Google that's just all of them.

---end quote---

The author immediately identifies that this isn't a real problem, by their own admissions that "Most people, I imagine, grab whatever's cheapest, or closest, or whatever they always use." Yea, most people when they have a headache and know that most painkillers on the market will result in about the same degree of relief, don't bother to cross reference a medical meta-analysis, because they have a headache and if the $0.01 worth of aspirin doesn't make it feel better they will just take a second pill and eat the penny.

I like the author's conclusion about how to quickly validate business ideas, but even in the title the author still holds firm to the belief that this was a "fantastic startup idea" even though reality seems to think otherwise. Was this such a great idea, do most consumers actually want to review a meta-analysis when picking their OTC medicine, or do most people just try a few things, get influenced by advertising, and purchase the most reasonably priced medicine they think will help. I am just a single data point, but I don't normally feel naked and unscrupulous when I just read the symptoms that a medicine treats and pick one, and that strategy generally works just fine.

Solution in search of a problem and also in search of humans that act in this weird atypical fashion.

Waiit - "You have a mind-shattering headache. [..] What do you reach for? " - isn't this where you're supposed or even required to ask the pharmacist? You know, the person who has the years-long training to know drug effects, limitations and interactions with other drugs?

I'd say yes but the last time I was truly sick (flu) my roommate went to the store and asked the pharmacist what to give me and came back home with a bunch of homeopathic sugar pills. My trust in other humans is pretty low for this sort of stuff.

That's pretty awful. On the flipside, here in Montreal I've been very impressed with my local pharmacists at the nearby chain drugstore. They pay attention to drug interactions to the extent of flagging a risk the doctor missed but in retrospect admitted is valid, give me a consult every time I get a new medicine, remember things well enough that I was once asked how my wife was doing with her own recovery from something, and so on.

I used to live in the US and I'm certainly not used to that either. I wish that kind of great pharmacist care were more widespread as a norm.

That's sad

In many countries the pharmacist is a doctor and they are there to handle simple cases. Thailand, Singapore, Malaysia all seem this way. VS the USA (and Japan?) where all the pharmacist does is handout the medicine some other doctor prescribed.

For some reason, people seem to have evolved to consider nicotine products (for instance) more opposed to the abstract mission of a drugstore than homeopathic products.

I have never smoked or vaped or anything, but it's the homeopathic OTC meds that viscerally upset me, that they should be allowed on the shelves.

Isn't there a social benefit in letting the 10% (or whatever) of people that believe in them get a safe and effective placebo, for many conditions where that's all that's needed?

I see belief in the "placebo effect" as a mind virus. Because it is not just a justification for lying to patients, but entails medical people lying to themselves.

I get tired trying to explain, and if I'm not convincing anyone, maybe it's me who doesn't understand...but, I feel like the key is to ask yourself, if the placebo effect is something that you can scientifically demonstrate, how would you arrange a control group?

Failure to identify a reason for an apparent effect cannot be turned into proof that "nothing" has an effect. It's just a mental short circuit that people get trapped in.

> if the placebo effect is something that you can scientifically demonstrate, how would you arrange a control group?

Easy, just don't give the control group any medicine. Give the other group a placebo. If outcomes are better for the second group, it's evidence that placebos work, just as clearly as the usual trial provides evidence that medicines work.

> most people when they have a headache and know that most painkillers on the market will result in about the same degree of relief

I've not had this experience. For me, Ibuprofen works, aspirin and acetaminophen have zero noticeable affect. It's been that way my entire life. The other two might have some affect on fever? (no idea) but none on pain, at least for me.

Then for you, a meta-analysis would be even more useless, as your own experience is much more valuable.

Also, the author completely missed the fact that different people react differently to different treatments.

It's basically akin to looking at Google Analytics and trying to predict what your next website visitor will do. You'll have an idea of what someone's going to do, but you can't perfectly predict it.

Also, people feel differently about different treatments, for particular reasons. If your mom always took one brand of Tylenol when she had a cold, taking it might reassure you more than the theoretically optimal painkiller. Customers were quite happy to pay more for the exact same Ibuprofen labelled with 'Back Pain' or 'Period Pain', because they felt it worked.

If you think about it, the author's desire to choose 'the scientifically proven to be the most effective' is just another example of such a superstition. If you could convince him that a particular medication was the winner in the meta-analysis, he would probably objectively feel better, even if it wasn't actually true.

Yes, for headaches no one wants to read a meta-analysis, they just want to buy something quickly and feel slightly better.

On the other hand, for depression medication, they don't want their doctor to look at an online tool and choose the most effective antidepressant. They want their doctor to look at THEM and say "hmm, we'll try you on X but if it doesn't help with the intrusive thoughts we'll maybe switch it out to Y and up the dosage of Z". Or they want to tell the doctor what they think they want to be prescribed. They are paying big bucks to see a psychiatrist. Most of them are not on a self-optimization trip, they just want to feel better, and also feel like someone takes an interest in how they are getting on. Using a snazzy tool would probably lead to the patient being less satisfied with the doctor's service, even if they have slightly better outcomes by whatever questionable metric is in the study.

> The tool should have been designed (IMO) as a consumer tool, either a kiosk at CVS/Walgreens/pharmacies to assist with OTC med selection or possibly as a website with ads/referrals.

Sure but... would you really need any clinical data to make this valuable? A pharmacy could just make a decision tree with some common symptoms and do the same thing based on their own recommendations. Including data from studies is a neat enhancement but not necessary IMO.

I think perhaps there's greater value in the platform or methodology the author built for doing this type of meta analysis for medication. Could it be applied to other research fields I wonder. Could the software be licensed?

> A pharmacy could just make a decision tree with some common symptoms and do the same thing based on their own recommendations.

Sure, and I’d go to a pharmacy who did this based off of science instead of whoever paid for an end cap advertisement that month.

My million dollar medicine idea: print the dosage info in huge type on the bottle (e.g. "Take 2 every 8 hours") with more detailed info in smaller type on the back side of a peel away label.

The only time I use headache medicine is when I have a headache and that is the exact time I don't want to be trying to read a huge block of 6 point type to see how many I should take. (A couple of times in the past I've written the dosage in Sharpie on the piece of masking tape that I wrap around the bottle but I'd rather it was done for me as I'm lazy)

I'd be skeptical that would work.

A friend did a ton of user testing of improvements to a price-comparison site back when those were the rage. With some frequency the engineers would come up with a way to help people make a better buying decision. E.g., picking a TV is a problem, so they'd make a wizard that would ask you questions and then give you a recommendation. Problem solved, right?

Alas, no. Turns out most of these user guidance things wouldn't help, because people had no reason to trust the thing. They might go through the process, but their behavior didn't change. I'd expect to see the same effect with a kiosk. Most wouldn't engage, and those who did wouldn't weight the recommendation very highly.

A website would have an even deeper trust problem, and would add an SEO problem on top. Imagine a referral is worth $1 on average and you spend $0.50 on coming up with good answers, $0.49 on making sure you're on the first page of Google results, and take $0.01 in profit. You'll very quickly have a competitor with that spends $0 on research and $0.99 on being ahead of you on Google. Sure, their data will be garbage, but the page will be just as convincing to somebody who doesn't know anything, which is your target market.

I suspect the real outcome, as with many would-be startups, is that this is a feature, not a business. Somebody like Wirecutter or Consumer Reports could turn this into solid content that would be a nice addition to what they have already. They've already built a trust relationship with their users, and they don't have to specifically find people in the (very rare) moment of choosing a new medication.

It's true people don't trust new user guidance things, but if your product is useful you can build trust over time or not?

>The tool should have been designed (IMO) as a consumer tool, either a kiosk at CVS/Walgreens/pharmacies to assist with OTC med selection or possibly as a website with ads/referrals. I would absolutely choose a pharmacy over another as a result of them having something to help through that process, especially when I have a headache.

You're making a similar mistake to OP. You're considering the end user, the pharmacy, but are now ignoring manufacturers, suppliers and wholesalers. Why would a major supplier provide their goods to a pharmacy who ranks them last in their kiosks? Why wouldn't they give preferential pricing to pharmacists that give better rankings?

Perhaps large chain pharmacists have buying power here, but they don't need help attracting customers. Perhaps small pharmacists could benefit from this system to attract customers, but if the system ranks Pfizer products last they wont be in business for long.

Point is, perhaps you want it, and perhaps you think pharmacists should want it for you, but the entire supply chain does not want it.

This is the classic case of building a product that you hope will solve a problem instead of finding a problem first then building a product to solve it. The correct approach would have been to have those conversations with doctors before spending $40k to build the product.

I've made this very mistake myself but I was lucky enough to have enough runway to start over and talk to customers first then pivot the product to something that they actually need.

I call this the "I have an idea for a startup!" issue. You hear it all the time from family/friends. Where they tell you this great idea for a product they had. This is the wrong approach. What you want to say is something like "There is this really interesting problem that everyone in ecommerce is facing right now"

Uh no. He had a problem - how to choose a medicine - and built a product to solve that problem. The issue was that solving the problem isn't something people wish to pay for, not that the problem doesn't exist.

> Uh no. He had a problem - how to choose a medicine...

Did he though? He even mentions that customers generally pick whatever (because they don't really care what's the perfectly optimized headache pill they should take) and doctors thought it was neat, but no one really wanted to pay for it. Their solution (e.g. prescribing what they would typically prescribe) was good enough.

It doesn't really sound like there's a problem there.

Solving your own problem and then saying "Oh I found someone interested in my solution" isn't really what they mean by finding customers though.

I don't think most doctors or customers would even consider this to be a real problem. For some specific, rare conditions sure. But for Tylenol? Come on. No one is going to pay for that.

In other words, "solve a frequent, burning problem".

Tom Blomfield worte about how during YC S11 they had trouble growing at all. He chanced upon a customer who fit their product's user persona to a tee but at the end of a 20m conversation Tom realised the it wasn't really a burning problem for them at all.

And so they pivoted...to GoCardless.

https://archive.is/8IDcl

To be more accurate, you want to find a problem that people will pay (money, time, etc) for. He never tried to test that until it was too late. To be fair, this is a very, very common failure mode and I can imagine most people reading this story can identify with it.

If that is the problem, this is not a great solution.

I don't care if the medicine is perfectly optimal in terms of standard deviations above placebo (which I don't trust, since the data is not likely to be reliably comparable); I care if it sufficiently solves my issue with least hassle (side effects/access/cost/works for me personally). Placebo might just be fine, in some cases.

If I am not able to identify what medicine to take, I ask my doctor, who has years of training for that, already knows what Google sources are good, and can look it up in a reference manual.

> He had a problem - how to choose a medicine - and built a product to solve that problem.

No, he didn't have a problem.

He proposes that OTC is the target. The reality is, the problem is not very deep here. Without access to prescription medications, you basically have a handful of options that are more of less the same. It's like walking onto a rental car lot. You need to know what category you need, but everything in that category is pretty much the same.

He isn't solving anything for the physician market. His recommendations are very naive and lack the clinical backing to support actual usage. A physician isn't going to buck their go-to's because a random tool suggests something else. This tool needs to provide scientific backing for it's recommendations and why that recommendation is better.

Yes, this is a problem that a heavily regulated free market actually solved pretty well.

I go to my local grocery store and look at the painkillers they have. There are generic aspirin and Tylenol available for a price almost anyone can afford. There are branded versions which are still reasonably priced. There are reasonable variations such as with caffeine or with decongestant, again priced perfectly fairly. If you want to pay $2 extra for awesome marketing you can. In some stores you might find a 'natural' remedy - essentially a placebo, again priced ok.

Everything on sale is basically safe. Nothing has a business model that relies on addiction. Nothing costs hundreds of dollars. Nothing is adulterated or counterfeit. It's easy to get information about everything available. The vast majority people would be able to choose something to match their own needs (for example, they might be allergic to aspirin, or need to avoid Tylenol because of a kidney condition).

If you bought the worst painkiller on sale in the store, it would be fairly effective and not too expensive. The experience in most stores in large parts of the world would be very similar.

He had a problem, but then decided to solve a completely different problem. The process of an end user picking an OTC drug for a headache vs a doctor prescribing something are worlds apart. I'd personally love a site which compared all generic + branded medication available over the counter and ranked them based on some criteria. I'm sure there would be a ton of advertising and affiliate marketing potential for a successful one as well.

He still has the DB, so it should be possible to generate a static site based off of it with advertising.

Right, it also seems more monetizable than WebMD since it's more shopper-oriented, so ads / affiliate model should not be dismissed out of hand.

To be fair, when he started he didn't know he wanted to talk to doctors. That was after a pivot.

Also, $40k is downright cheap compared to most failed projects to improve medicine. If it were someone else's money, that is.

True, but he didn't talk to consumers before spending the $40K either.

Also I don't think $40K is cheap for an MVP. He could have used static HTML mock-ups, or even a few Power Point slides to see if his potential customers would have paid for that service.

The most important lesson in startups: sell it, then build it. Engineers have a bias for building, and tend to do it backwards.

Drug companies spend hundreds of millions of dollars showering doctors with incentives to prescribe their drugs. Additionally drug companies run ads where they literally tell potential customers what to "discuss" with their doctor. Honestly I think a day or so of research into how drug companies market their wares would of probably saved this guy $40k.

His work is interesting as a free product, maybe as a side project could of grown over time into a potential acquisition. Slapping a paint of React (or Svelte) on top of his Excel-driven POC and a "show Hacker News" post might of been a better way to go.

However kudos to him for not only trying to pursue the idea but also for the write up. It's easy for me to armchair quarterback this one, especially since my father was a doctor who got hammered weekly by drug companies, but I'm sure his blood/sweat/tears will pay off in other ways.

The current ideology of needing to go big fast kills a lot of decent ideas.

Everyone thinks there are two modes for a service: dead or worth $1B+.

Not to like bash you while you are down but you may have been able to do this while working and not spending $40k on contractors.

Not everyone needs to raise $10M, get a flashy office, employ 200 people and have all of the trappings of a "successful" startup founder; if anything, that's antithetical to what you should be doing.

No MVP, no testing their model, just straight to "Next Best Thing". It definitely hurts, I've been there too for my first "next big thing" and I doubt you'll be the last to get burnt by the image VC firms sell.

$40K is not much to put into a business. You'd likely be putting down more than that just to open a convenience store.

Sure you need to spend a lot of money to grow a $1B+ business, but the same is true for a $100M business or a $10M business and even most $1M businesses.

$40k is a lot to put in a business you have no idea if it can make money or whether it is needed.

It's a definite that people have to eat. I would assume most failures in restaurants and convenience stores are operations, not a critically flawed business model.

---

But yes, I guess $40k is less than $2M, but it also takes $40k away from another venture and possibly having to go back to a day job to take another run at something.

All because you didn't validate the problem before cutting off your income source.

You would not need to put up a lot of money if you just came to a market to sell some vegetables. For less central areas you could potentially sell stuff for free. Convenience store is a bad example as these are heavily standardised by now: we all expect a high diversity of products and thus the equipment needed to store it

$40K is not much to put into a business but it's a lot to spend before you validate your idea.

40K is a lot to spend on an idea you didn't even try to sell or check how to make money on it.

I thought I read this here before: https://news.ycombinator.com/item?id=21947551

I still feel the same about it[1]:

> Don't shut down the website. Keep it running and make the data available in read only mode and add a donation button (alternative to ads). It should cost $5 a month to host on linode or digitalocean. Don't let the money you spent go down the drain.

[1]: https://news.ycombinator.com/item?id=21951604

>Make something people want. It's Y-Combinator's motto and a maxim of aspiring internet entrepreneurs. The idea is that if you build something truly awesome, you'll figure out a way to make some money off of it. So I built something people wanted. Consumers wanted it, doctors wanted it, I wanted it. Where did I go wrong?

I'm not sure he still gets it. He didn't built something people want (even less so want enough to pay for).

He built something he merely thought people should want. In the end, people didn't really want it.

If it had succeeded and was ad based I'm certain I would have used it occasionally assuming I knew about it.

Some would like it, however the problem is that you wouldn't use it enough for it to be worth it to him to build it. Quote:

> Turns out the world's biggest health website makes about $0.50/year per user. That is...not enough money to bootstrap GlacierMD.

Does not always work like that. Did people „want“ cars when there were only horses. Or Facebook when there was just email. Sometimes you build something and only after people realise that they want built thing.

>Did people „want“ cars when there were only horses.

That's not the right question (as based on the well-known quote).

The right question is: if I showed people a car on actual paved roads, with its load capacity and speed benefits over horses, would they jump for it?

And they did.

But for this thing, they didn't.

This person didn't merely talk about his yet un-made app to some people who couldn't even imagine what could be.

He showcased it, working in full form, and different target audiences could not care less.

Thats not entirely true. You can't build car, paved roads, load capacity in one go, you got to make the first horse cart without the horse.. and many would object saying horse is literally free, why should i pay for fuel. where would i refuel wheres the horse can feed itself. horse can drive itself if necessary etc..

There is no single formula for success. Sometimes you are too early and you fail. Snake oil startups are getting billion dollar valuations while a legitimate idea may be rotting in another corner of the world. It's a combination of lot of things. Sometimes I feel many of the yc companies are successful not because people are jumping to buy the license, but because they have a network effect and are vetted by a community and perceive it as the better option while it may not be such. Connections, Money, Advices, Talent, Idea, Marketing, Sales, Management - every single thing is part of the puzzle.

Somtimes you are lucky to get the market fit right of the bat, but many times you have to try hard and struggle a lot before you even have a chance to have a glimpse of market fit. In the case of the author of the blog, if he had build a hype around his product like this is the next biggest thing and got a network effect going (your peer doctors are using it) it may have turned out way different. He wouldn't need to charge it initially or have a free plan that is good enough to get them hooked etc. Don't know there are thousands of variations possible. He simply stopped too early IMO.

You're looking at the solution. Not validating the problem.

Think about the problem horses/cars were solving. It's clear people 'wanted' that problem being solved just by how prevalent horse-as-utility were.

Yes, they wanted cars before the cars got invented. The wanted greater speed, more comfort, less trouble. People didn't care whether it took the shape of a "car" or of a "sdflkjsdkj". As long as the invention satisfied their needs, people were willing to pay for it. This stands true always.

This was a great read!

Doctors definitely don't need your product, as they pointed out.

But consumers do!

I think you should have sold to consumers.

Having three doctors in my close family, I often found that whatever the tired doctor at the hospital is recommending is in contrast with what a doctor who care about my health (and is willing to look on their proprietary platforms) is recommending. That's why you ask different private doctors' opinions when things matter.

I found myself googling studies on what's the better treatment for $x and I would pay for better-than-a-doctor-advice backed by actual studies.

I think the problem, in your situation, is the scale of the operation.

If you want to do B2C, you often need investments and you can expect to become profitable in a few years.

Either find some funding or fix your cashflow.

That was my reaction too. It seems like GlacierMD might have been successful if it had focused on SEO. Whenever someone types "what is the best headache medicine", GlacierMD should have been at the top of the results. The landing page would show the results of the studies for free, but would also let the user sign up so they can enter allergy and preference info and customize the results to them. Nearby doctors would advertise on the site.

So, in essence, before spending any money, I would go out and interview doctors to see if they're willing to pay for a channel of new customers. I would also try to gauge how many customers would be willing to sign up and enter their info so local doctors could advertise to them. And, of course, I would check with a lawyer to ensure I'm not bumping into HIPAA.

Yeah I agree with your comment. Also, he could host it somewhere cheaper after launch and just keep the site up. Getting high up in the SEO isn't really built over the course of one day.

I am sure he could self-host it or rent a cheap server somewhere. AWS is crazy expensive so it seems kind of unnecessary to host it there if you are low on cash. I host a site with 40 TB worth of images and my bill is probably way less than his was.

I think the data could be beneficial to software vendors, maybe create open source dev kits, so they can pull in raw data or graphs.

Then doctors get the product without having to sell to doctors. they are less tech savvy than a software vendor, and they also don't want more tools to have to use technology wise.

It sounds like GlacierMD not only presented results from meta-analyses, they made a platform to make them easier to do and did their own but didn't publish them. If you built a platform to make doing meta-analysis easier, you could collaborate with medical researchers and institutions who are publishing meta-analyses. Maybe give it to them for free at first, maybe get an author credit, some PR, and more sales leads. Researchers can probably use grant money to buy access to your platform and still save money on paying analysts to do things manually.

Precisely what I was thinking - a tool to analyze medical studies, compare, and provide easily digestible results is a tool for pharma, not providers.

> So I built something people wanted. Consumers wanted it, doctors wanted it, I wanted it. Where did I go wrong?

A lot of people in this thread are talking about the woes of ethical doctoring, the problems of healthcare, and so on. The reality here is that no one wanted it -- at least no one that he was selling to. I can relate. I've been there -- building something for months and then hearing crickets. The Y Combinator motto -- "Make something people want" -- is actually pretty tough to get right. In fact, it's the crux of any viable business. The reason why 9/10 businesses fail is largely in part due to the fact that they aren't making something people want.

Prospective customer interviews suck (especially for technical folks), but they are integral to building a product people want. Doing them before an MVP is crucial.

>I had literally nothing to say to that. It had been a bit of a working assumption of mine over the past few weeks that if you could improve the health of the patients then, you know, the doctors or the hospitals or whatever would pay for that. There was this giant thing called healthcare right, and its main purpose is improving health—trillions of dollars are spent trying to do this. So if I built a thing that improves health someone should pay me, right?

I think the author makes the wrong assumption that his product does improve health and that the current practices of GPs or other health professionals don't (as much as they could). He misses the most important thing of any professional and successful health care worker: experience.

If a doctor has a high success rate with drug A on their patients then even if some website tells them that drug B _could_ be slightly better why would they try it? There has to be more of an incentive than _just a website saying so_. The doctor has a high success rate with drug A, has seen it working many times and knows everything about side effects or maybe the lack of those. If nobody has complained and it works miracles then it would be foolish for the doctor to start prescribing a new drug simply based on a random website's recommendation. Also doctors do have to stay current and they get medical journals sent home, attend training courses and conferences and learn about new stuff when it's actually relevant and needed.

> So in July 2018, nine months and $40K after starting GlacierMD, I shut it down. [...] The idea is that if you build something truly awesome [...] Where did I go wrong?

$40k and 9 months is not that much. Either you gave up just before you figure out how to make the business work or maybe GlacierMD wasn't that revolutionary.

Why not open sourcing and publishing all that work? You can still recoup some money via ad money even if it's not the millions you were expecting and at least you have something to show off.

> "Hmmmm," she said, picking at her fingernails. "Not directly. Of course I always have the best interests of my patients in mind, but, you know, it's not like they'll pay more if I prescribe Lexapro instead of Zoloft. They won't come back more often or refer more friends. So I'd sorta just be, like, donating this money if I paid you for this thing, right?"

this is the problem. the incentives in healthcare are messed up. doctors are paid for their time, not for their outcomes. if a patient comes in and is prescribed a therapy, and they don't have to come back, the doctor should receive more than if the patient returns because the therapy had an issue.

Maybe we should also pay programmers based on outcome then, and you get less if your % of bugs is higher than the team average, sounds like a great idea, eh? Maybe a day without pay if you bring down the live service? I'm pretty sure this wouldn't work. The first place to put it in place, be it a hospital or a tech company, would just have most people quit. Sometimes things are hard and people need to know the business has got their backs as long as they try their best. Reducing pay on bad outcomes is institutional blame culture on overdrive.

If you paid doctors by individual outcome then the incentives would be even more messed up, since no-one would want the tough or ambiguous cases.

This is opposite of the desired outcome, which is that the best doctors see the hardest cases.

Always pay professionals for their time. Patients are not widgets.

I was horrified when I read that line too. I have this crazy idea that most doctors go into the medical profession because they want to help people - her attitude is awful, IMO.

Between the lines, the real doctor perspective here is "Medicine A, which I have been seeing success with for decades, is (according to this site) less effective than Medicine B. I really trust Medicine A, so I don't think this is enough of a nudge to make me change my mind necessarily."

I would have absolutely used this -- as a consumer tool. I _would_ even pay a (small) subscription for the service, as I'm an avid user of WebMD who is more often than not overwhelmed by potential treatments for any given symptom, but I admit that most likely would not scale.

I do, however, think it makes much more sense as a consumer tool rather than a doctor tool: if I've already gone through the trouble of making an appointment at a doctor, who I trust, I've either given up on trying to resolve the issue myself and will trust their prescription, or I have no idea how to approach the issue or symptoms I'm feeling as communicated by WebMD. Your value proposition is saving the consumer time and effort of booking an appointment to find out how to best fix a minor-medical issue they (or, say, their child) may have. Is that a sustainable proposition? I don't know, but it makes much more sense to me than as a tool to tell doctors how to do their job "better".

I'd imagine if you went the WebMD consumer route, too, once you reach a certain user-size, you'd have no problem partnering with pharmacies (or even pharmaceutical companies) to offer coupons for some OTC medicines in exchange for a referral fee, as is standard in the industry (a. la. GoodRx). Then you're providing value to the consumer in the form of recommending useful medicines, providing the consumer a way to get that medicine for a lower cost, saving them time and money spent at the doctor's office, and providing the pharmacy/pharmaceutical company a new customer.

I like your idea, but I’m confused why you didn’t try the webmd competitor route.. 50c/user that webmd makes is pretty good really , with good SEO + content marketing you could scale to enough users to make things interesting

Exactly. This was the right path in this case. The value is for the consumer, if he could attract just 100K visitors per month, that's $500,000/year.

Not only that, he would have a highly targeted group of users that he could tailor all kinds of products and services to.

You need 1mil visitors for 500k.

Indeed. Which what 100k per month comes to, give or take :)

You’re both wrong. 100k visitors per month ≠ 100k unique users per month. Very different things!

You're being super nit-picky. It should have been assumed 100K "visitors" meant uniques, but I knew someone like you might make this comment. 100K * 12 = 1,200,000 visitors. I added a 20% margin to account for return visitors.

I’d like to see you tell your founder or CMO that the difference between monthly visitors and monthly unique visitors is “super nit-picky”.

The 20% margin is risible and those who live by SEO die by SEO: https://cdixon.org/2011/03/05/seo-is-no-longer-a-viable-mark...

Well, quite! But then I pretty much assumed it was a shorthand amongst peers so (over) explanation perhaps wasn’t really necessary...

IMHO most articles about medical conditions should be banned or severely downranked by Google, as they create far more problems than they solve. Over the years I've freaked out several times because I googled some symptoms I had and of course was presented with a slew of SEO-optimized articles that associate them with serious diseases. The thing is, those articles are mostly written by marketing folks that rephrase stuff from Wikipedia or medical literature, but without knowing anything about the conditions they write about. The sole purpose of these articles is to increase the page rank of a website and generate page views that can be monetized. Of course it's like that with almost every topic on the Internet, it's still highly unethical though as it incites unwarranted, existential fears in people.

Presenting medical trials and statistics to an uninitiated reader will probably also cause more harm than benefit, so it's good he canned this idea.

I have very similar experience in very different field.

About 8 years ago, we've built a SAAS for purchasing material for companies using auctions. The idea was that if it will be super easy for companies to do reverse auctions and use them to buy material they will save a lot of money (proven concept in those who actually used it).

But what we learned after some time was it didn't matter. Because people in the companies responsible for buying material didn't really care for how much they buy it. They care that they call their favourite supplier have a nice chat and then order whatever they need. And those suppliers throw some "gifts/rewards" to the package here and there. And without appraisal of those people the business owner never said yes.

The thing was that those people who effectively decided if they will use it or not didn't benefited from it at all.

We haven't lost so much money on that, but the lesson learned was huge.

The author of this one quit too soon or his product, despite his posturing and promoting was poor. 1st off he approached an American Drs practice and they said it doesn’t pay for them to use it. He could have gone on to approach a different country, UK, Australia, Canada to name but a few hell France, Germany - the list of countries with a national Health service with KPIs for exactly this purpose.

Moral of this story is if you got a good product don’t quit just because America can’t see the value lol. The irony that the American healthcare system didn't value this overwhelming.

As a few people have pointed out now, the insurance companies would be the customer. If your software allows patents with disease X to recover 5% faster and they spend a billion on disease X claims each year your software instantly has 50m value to them. You should be able to find a bunch of examples of how much Americans spend on different medications a year, extrapolate the differences in recovery rates from correct and incorrect prescriptions to that value.

Additionally, I'm sure they audit Dr prescriptions, if you get the insurance companies to use your service in the audits the hospitals will also purchase it to pre-audit themselves. Perhaps the drug companies can mandate its use. If you can't get a business to want your product, make them need it :)

You say “ So I built something people wanted. Consumers wanted it, doctors wanted it, I wanted it. Where did I go wrong?”

The fact that the people giving you the money I.e doctors / hospitals didn’t want to pay for your product proves you built something no one wanted. Wanted enough that they were willing to pay you money for it. Or go out of the way to find the budget for it.

Or the incentive systems in the US are completely screwed up when it comes to healthcare.

I was going to point out Kaiser, an HMO that provides both insurance and services. It's reputation is mixed. HMOs like that, or maybe the VA, would be perfect customers because their interests are mode aligned. Private practice doctors, not so much.

Nah, everyone wanted it, but for free.

I think the only thing that went wrong is the budgeting, OP might have made it had he dumbed it down a little and made it 100x more simple (on the backend).

It's been mentioned a couple times on the thread but I'll third it: this post succinctly captures the problem statement of _The Mom Test_, which is a short book about how to structure your product/market fit work to cut the crap and lock in on a product that people will actually pay money for. It's good, more people should read it.

Seconded. Do not start a company before running it through the Mom Test.

I love the author's writing style. I found it kind of funny. But there's a lot he could've done better.

My top recommendation would've been to find someone in medicine as a co-founder. This could've helped him find flaws in the economic model faster. I still think his project has amazing potential, except that it's going to have to go through a couple of iterations. For example - what if he somehow integrated this information to make it so that it's easier to connect customers to future doctors and take a affiliate fee?

Also imagine if we had a better medical search engine. There's so much information in medicine that is still blocked off to consumers. The author mentions that consumers do not want to go through the details - maybe we can make something that can dumb it down for people and put a directory of doctors who can help explain further?

It felt like there was a missing chapter!

Consumers don't see the value at all.

Doctors see the value, but won't pay for it, because it doesn't change their bottom line.

Whose bottom line does this change then?

HMOs, insurance companies, organizations that want to improve the health of a population in aggregate. They'd pay money to have healthier people (on average) to provide the service to their docs (who see the value) to use for free.

Well, maybe; but I was hoping to see that theory validated in the last turn of the wheel.

Insurance companies are the real target here. If they can potentially manage to get doctors to prescribe more efficient medication, then they will end up needing to reimburse less and improve their bottom line.

Exactly. Some 23andme clones are aiming at insurance companies and national healthcare systems. I know at least two of them personally, with investments from big pharma.

If you convince say, NHS, that this service would make their medical practice more efficient they would totally buy it! For this you'd need to hire MDs and play lots of regulatory tricks. Plus, after that, probably private medical offices would follow the lead.

Companies doing genetics-based prediction are trying to fight the same uphill battle, and I think it's one worth fighting. So don't give up!

Me and my fiends went a very similar path with a different product. We made an Android and iOS app that would help learn programming by wring actual code. Users loved it. They were spending 10 minutes in the app on average! Some spent over 2 hours a day and would finish all the content in two days. Even though, nobody wanted to pay for it. We figured, we would never be able to bootstrap with ads alone so we had to focus on something else.

He needed a co-founder who has experience in health care. As a Clinical Pharmacist and a developer - this idea would automate part of my clinical pharmacist job and is viable. He needed to sell to hospitals and PBMs as a tool to automate formulary management.

Selling to those entities are really *really* difficult and there is no good distribution channel, but doable with an experienced person who sold them stuff before.

Previously discussed: https://news.ycombinator.com/item?id=21947551

Weird that the blog post is dated Oct 2020 but was apparently discussed on HN in Jan 2020...

IIRC blogs with fresher dates are ranked higher, or maybe people just click them more. Just the author practicing low-level SEO spamming, I guess.

Really interesting to read. We can all now be captain Hindsight, but no point in stating the obvious, as long as you and others reading it have learned from the story.

Although, I have to mention, reading the article and seeing mentions of MVP and no testing or concrete business plan, I remembered the book "Disciplined Entrepreneurship: 24 Steps to a Successful Startup"[1] and I loved how Bill Aulet in the book mentioned the MVP model and gave his own thought:

Instead of everyone rushing MVPs, he recommends to people to rather chase an MVBI, a minimum viable business idea, since as he says, the purpose of a business is to create value and capture it, since you are not running a charity.

I must say he convinced me, and changed my line of thinking, it does make sense when you think about it, you cannot build the product for the sake of the product.

I also like the quote "Real entrepreneurs fall in love not with their product, but with their business plan."

[1] - https://www.goodreads.com/book/show/18652777-disciplined-ent...

'"Oh, uh—hmmmm," she said. "I don't know if we can spare the budget here, to be honest. It's very fun...but I'm not sure if our practice can justify this cost."' - Susan (psychiatrist)

It just goes to show you how haphazard and lacking of scientific method the psychiatrist field is.

'"And that isn't worth something? Prescribing better treatments?"

"Hmmmm," she said, picking at her fingernails. "Not directly. Of course I always have the best interests of my patients in mind, but, you know, it's not like they'll pay more if I prescribe Lexapro instead of Zoloft. They won't come back more often or refer more friends. So I'd sorta just be, like, donating this money if I paid you for this thing, right?"' - Susan again

This is an absolutely disgusting response to me. They don't care because the costs of shitty treatment - medications that fuck a person up - are externalized, the practitioner/psychiatrist doesn't have to deal with whatever severe "side" effects the medications they'll experiment on a person with.

A few points come to mind:

1. Good thing he wasted $40k that he actually had rather than $400k he might have borrowed. Of course you could argue that "no risk no reward", but still.

2. The idea itself seems to stem from a weird US-consumer-centric bubble. More specifically

  * "Pain killers"? Those medications have different active ingredients with different effects, only part of which is some kind of pain (and not a single kind). Some are anti-inflammatory, some are analegsic etc. Those are really different things.
  
  * Different people have different reactions to medication, even within the same group of active ingredients. One is better served by choosing medication based on past experience is similar situations.

  * If it's not clear what you should be taking, you probably want to have a doctor examine the entirety of your symptoms and make recommendations.

  * If some kind of medication is consistently superior to another for almost all people and scenarios, the superior one will likely drown out the inferior one in terms of availability and public awareness. The popularity of multiple kinds of medication usually (?) indicates they are preferable in different scenarios.

3. When I first saw what the novelty was supposed to be, I was unimpressed. I mean, ok, it's not trivial to do statistical analysis in R - but it seems like anyone with some R and statistics experience can do it. It doesn't seem like there's a lot of secret sauce. There might still be commercial viability here, but it's at best risky.

4. What was the initial business plan? Revenue model? Who was supposed to pay for what? Specifically, suppose one person paid for a recommendation for pain medication. Ok, so they got it - but it's not going to be a secret. It's now public.

Healthcare is tough. There are few markets with more inertia and misaligned incentives than this.

With respect to pitching the product directly to clinicians, OP may have misunderstood his target market or overestimated the value proposition.

It seems like a very big request to ask a doctor to deviate from a treatment with which they already have a high level of experience.

I think that people are probably misleading themselves if they say they want it, but aren't willing to pay for it.

People fork over money for IAPs by the millions per minute, because they want the results.

I think if it were truly that desirable, people would have been happy to pay for it. I just don't think it provided that much value to the customer.

Mostly they're being polite (and showing they understand perfectly well how they could use it, so no need to explain its purpose any further thanks). They're not really misleading themselves...

Ultimately the author's problem is that he was showing Susan a clever project that looked much more attractive than notes she'd usually consult rather than asking her how many times a year she wasn't sure what to prescribe and how difficult it actually was for her to get a sound recommendation from a source she trusted.

Even if he was trying to demo a finished product rather than figuring out if it should exist, asking those questions would still have been more likely to put her in the frame of mind where she asks about or spots the really difficult to ascertain info that has professional liability repercussions rather than just liking the UI and being impressed that this programmer is citing the Hamilton Rating Scale for depression)

Original post from January 2020: https://news.ycombinator.com/item?id=21947551

> Make something people want

That's a charity.

> Make something people want __and will pay you for__

That's a business.

You learnt the hard way.

> Make something people want __and stuff it with ads__

That's a business.

IMHO "Make something people want" implies that people will want to pay you for it[1]. If people want it, they'll come asking you to take their money - if their response is "meh, I'd try it for free" then I'd argue it's misleading to tell that they want it, the most you could say is that they are interested. I mean, he demoed this stuff to people and none of them seemed to say "wait, can you give me access to this right now? Pretty please, I'll be your pilot test user, but I want to have this?" - I can't see any evidence that would justify the author claims "Consumers wanted it, doctors wanted it". The doctors called it "What a fun project!", the doctors called it "useful", but none of them wanted it the way they might want a cup of coffee i.e. being willing to give something at all to get the thing they want.

[1] the other limitation is whether they can afford to pay you despite wanting it; there are products/markets where the target audience is eager to pay but their collective budget is too small for the scale you need.

This reminds me of an idea I had days ago of using an AI to tell poisonous mushrooms from edible ones apart, from photos.

Then I realized I don't want to be responsible for anyone's death.

What if you limited the range from I don’t know if it’s poisonous to X% likely poisonous vs is/is not poisonous?

People empirically do not understand probability in a useful way. Folks will, in aggregate, read that as "X% poisonous", not "X% chance to be kill you."

"Oh, that's only 25%, it'll be fiiine."

Even if you can avoid being liable for it, I'd think most people wouldn't want someone's death hanging over their head for reasons beyond the legal.

Definitely a neat idea, but how would this work? Most identification starts with a spore print. What if instead of an AI to identify mushrooms it was software that used a classification system like what is described in "Mushrooms Demystified".

Edit: Should clarify most reading I have done on mushroom ID involves taking spore prints as an essential step but it's not always the rule.

Isn't the volume so low on that and the latency requirements so undemanding that ML is overkill to get the answer as well as the accuracy issues with false positive for "not deadly" being somewhat asymmetric?

I still think of ML as having a niche of "mostly right is ok" and useful for very, very large volume otherwise one or more people does better for less. That may change in time. Self driving cars are very close now they say, but i suppose we've been hearing that for a few years without getting there.

Already exists, I heard about it on NPR the other day (perhaps the idea got implanted into your subconscious) https://mushroomai.ml/#mushroomai

I don’t believe modern state of AI is good enough for the job. At least not yet.

A human expert might look at your photo and instead of saying “you may / should not eat that” start asking questions, about local ecosystem, climate and weather. Or they might ask you to shoot more photos, e.g. bottom of the cap or microscopic photo of spores.

I’m not an expert in AI but based on what I know they aren’t smart enough for that, at least not yet.

Yikes! Some species require microscopes to differentiate. Thanks for not making that app!

"Of course I always have the best interests of my patients in mind, but, you know, it's not like they'll pay more if I prescribe Lexapro instead of Zoloft. They won't come back more often or refer more friends. So I'd sorta just be, like, donating this money if I paid you for this thing, right?"

But you know who would pay more if they prescribe Lexapro? The manufacturer of Lexapro, and that's who he should have pitched this to. And they have the marketing muscle needed to make it scale.

That said, there's all sorts of conflict of interest/slippery slope issues going down that path, and I'm not sure they would buy this either, but they'd still have been his best bet.

He missed the obvious point: sell it to the pharma companies as a platform to promote their products. Guy could've been a zillionaire by now.

More and more developers these days give their work away for free and create a Patreon. I don't know if it would work in this case, but if enough people got wind of the idea, perhaps it would at least equalize. Not all brilliant ideas are meant to be businesses, so why not give users the options to support something they find useful? You can easily thermometer the demand this way, and if there's a [community] need that you are fulfilling, supporters will show up. You can always hide more advanced features behind the Patreon subscription levels, too. It sounds like what you are really after is a community that vets health knowledge, and this could be one cornerstone of that.

“Doppelganger has created value for the customer but not for the company.”

Tell that to “Celebs”, which’s sole app function is the described and makes ~$200K MRR.

That’s shocking to me. They have subscriptions? Why is someone subscribing to an app like this?

Looked at things on the App Store. It’s not clear what a subscription gets you but they’re charging $4.99 a week for the premium features.

Yeah, I was going to chime in and say "I bet Doppelganger would make more money than GlacierMD." Cynical content warning: Precisely because it's not geared toward helping others, which no one gives a shit about, and is all about your photo of yourself which everybody obsesses over.

40K is pretty cheap for an education these days. Especially a really top notch education of the real world, which this guy got.

Fantastic idea? This is AI snake-oil. If this guy literally thought he could build an auto-meta-analyses that could magically prescribe the right OTC to fix your pain based on a few button clicks, he's trying to play God.

It’s another one of the AI use cases where the problem is built around a solution.

The same url was discussed Jan of last year:

https://news.ycombinator.com/item?id=21947551

Why does the blog-post date say 'October 18, 2020'?

very odd, here is the original when it was published

https://web.archive.org/web/20200103170008/https://tjcx.me/p...

I have read this before but rereading it realized a more fundamental flaw. The author built an amazing tool that's actually worth a lot of money, but not for patients or doctors! The healthcare data analytics industry would kill for a version of this tool done well. Pharma companies and consultants would pay tens of thousands (if not more) to access a tool like this that's kept up to date. He could have sold the IP and rhe data and the protocols (or get acquired) if he'd networked a bit more in this side. It's probably still not too late!

Instead of marketing this as a tool to better select drugs, I would market this as a tool to better justify drug selections.

When I go to the doctor part of me is always skeptical - are they prescribing this because it is actually best for me, or did they just prescribe me the default and don't care if I have to schedule another appointment in 2 weeks to get a different prescription? Between insurance politics and the difficulties of keeping up with research, there are a lot of perverse incentives out there. I'm certainly not alone in this. Even if I knew for certain the Doctor's interests aligned perfectly with my own, the fact is I am a non-expert and putting my faith in something I am unfamiliar with and don't fully understand is scary.

Now imagine if a doctor could instantly put my fears to rest. When I start asking questions, he can instantly put a nice looking piece of paper in my hand telling me why his decision is the best option for me, personally. Maybe I don't totally get what all the numbers mean, but I see my readings, my symptoms, there's a graph on it and the drug prescribed to me has the biggest bar. If that were an option, I'd never use a doctor who didn't do that. Even if it didn't lead to more patient referrals, certainly it would save doctors time answering all the "but WebMD says..." questions, and a doctor's time is worth a lot.

I think it's a great idea but looking at the product I'd say that the value proposition isn't enough for anyone to buy a subscription. Maybe an ad supported website (like WebMD) that licensed professionals could check to validate their own assertions.

Also "$0.50/year per user" is actually a great amount if the site was made for a global audience.

And as an AWS cloud engineer I'm actually curious at the AWS cost breakdown. If done properly this kind of data processing and site hosting costs would be trivial.

Yeah, this is the kind of project that could fly really well as an "indie hacker" style. Just one person running it all. Today's cloud tools allow that.

edit: but I don't think that's what the author wanted. Not really sure what they wanted, to have full success by month #8? Drop it all for an idea not even tested yet sounds very very amateur.

Oh man! I'm laughing and crying at the same time. I had a very similar experience in my first "startup". It hurt me so bad, both financially and emotionally, that I could do another "startup" for 7 years.

Luckily, the next time I came across Steve Blank and YC Startup School materials and instead of burning all my cash over a two years I managed to FAIL my next three startups withing few weeks with a negligible cost.

Those books and articles saves your life and money!

This is yet another cautionary tale for founders.

I have to say, this sounds to me like trying to market an expert system for experienced woodworkers to choose the best chisel for a particular task based on the type of wood, size of the cut, etc. Practitioners have strong opinions regarding the implements they use; they use them in concert with other implements and modalities; and the small differences that may emerge from this kind of analysis are likely to be swallowed up by differences in copay or other things besides pure efficacy.

I was reading this as a licensed professional in a different field, and trying to imagine someone without my licensure trying to sell me an app that made my decisions for me. Let’s just say this was always going to be a tough sell, no matter how good the economics looked.

I loved reading this, but it doesn't seem like he thought this through all the way. For one you'd want to figure out how you plan to monetize anything you expect an roi on before you start spending all of your money. Two giving out medical advice feels like a giant cluster of potential headaches, and liability. A doctor who trust your website and then maybe one of her patients has a bad reaction, may have grounds to sue you particularly if you charge a fee to use the product.

Amusingly enough, I've got a document that's essentially a list of ideas where the business model is spurious at best and on that list is a solution almost identical to what was built here.

I appreciate that someone took the dive and then shared the process, pain and failure of seeing it through to its conclusion. The only thing missing is the attempt to raise VC money to scale it to some sort of expert system you can sell to enterprise customers.

Turning the product described in the article into an enterprise product sounds like nonsense to me. The product should arrive to the user that benefits the most. Isn't that people in general?

> So I built something people wanted. Consumers wanted it, doctors wanted it, I wanted it. Where did I go wrong?

There's a HIUUUGE difference between what people say they'll do and what they will actually do.

Yes. There's also a huge difference between what people want and what they are actually willing to pay for. I bet many of the people interviewd by the author would have be happy to use the product on occasion if it had been free.

I think this was posted already (a couple of years? ago)

Anyway, yeah, it wasn't a good value proposition (and is it just me or GlacierMD is a bad name for this)?

Because in the end, it doesn't matter much if you don't get "the absolutely best choice of drug". If it solves your problem great, if not, doctors can exchange it, but that's when the drug or dosage is bad, not "not great".

From Glacier, it's a very short mental route to Titanic. So, yeah I'd say naming your product that brings up thoughts of the Titanic is not great.

Iceberg != Glacier. AWS manages to market Glacier without people expecting S3 to sink to the bottom of the Atlantic

Good point, but AWS Glacier also has a very clear connection to the word since it's a cold storage service.

I would definitely look at antidepressants on GlacierMD, because I think I want to take one but the options and side effects are all over the place

yeah if it was something like "we can recommend a cheaper drug" then there might be an incentive... but also that's not how drugs are even priced. The healthcare world is such a mess. Charging doctor's offices for SaaS is very very tricky.

> But in fact everyone gave me some version of Susan's answer. "We just can't justify the cost," a pediatrician told me. "I'm not sure it's in the budget," said a primary care physician. "It's awesome," said a hospitalist. "You should try to sell this!" Ugh.

In B2B People pay either to reduce costs or drive sales. It’s as easy as that.

I didn't get it. He starts as if this is b2c kind of business, but then shows screenshots with service aimed at drug researchers.

If it is b2c: make an app for the phone.

Screen1: human body, where you can choose places where you feel unrest.

Screen2: describe what is wrong

Screen3: "based on your answers and your geolocation you can get these 3 remedies which are safe" or if something extreme is going on "please see the doctor, in the mean time you can use aspirin to relieve your pain"

*I usually call mom when I have headache, temperature or diarrhea cause I constantly forget which one should I take. I'm 30:-( I would use such app for these hints. For anything more serious I would go to doctor and wouldn't self medicate.

Monetization(Hard): Subscription for extra features like tracking of drug dosages and keeping track of your home stock. Would probably be beneficial to someone who takes a lot of drugs all the time.

Monetization2(Medium hard): deals with local pharmacies to be able to provide recommendations where to buy a specific drug advised by the app.

Based on screenshots what he built is b2b app for doctors?

one more thing 5 minutes later: I was travelling a lot past 4 years and would be great:

Buy international insurance within an app and have a list of local hospitals I can go to

Be able to "download" integration with local pharmacies like it was with maps.me These can be sold. If I'm in India and got Delhi Belly I would pay 2-5 USD to find nearest pharmacy/hospital or order drugs to my hotel

Sounds like $40k spent to learn a lot of the useful stuff you'd learn in a basic MBA program. Probably a better deal than the MBA :-)

I can fully relate to this experience. I think it's no coincidence that the businesses which are the most useful for society are the poorest. Their target customers have no money. Value producers tend to be honest people. They believe that they have to work to get money. That's why they can't get any. Those who have direct access to the money printers know this better than anyone. You need to be a fraudster to make money in this economy. I learned this the hard way working in open source. Now I work in finance. If you can't beat them, join them... Help accelerate the financial collapse.

Education, medicine, software which lets people work more efficiently and save time; they're a waste of time financially. The fiat monetary system wants more bureaucracy, more jobs, more surveillance, more manipulation, less efficiency, less free time for everyone.

Most people who made money in this scam system are fraudsters, attention hoarders, legitimised money launderers, risk externalizers...

Resonates with me big time. I also bootstrapped a digital health startup called AnywhereMD with a team of physicians. At first it was a triage app that evolved into a messaging platform. The triage had a similar value proposition; get people to the optimal point of care and they will avoid unnecessary and costly ER visits. Self insured employers and accountable care organizations would reap the savings and therefore pay per employee or insured per month and drive people to use AnywhereMD before deciding where to seek care. We were bootstrapped and had a hard time convincing large companies that we were legitimate since we were so small. The biggest parallel I can see is that in healthcare, cost savings and better outcomes are obscured, hard to see, delayed etc.... So it’s not easy to prove value and then capture some of the value for your self.

The graph about the curves (escitaloprame, sertraline) seems like a joke. The curves bending upwards indicating that everybody will be worse? Looking at the raw data (points in the graph), there is really no indication for that and if I were to see that I'd prefer the raw data.

Like other comments I think the author addressed the wrong market.

Essentially he developped a tool to facilitate meta-analyses, both building them and then presenting the results.

So, his customers should have been people who usually are investing in meta-analyses. Doctors are not investing, merely passive consumers of existing analyses they can find.

So, research teams should have been his target, probably mainly in academia and in public health bodies. He could have grasped public funds.

The right market was researchers, i.e. in real life usually the ones owning biology PhDs (and often professor chairs), and not the physicians with medicine PhDs, who are the engineers of the human body (engineers applying known solutions, versus researchers looking for extending or consolidating the knowledge).

For an article written in October of last year it sure is funny that there's an HN discussion about it from last January...

https://news.ycombinator.com/item?id=21947551

Healthcare system? More like moneycare system.

I like his idea and would've also been naive and idealistic to believe doctors want your best. However, you learn the hard way, even in Europe: doctors want your money.

If they get a lot of patients their goal is to minimize time with you. 5 minute consultation, prescribe an antibiotic without ordering a lab test to determine if it's even bacterial, next!

If they don't get a lot of patients, then diagnose something that requires multiple treatments or prescribe something that kinda helps but not completely. They'll come back to you sooner than later and TADA you have a "subscriber".

It is tough to see one's project(s) fail but in hindsight, I wonder if the author could have gotten feedback from customer even before writing a single line of code to do the meta-analyses (let alone paying contractors). This is on the lines of "simple but not easy" but here goes anyway:

1. Identifying "who am I solving this for"

2. Identifying "who might pay for this"

3. Prototypes , mockups of the solution

4. Getting out of the building (GOOB) and talking to the end customers (the ones who will pay) and validating 1 & 2 before going full steam ahead

This sounds like a classic Solution In Search of Problem (SISP)". I think the author has a great idea but it just took so much effort to find the problem. When nobody jumps out of the chair and starts writing a check, then it is time to start iterating the MVP or pivot. I'm pretty sure there's a problem in the world that needs this solution, but it just take inundated effort to find product market fit. Instead, when you start with the problem it is somewhat easier to get to the point of the product market fit.

> We worked at a startup that leveraged autonomous blockchains to transfer money from naïve investors to slightly less naïve twenty-somethings. There are worse gigs.

At least he's being honest about it.

> We worked at a startup that leveraged autonomous blockchains to transfer money from naïve investors to slightly less naïve twenty-somethings.

What does it mean to “leverage autonomous blockchains”?

Yes, I read this article several years ago. Why is it reposted with a recent date?

Anyway, besides all the BS the author wrote, I didn't see the business model. What was the basic - dumb business model?

So you find which is the best painkiller. So what? How do you charge users for this? This is one-time info. Once we all agree which is the best, why would anyone pay you for this?

And why the fuck did you draw diagrams on your parents' windows? Have you watched the facebook movie too many times?

> There was this giant thing called healthcare right, and its main purpose is improving health

My reaction to this was:

laughing: Oh, how adorably naive.

followed shortly by:

crying: Ugh, we're so fucked.

This quote.

> We worked at a startup that leveraged autonomous blockchains to transfer money from naïve investors to slightly less naïve twenty-somethings. There are worse gigs.

There is so much to take out from this story for developers who want to do the same.

Marketing is hard, making people pay for something is even harder.

Your idea about "fantastic app" might not be so fantastic for target group. Problem is there is so much more to any business that we don't know from the outside. There are things that unless you are in the group you are not able to notice because those are "dirty" trade secrets.

Seems to me the product here is the database.

Doctors don't want another product to learn add to their already saturated load.

Most doctors and dentists I know are very non-technical actually. I know a dentist who could barely work email.

However there's some software they probably use quite a bit, if those vendors were to license the data, and maybe the graphs generated by the data - THAT is the business model that would work imho.

This seems like it might be more valuable to a health insurance company or a government health agency than to doctors. Basically the tool tells them what to prescribe to be most effective so customers don't have additional future claims, and this saves the insurance company money. They can also offer it to doctors who have agreements with them as an incentive to sign on etc.

Seems like maybe you would have been good to talk to Kaiser or other companies more directly incentivized towards positive health outcomes?

Why dont you create a platform to save such as data sets. etc. and sell that. I think you were selling the wrong product of your startup.

If you want to see an index of meta studies, you can check out the Cochrane Collaboration: https://www.cochranelibrary.com/

It’s mentioned in the book Bad Pharma which has a pretty good breakdown of how drug marketing to consumers and medical professionals distorts decision making.

See that sertraline vs. Lexapro curve? Pretty similar. Pharma is very competitive. Successful pharma products are so close doctors use intuition, experience, and hunches to break ties. Also, they’ve been shown detailed one-of curves by pharma salesladies, EACH of which purports to show that THEIR drug is superior.

Then there’s that $0.50 revenue/user problem.

As a researcher interested in improving evidence synthesis, I would have loved to use this. Can't this still continue as a volunteer service/non-profit/research project? It reminds me of Examine (https://examine.com/) with broader application potential...

Why not ask funding from the NIH, or sell access to the companies doctors use to get their suggestions, like UpToDate.

I can see why doctors themselves wouldn't pay for the service; it's a net loss of money in the business sense, and they already believe they're prescribing evidence-based treatments, maybe just not THE best evidence-based Tx.

Just an interesting thought, wouldn't the medicine companies be the better place to market this product? This fits the marketing principle of 'benefit to all' the author has highlighted. They could have probably used it to analyze how their product fares against their competitors.

My startup journeys were not successful in the eyes of the world in terms making money or blowing up. Yet the experiences, people I met and most importantly the skilled (developer/designer as a living) I learned made it successful for me!

It was the best way to learn how to code for me personally.

I think the idea is good. Something like AI suggested drugs. But I guess that is hard to do the right way. On the other hand if I really care for the a new drug I take. Just using google scholar and quite often you have some summery studies which compare different meds.

How did you write 200k lines of code in nine months? Was that all functioning code? Was some stuff generated? Retaining ~730 LoC per day including weekends for nine months is a really crazy pace and I'd love to hear more about that workflow.

In a nutshell, this is just another story of coding something cool. Just to realize that no one will pay for it 8 months after.

YC also says all the way through "Talk to your users/prospects".

That's how you "Make something people want".

I mean, I think that the people to target with this would be insurance companies, public health organizations, or employers. You're generating value for them. Unfortunately, those folks aren't very easy to sell to.

Hey MS, Google, Amazon, Oracle, and countless others invested billions into health-it and made no observable impact. 40k for one of the most popular pharmaceutical site really is a wonder to behold, sorry about the 40k

He quotes Strategic Marketing Management, which is a great book if a bit academic. Another great quote I always remember from that book:

"Marketing is not about selling a product, but about figuring out what product to sell"

Perhaps this would have been a great business model to start astroturfing big pharma by focusing on SEO, allowing users to upvote meds that worked well for them and letting big pharma pay to bump theirs to the top or highlight it or something like that. Or for a given class of med, charge big pharma or individual chemists a certain amount for referral back links.

As a consumer your flow is probably naturally: what type? which brand? where is cheapest in my area?

There are plenty of energy comparison companies making money that way. They have a lot less tangible data.

I’ll be honest - the tech is so cool you nearly made me want to reach out and buy where you got upto. 'Nearly' is code for, yes I'd be interested, but I don't have enough $$ to make you interested vs the work required to build a profitable model.

SEO on medical topics without expert backing is a notoriously uphill battle as Google can arbitrarily apply heavy penalties to the search results.

> We have very high Page Quality rating standards for YMYL [Your Money or Your Life] pages because low quality YMYL pages could potentially negatively impact a person’s ... health

> The Lowest rating must be used for any of the following types of content on pages that could appear to be informational:

> * YMYL content that contradicts well-established expert consensus.

> There is no evidence that the author has medical expertise. Because this is a YMYL medical article, lacking expertise is a reason for a Low rating

Source[0] is 175 pages. Grep for "medical" or "YMYL" for more details.

[0]: Google Quality Rater Guidelines https://static.googleusercontent.com/media/guidelines.raterh...

Sounds like this might work to offer it to doctors as a white label phone app or pwa, that each customer (doctor or group or clinic or hospital etc) skins and gives to their own customers or sticks on their own website.

It has the recommendation engine and the doctors marketing looks and some forms of low-friction way to contact the doctor (generate billable service) from email to video chat.

Doctor pays for the startup skinning, any feature changes or updates, and a subscription which pays for ongoing development of the engine and the datasets, maybe some related hosting.

Doctor maybe gives the app to tge users free, maybe charges if they think they can pull it off, and generates sales from the in-app "help me" buttons.

Could have a few quasi-generic versions geared to different kinds of practises that smaller practices could share.

Would it make sense to go to whichever drug company ranked #1 and ask them to sponsor?

The key would be to remain true to the message if the #1 changed. In all cases you would put the ranking up before asking for sponsorship.

Just want to say, if I ever get a headache, I don't try to block it with a pain pill, but do what my body's asking me to, very loudly, which is to sleep and rest.

Analysing studies and providing summary on various topics is what examine.com does. I am a happy customer, so there is definitely a way of monetising meta-analysis.

This seems like it needs to be a nonprofit database - perhaps funded by government grants or industry sponsors - rather than a for-profit growth-oriented startup.

And why did you quit your job? Always validate the idea first and validate it only by making lots of people pay you for it. Not by asking your current boss.

Maybe its just me but I would have approached drug companies to partner with their marketing teams. THAT would have been something worth paying for TO THEM.

Never approach marketing teams unless you’re Theranos sized courting Walgreens.

Corporate marketers hate risk. It’s a great way to lose your cushy job. What I have seen happen is that they embrace innovation, parade you around their offices and never close on a deal.

I saw this happen. The founder of BuyYourFriendADrink was the doyenne of Diageo’s hallways but after burning six months of runway no revenue producing contract was inked. Some corporate politicians simply exist to take meetings.

Funnily enough, the best apps that that use ai to find your celebrity look alike are making hundreds of thousands per month in subscriptions and ad revenue

Have you contacted any doctors or hospitals in Europe? We do take healthcare more seriously here, so I'm wondering whether you get the same response.

The monetization part of the idea is as important as the idea. A lot of first time entrepreneurs waste time on a project that is impossible to monetize.

> It had been a bit of a working assumption of mine over the past few weeks that if you could improve the health of the patients then, you know, the doctors or the hospitals or whatever would pay for that. There was this giant thing called healthcare right, and its main purpose is improving health—trillions of dollars are spent trying to do this. So if I built a thing that improves health someone should pay me, right?

> I said goodbye to Susan and tried to cheer myself up. I had ten more meetings with doctors all over the Bay Area—surely not all of them were ruthless capitalists like Susan. Maybe they would see the the towering genius of GlacierMD and shell out some cash.

This is pretty accurate, unfortunately.

I can't see this being useful for consumers directly. I'm not going to pay to optimize the efficacy of the pills I get for a headache or heartburn or bloating or whatever. If I choose one brand/compound over another and it's 20% better...what's that worth? $5/year? Maybe?

I can see this being useful for problems that are not trivial. Heart medication. Cancer drugs. HIV medication. The sort of stuff where a 20% difference in efficacy is not the difference between me having an extra cup of coffee or not—it's the difference between an extra year on medication and worse side effects or not (or even life and death).

But even then, you're not selling to patients—it seems super unethical (or at least immoral) to profit from folks with serious illness who want to know what's best going to keep them from dying. You sell this to the healthcare industry. But like the post says, doctors don't...really care. Not that they shouldn't care, but they don't benefit personally, and the patients aren't going to pay them more for it, and the industry will (slowly) get better on its own in the long term. That doesn't feel great.

So to me, it seems almost obvious that the customer here is the pharmaceutical companies. They've got money to burn. They pay HUGE amounts of money for display ads and instagram stories which literally don't even tell you what the damn drug does. During Ru Paul's Drag Race this past week there was an ad for a medication which didn't mention what it treats or who could take it.

Make the data free to consumers and doctors. Hire someone to call every cardiologist and HIV specialist and oncologist and whoever else in eastern standard time. "Hey, I've got this thing that will tell you what's likely the best drug for patients. It's free to use, and we'll send you a free mug after you use it 100 times." Then you call up every pharmaceutical company you can find and say "Hey friends. We'll put up a profile page for each of your drugs and make sure they're represented with independently verified data. We've called X thousand doctors who will be using this. Pay us $Xk/year and we'll make this happen."

You could probably do the same thing for healthcare companies (e.g., Kaiser): "pay for a license for all of your doctors and you'll have better health outcomes for your patients."

This is obviously untested, but if you exclude the doctors and the patients as potential customers, you're left with a pile of companies with a lot of money. I would love to hear from OP why this wasn't something they pursued.

What about selling it to government which provides healthcare to maintain it and make it free for their healthcare system?

Well written - that was an entertaining read.

So I built something people wanted. Consumers wanted it, doctors wanted it, I wanted it. Where did I go wrong?

I think it all went downhill where he refused to setup a fake business model with fake revenues and customers, pour anther 10K in online astroturfing touting the marvelous startup clearly heading for unicornicity, then immediately hit all venture capitalists and investment funds with financing requests.

This is a big, once in a life time, fake medical company idea, he could have easily gathered capital in the 9 figure range before retreating away on a tropical island to seastead.

I am sure that by mining data from people's phones he could have sold users data to the lowest bidder...

There is a book titled Powerful Medicines that proposes including efficacy as a factor in new drug approval.

As a user, I'd definitely pay a moderate subscription for this service. I think it's super useful.

Does Wharton’s book really say that?

If you just make value for the company it is a scam?

If you call it entertainment purposes I guess its fine

Not directly. Of course I always have the best interests of my patients in mind, but, you know, it's not like they'll pay more if I prescribe Lexapro instead of Zoloft. They won't come back more often or refer more friends. So I'd sorta just be, like, donating this money if I paid you for this thing, right?

Coming from a Doctor, this is extremely worrisome and borderline inmoral

To me at least that was an impressive amount of working your butt off for something you believed in.

He paid $40K for a real life course on how to do market validation that he’ll never forget.

I think he could had licensed this tool to another site. Maybe even webmd would want it.

Why not try to fund it by selling pharmacies, or with some kind of affiliate program?

wonder what wouldve happened if instead of trying to make money he focused on getting venture funding and growing big as a consumer site? seems like alot of companies these days dont have a monetization stategy.

Doesn't this count as "practising medicine without a license?"

The author has a great sense of humor. This was really fun to read:)

should have made it enterprise software for HMOs - they profit when their population has better outcomes. Or sell it Epic or something and they'd package it into their systems.

I enjoyed that read.

Sounds like it may be a familiar refrain in the Startuposphere.

You were lucky. Good on you for getting out when you did!

The author can write a story than a lot of writers.

A tale that is foundational and reoccurring.

You need to write books, really.

The product is good. Keep going.

this lines up with my experience - naproxen sodium is the best otc pain reliever.

Before spending all that cash or even incorporating, he should have pitched his idea to VC in the health space.

The author brilliantly makes the case against single payer healthcare. Bravo. Job well done, sir.