How to Kill a Startup Idea with Google Keyword Planner and AdWords: A Case Study | Hacker News Reader

How to Kill a Startup Idea with Google Keyword Planner and AdWords: A Case Study

psl.com

568 pointstreblig188 comments

So...to me this article illustrates the dramatic differences in objective and mentality between bootstrappers and VCs.

This VC looks at this idea, figures out that in order to own the market and make $fuckton, they'd have to spend > $fuckton. Decides to bail.

As a bootstrapper of businesses, I look at this case study and go "well yeah of course it'll take $fuckton because you're haven't focused on a narrow enough niche". Which of course they haven't because you can only make $shittons from niches, not $fucktons (where $fuckton is up to an order of magnitude larger than $shitton).

It's a very different set of expectations. $idea might not be investable as a VC, but if you can really define a specific niche, get a toehold, and patiently build from there, $idea might have merit for a bootstrapper.

(All that said I don't think this particular idea has legs based on the reasons identified in the "Regarding competition" section and also from my own limited experience in the music space).

Ansel from PSL here.

Absolutely. Because we're venture-backed and are spinning out venture-backed companies, we are limited to billion dollar ideas. It's not uncommon for us to kill great ideas that could "only" make tens of millions of dollars.

We're hoping through blog posts like this and other means to be able to share more of them because we want those companies to exist, we're just not set up to create them!

Hey Ansel, thanks for responding and for validating my train of thought...I don't have that much exposure to the VC world so the things I suspect far outweigh the things I know. :)

feel free to start sharing those "small market" ideas haha

After they've been vetted of course ;)

> we are limited to billion dollar ideas

How did matchmaking for music lessons get into the discussion as a billion dollar idea?

Referrals for tutoring in any subject (math, reading, music, etc.) would be a bigger market, but even then it might not be a $1B company.

Well I was right there with you, thinking it was a completely shit idea, but a little math shows how it at least popped up on the radar (and I don't think its quite so bad myself after some thought):

100,000 music instructors, $60 per hour, at 365 days per year is roughly: $2.19 billion gross revenue per year-hour in the segment. Assuming, the average hours per day are like 3.5 (I can't imagine folks doing this are giving lessons for a full 8 hours per day, 5 days a week), that gives a total market of like: ~$7.65bn. Assuming they could earn a 20% stake in 100% of the market, that's like $1.53bn.

Looking at the numbers, I think there is likely more potential than they realized[1]. I would suspect that by and large, digital advertising is capturing almost none of the market as kinda proven by their analysis. Looking at how low the frequency of searches were compared to the sheer number of employed individuals tells me something is off as a whole with the analysis.

That is to say, digital advertising isn't showing the volume of or demand for musical teaching service because it's highly likely everyone believes it's incapable of doing so or adding value... That (driving the market online) is the problem to be solved in the space, because it sounds _hard_, and solving hard problems tends to be the way to make a lot of money. The other opportunity I would look at is, sort of from the other side: how to fix (what I assume to be likely) most music instructor's "underemployment" problems (i.e. inconsistent, few, or not enough hours). I imagine this angle would probably net more gross sign-ups as well since users (instructors) would be advertising the platform for you.

With all that said, you'd have to assume there is latent under-served demand (25-35%) in the market too to bother with anything I've said... that 25% is essentially what would be left over for others after you've captured the niche.

One last thought, for the VC since it looks like they're reading HN. Have you thought about doing retroactive market analysis where unicorns now exist to see what the market looked like then (e.g. short-term vacation rentals in ~2006)?

Might give you some strong signals for where there is potential...

[1] They are indeed a VC so it may not be in their best interest (time value of money, opportunity costs, etc)...

Edit: Adjust some bad arithmetic.

$60/hr and booking 3.5hrs per day all 365 days of the year are all really high estimates IMO.

That's $76k/year. My guess is that only the top 1% of music tutors are making that much per year.

And only 100k tutors in the US is an incredibly small market.

Yeah booking 10-15 hrs a week at $35/hr for 40 weeks would probably be a much more representative sample (Thats $14,000/year per person on the low end)

> That is to say, digital advertising isn't showing the volume of or demand for musical teaching service because it's highly likely everyone believes it's incapable of doing so or adding value

An ad for online one-to-one art lessons has been popping up in my Wechat moments. Judging by the ad copy, they feel they have two strengths:

1. No need to transport your kid to and from the lesson.

2. Your kid will like this more than classroom-based lessons.

I conclude that the market they're trying to serve is parents who would like their children to learn to draw. But there are a lot of those, and their selling points seem pretty reasonable.

> Assuming they could earn a 20% stake in 100% of the market, that's like $1.53bn.

That makes no sense. Why would/should instructors+students continue to give them 20% commission for follow-up classes beyond the first one? It seems like just wishful thinking. The app store situation is very unusual (monopoly/monopsony), and does not apply in this context.

Yeah, this is the problem Wag has been having... and they ended up trying some draconian things to try to prevent people from working outside the system.

These market place ideas don't work nearly as well for businesses where there are long term, repeated, interactions between seller and purchaser. They work best when there is a one time exchange, because a market is really getting paid to make an introduction.

edit: I see now that many people made this argument much more eloquently in other comments on this thread.

By providing some small value-add products that are not possible easy for individual instructors to provide, but customers like to see. E.g. nicer scheduling of lessons or a way to share sheet music with the customers?

Very few people would give up 20% of their income for "some small value-add products."

With platforms like these the choice isn't really up to the supplier, but to the consumer. The consumers come to expect those value-adds and so the suppliers have to switch to the platform (where the revenue share is opaque to the consumer) to get gigs.

The 20% cut also does sound rather extreme, but if you build a very good product where you can end up charging even more than the usual price, the 20% isn't taken away completely from the supplier, and might end up more like a 10% actual cut.

With platforms like these the choice isn't really up to the supplier, but to the consumer.

Once the platform has a significant number of suppliers, sure. If the platform is new then attracting suppliers is the number one problem to solve, and that's really hard if it's too expensive.

Isn't that what VCs are trying for? Or rather, the nature of their investments? We're talking hypotheticals to analyze the potential of the business, saying it "makes no sense" doesn't mean much. We're trying to gauge an idea and to do so we have to be able to look at the potential.

If there is a medium to high drop out rate, and teachers are constantly looking for new student to fill their schedule (even 30% of their schedule), it could well be worth it.

It also assumes that all discovery of teacher is done by direct searching instead of word of mouth - I know some one Phd in music, London session Musician who's played on top 10 hits he wont need to advertise

This is basically a hyperlocal seo play not sure why PSL seems obsessed with PPC only paid is not the only channel - go talk to some one Like Norm at DAC group

Experienced guitarist here

$60 / hr is not awfully high, lots of private tutors in the western world charge that, but note, that's it the west. There are also hordes of tutors charging as low as $5-$10 / hr in the west. It is unfortunately a business that is absolutely overrun with undercutting.

I think in reality, the only tutors you'll find that can charge that much, consistently, are professional musicians (i.e doing it for a living), and those with a degree.

Neither does it help that free alternatives like youtube and IG vids are in great abundance, which again can help to bring down the perceived value of lessons ("Why should I pay $60 when this great musician with 100k followers is putting out lessons for free? / on patreon for $10 a month")

There's probably some kind of market, but I'm not sure it's a billion $ market.

I'm a member of various FB musicians groups: buy/sell, discussions, theory, etc. and you see small businesses and startups like these come by all the time.

Not sure about now, but I was a musician with a variety of instruments for about a decade in my youth. In my experience, private instruction was always locally advertised. For instance, I had lessons on oboe, violin and piano. The piano teacher was a friend of my mothers, oboe and violin I had a class at school where the instructor suggested where to get tutors, typically either former students in the local orchestra or students at the local college music program. To me, an app would have to improve upon this relatively convenient method of finding instructors. Typically these instructors come with the implicit approval of a trusted person as well.

> How did matchmaking for music lessons get into the discussion as a billion dollar idea?

This is a near perfect example of how "big business" can kill the spirit of entrepreneurship.

The idea (or VC cocaine induced fever dream) that every idea should be a billion dollar idea is exactly whats wrong with the current tech climate.

Brainstorming and validation rarely starts with a perfect idea. It starts with some about of demand, an unsolved problem or an inefficient market. We start to dig and as we talk to potential customers we get a better understanding of the problem and pivot accordingly. Sometimes we end up unearthing a great idea, sometimes not. It's worth remembering that finding billion dollar business narratives often involve some amount of market creation or expansion. It's not just "can we capture enough of the market that exists". It's "would customers pay more for a better experience?", "would more people become customers if the process was easier? Or cheaper?".

- Ansel from PSL

Does this apply to the vast majority of VC's? So if I want to do a yc pitch it needs to be able to have unicorn level potential? Who would you target as investors for tens of millions ideas if so? I'm guessing a larger focus on angels and IB's after a more proven, traditional profit model is shown (often a much different model than VC-startup profit model).

Yes, VC returns are dominated by the huge hits, so everyone is chasing those. Google almost certainly returned more than every other investment made that year for example.

For a tens of millions pitch go to Earnest Capital who will be perfectly happy to invest.

https://earnestcapital.com/

Yes due to how the economics of traditional venture funds work, in order to achieve the returns that limited partners expect, investments are made in companies with unicorn potential. Some will do well, most will fail, sometimes one will actually become a unicorn but because you don't know which is which you need every at bat to have a path to a home run.

This means that if your business is the right shape, VC is a fast way to raise a tremendous about of money.

But that's NOT to say that every company can or should raise a tremendous about of money. There are less traditional VCs, angel investors, debt, and other financial vehicles that can help businesses succeed. There are tons of examples of companies worth hundreds of millions of dollars who raised little or no money to get there (Webflow and Atlassian are two top-of-mind examples).

- Ansel from PSL

Have you found many billion-dollar ideas by following this ad-testing methodology?

We have spun out 20 companies over the last 4 years which we are confident meet this criteria using this methodology. All are venture backed (led by investors other than us) and still operating (except for 1 which sold to Nike last year)

- Ansel from PSL

I understand the logic, but I can't help but feel that this type of assessment is really shortsighted.

Would Facebook have passed this "test"? There was Myspace and Friendster dominating. How about Instagram? Why would anyone want to share photos elsewhere when everyone posted their lives on Facebook?

We all know how those stories ended.

Your firm's method only addresses the current market in current conditions. Having the foresight to see the currently unseen before anyone else is what yields amazing results. And if this is the main reason for killing something like this off... let's just say I'm glad I don't have my funds there. :)

You're also ignoring the probability of success. This project is only worth $shitton ($30 million) if the project were perfectly executed and other market participants didn't dynamically react to this new entrant. This isn't the likely outcome.

The expected value of this project is probably only $5-10 million once you factor probability of success into account and thus not worth the time and effort at trying in the first place. A $5-10 million E.V. project is very much worthwhile for two founders who wanted to bootstrap though!

One reason VCs target billion dollar ideas is that you'll probably fail. But in the unlikely scenario that you succeed, it more than makes up for the 10-20 other projects in their portfolio that DID fail.

You won't even make 1 penny, if you can't break even on your user acquisition costs. That's what usually breaks a business not the fact that it's not profitable enough, but that's it not profitable at all.

But music lessons are an ongoing (hopefully) expense. If your pockets are deep enough you can run out all the other players and capture the market.

The lessons are an ongoing expense, but the principals will rather quickly eliminate the middle man marketplace after meeting and being happy with the connection.

Unless you can provide other useful services to help the teacher to manage thier students. Maybe scheduling, accounting, payments.

In that case it seems a lot easier to skip building a marketplace and just create booking software for music teachers.

Though you'd probably find that most were happy with Google Calendar, Xero and PayPal / Square.

^ Exactly - this where 'commoditizing your complement' comes into play: https://www.gwern.net/Complement

As your niche gets smaller, you will be able to focus your marketing a bit better. This should result in lower user acquisition costs.

That said, if you want to specialize in $shitton projects you should be picking ones that don't need to be perfectly executed. There's a literal shitton of $shitton projects out there that you can half ass and still make a $shitton from. In my experience the best method at finding these $shitton projects is throwing shit on a wall and seeing what sticks.

Often though, many successful VC startups start from a niche and expand from there. For ex: uber started with luxury black car service only in SF (a niche) and then expanded into normal cars, carpools, food and the rest of the world. I think the real requirement is a foreseeable future that you could potentially expand into a bigger $billion market.

Small niches that have proved themselves are often more attractive from an investment standpoint.

Had a similar experience running a consulting practice for a global software company. Our practice revenue was growing with profitably around 20% EBIT. Software sales where not growing as fast. As a result consulting was an increasing % of regional contribution income but at a lower margin. So my practice was lowering regional EBIT margin. We got told to slow down. I put an amazing guy in charge and resigned.

Millions of incremental profit... unwanted. Of course it's logical given shareholders, but remains strange all the same.

There's a large opportunity cost to that strategy. If you allow consulting to make up 80% of revenue by default the business will inevitably turn into a consulting business, not a software business.

Consulting can be a great way to bootstrap but you need to know when the tail is at risk of wagging the dog.

money is money IMO. If the business turns into a consulting business because that's where the money is then so be it.

during a business downturn, it can be hard to get consulting contracts as everyone is under pressure to cut costs. That's when having a viable software business can be helpful.

i totally agree, i was being overly terse. To me, if something is working well then you should do more of it but, yes, you have to manage risk and see the whole picture. Like everything, it's a balance.

It's counter intuitive but in downturns consulting can tick up. The first place companies go to look to reduce cost is headcount but the work still has to get done. In come the consultants to implement a new system to increase efficiency and reduce headcount (it rarely turns out that way though). Also, I think it's easier to finance money for consultants than FTEs because of where the expense falls on the accounting books.

Using that case study, getting $21k in revenue for $80k in expense makes zero sense ever. That doesn’t even count the actual startup cost to build the platform or even get the “talent” to deliver the lessons — which is an expense not to be ignored. So the real $80k is probably much higher because for a marketplace to exist, you have to attract buyers and sellers. And a 30% commission means that music teachers would have to sell at 30% more than they would for their own off-marketplace customers. And, unlike app sales, 1-1 music lessons don’t scale, so that 30% is a real hit because they can’t really make it up in volume: 1 hour can only yield, 1 hour of lesson for one person. I know this problem well: I founded a therapy marketplace that has been surviving for 10 years now, but it took 8 years to actually become profitable and still, just barely.

Real world music lessons scale better, with many beginner classes at Guitar Center (or private centers/homes) scaling to 5ish students per teacher.

It's not as high quality as 1 on 1 instruction, but the benefits of scale cannot be understated. Offer lessons at 1/3 price but make double the revenue.

In that case it's the regional "Guitar Center" or "Yamaha Music School" or whatever that provides (i) a marketplace that also lends credibility to the instruction and (ii) scaling via group lessons and as a result take a cut from the pie.

I’m a bootstrapper and I also use a similar metholodology to this article.

I only go niche if the larger addressable market is big enough. It might be okay if the initial niche offering only makes $5k its first year, but I should see some path to expand to $500k+ per year or the opportunity cost is too high.

Totally makes sense for both sides really.

VCs look at this and are like "We can't do that with run-off-the-mill webdevs and marketers". And they have a formula to express what to expect from this reasonable effort model.

Comes a founder with awesome experience in pedagogy, a reputation in, say, Montessori teaching, and publishing records. She knows 10 musicians who could potentially teach, 100 potential students to bootstrap the idea.

VCs re-reun the numbers with these new assumption and discover a potentially ten times higher return after Q1.

VCs and bootstrapers and founders make different assumptions in efforts and time and need each other, fit different niches.

So these guys have to read up on some articles about multisided marketplaces which what they were testing. Starting a matching platform between consumers and teachers is not a feasible plan because this is a longer term relationship and a platform is easily cut out. A platform like uber works because nobody has a regular cab driver. Supply and demand matching is also terrible for dog walking platforms. Don't spend time testing all kind of startup ideas with complicated technique when you don't know what multisided marketplaces are and network effects.

Can you clarify what you mean by "supply and demand matching is terrible" for dog walking platforms?

It seems like if the primary qualifier on marketplace success is the ease at which two parties can cut out the platform, dog walking would not pass.

Perhaps surprisingly, dog walking platforms like Rover do experience alarming levels of defection (when participants do a side deal and cut out the platform). The dynamic seems to be that you generally take your dog out for a walk at a similar time every day, and this regularity (and comfort with a specific dog walker that you've grown to trust) leads to more frequent defections.

Excellent point. This would have been reason enough to kill the idea irrespective of the CAC calculation.

Speaking from extensive marketplace experience: A pattern of long term relationships will lead almost surely* to supply-side defection in your marketplace, which is when the supply leverages its relationship with the demand to set up a side deal that cuts out the platform. This can quickly take both supply-side and demand-side churn to levels that are an existential threat to the business. And of course, fraudulently inclined supply side firms may turn defection itself into an optimized business process if it's lucrative enough to do so.

High defection rates ultimately killed Poppy (babysitting marketplace) and Tutorspree (tutoring marketplace), and they continue to plague even platforms like Rover and Airbnb despite (or because of?) their scale.

* Despite what many folks believe, it's is possible to run a successful relationship marketplace without significant supply-side defection; you just need to structure it to leverage the relationship, rather than trying to break the relationship into a set of one-off transactions. (As an existence proof, I'm running such a marketplace right now.)

I’ve seen some writing on what makes marketplaces resilient to “supply side defection” as you call it, but I’d love to hear your take. Specifically I’d love to hear more about structuring to leverage the relationship, or anything else you think is important.

What specific structural things are important?

Generally speaking: your participants will use you when they need you and cut you out when they can do so cheaply. This isn't an indictment of them: it's completely rational for them to do this. But that means it's up to you to add value to the transaction, until the perceived cost of defection is higher than the platform margin.

Airbnb did this beautifully: they set up insurance for all their hosts, and so were able to frame "stay on our platform" as a safety directive.

You can also go the opposite way, but that usually leads to user-hostile antipatterns. E.g., Freelancer tries to detect all the ways folks might send each other phone numbers on their messaging service. This is both annoying and easy to circumvent.

If your rate of defection is high enough, no amount of tactical tricks will save you. Incentive and ingenuity together are too much.

When you're small and still have structural degrees of freedom, target the incentive by adding value to your offering if you can. If the problem only appears when you're big, you may have the resources to play whack-a-mole.

If you're small, have high defection rates, and also can't economically add value to your offering - it's time to move into another line of business.

Upwork (formerly elance) is a good example of a platform which could leverage the relationship by providing better intermediary services (like dispute resolution). In practice, they don’t, and there is no real reason not to defect.

Yes, you even see this on marketplaces like Ebay, etc. Once you've established a relationship with some buyer / seller, it doesn't take long before said marketplace gets cut out from the equation.

I sell a lot of expensive stuff on the web, and I'd say 80% of my buyers are repeat buyers that started contacting me through mail after a couple of deals.

From guest point of view, Airbnb defection defense is pretty annoying, as it doesn't let you post URLs, addresses, or phone numbers prior to booking. Also pretty easy to circumvent.

How does supply side defection work for airbnb except in very edge cases, such as a long term rental. Even then it turns into something like apartments.com

That's exactly how it works. The fact that it's an edge case on Airbnb is what makes the resulting churn survivable.

Well it's pretty much driven by customers but same diff really. People who book a night or two then realize they want to extend and ask to pay cash, mostly. Someone who's already stayed with you just texting the next time, etc.

It works since Airbnb is mostly there to hook you up and protect from crazy people's wanton destruction, so once you know you're dealing with safe people and you've gotten the review, eh.

Wouldn't really be worth the penny-pinching, except people tend to forget about the long-term booking discounts and actually end up paying more than they would've just using airbnb.

How would you rank UpWork in this respect?

Upwork offers payment platform/record keeping and will dole out harsh punishments if caught leaving.

That's the value they provide and the stick. It works because trusting third world developers is hard. It fails for more local developers.

I wonder if it would help to rank a provider based on retention and CSAT and give more incoming leads to providers with more in-app retention?

How do you explain Angie’s List then?

Don't they also live off once in a while kind of things? A plumber once through the site is a much larger chunk of $ even if I call the same plumber the next time.

Two things come to mind (as a homeowner of a fixer-upper).

First, a lot of home repair (not all, but a lot) is small, simple, and fairly commoditized. I used Homeadvisor last week, for example, to repair a broken window, and I wanted it fixed basically as soon as possible. I really just needed a guy with the tools to come in and do the job fast, and Homeadvisor found me that guy quickly.

Second, it's hard to find a good contractor. I'm on my third or fourth plumbing change right now, and have used different plumbers each time. They were all fine, but I haven't yet found someone who I'd absolutely call first. Homeadvisor gets my business until that person comes along.

I don’t have a regular “Sheetrock repair guy.”

Not only that, I’m amazed that VCs call the awful ads they tested “marketing”. What they created is simple, uncompetitive, uncreative advertising, and that kind of thing doesn’t stand a chance in today’s world. Of course it performed terribly. Marketing today takes just as much creativity as the product you’re selling. Actually, I would argue it takes more in many cases. It’s like building a race car and putting low grade gas in the tank. Theres a surprising lack of business acumen here.

That's why they accounted for much higher CTRs in the long run by optimizing the ads. Also, I'm no expert but I would imagine you shouldn't base your market research on getting lucky with a viral ad.

If the ads were higher quality, the click through would improve but conversions wouldn’t. So it would drive up the acquisition costs even more — further making the point that this was a loser idea.

Hmm, this is not how improving ads has ever worked for me. Better ads = better performance.

higher clickthroughs to the landing page would have zero incremental effect on conversions? How do you figure that?

why would anyone bother to A/B test ad copy or landing page CTAs if this were true?

We specifically avoid "creative" ads. We know how to make ads more effective but that's for when we have a business we're trying to grow, not for when we are trying to determine if there's a business at all. We want to distill a very specific value proposition and target demographic in each ad. If the value prop resonates, they will click and they will convert regardless of ad "creativity". If we were just trying to juice numbers, we wouldn't learn anything whether and which value propositions are effective.

- Ansel from PSL

Just wrote a similar piece like yours which brought up another thought: Do you think disintermediation is solvable when it's a BTB (so all parties are businesses) and you put heavy penalties on disintermediation?

This might come off as awfully naive, but this doesn’t sit right with me:

> XYLO's mobile music lesson technology is the easiest way to learn new music, right from the comfort of your home. Our platform can teach your child any instrument on any schedule, and let them replay lessons forever.

As far as I understand, this is... a complete lie? You just made some crap up to see if you can make people click?

It's a lie, but only in a sense that they don't have a product yet. But that's something that they could have striven for if they were funded.

Can't you say the same for financial fraud?

"It's fraud, but only in the sense that we don't have the profits yet. With some investment we'll surely be able to make that much profit and more."

Also, it's not just a landing page. They take PII. I'm honestly a bit surprised this is legal.

It's understandably naive. As I commented in another thread, many many years ago there were discussions here on HN on doing just this. Most of us agreed it was immoral behavior. These guys are endorsing it.

The argument from the other side was: Yes, you are making some people angry. You're going to piss off 300 potential customers so you can learn enough to get 300,000. You're not taking money and if you protect their email addresses (don't sell them, etc.) then it's worth it because you can save a lot of money and time.

You have platforms that thrive on this : Kickstarter, Indiegogo.

They are filled with concepts, prototypes, what ever that's sold as end products.

Yet they ask for money, here it's offered the possibility to get more information.

Yes, there are many people who lie. That’s what makes it a problem, just like with other types of pollution.

But in my experience your characterization of crowdfunding is mostly wrong. While projects tend to describe their product as they intend for it to be in final ready form, the whole concept of crowdfunding makes it clear enough that backers help fund something that isn’t yet a reality, and that there is risk involved.

This is very useful insight that I want to apply to my own planned side-projects (like 30-something domains registered as ideas...) Too often I've thought "that's a great idea" and then focused on the tech of implementation rather than the viability.

At that point it's just a hobby, or skill-sharpening exercise.

Regarding the example shown, I would think that "uber for" ideas (meaning facilitate a connection between service provider X with service need-er Y, take a commission on the transaction) sound as though possibilities are unlimited. However the reality may be that not everything can be disrupted in this way.

Reading this article and seeing these replies, I feel like a communist who accidentally walked into a Davos hotel lobby during WEF.

What about just making a website to connect learners and providers? This commission-taking, the Uber model, such effing rent-seeking.

I mean I also don't work for free, but things like "Uber-for-dog-walkers, but we'll fine you much money if you arrange dog walks outside of our service. Also we don't vet our walkers so they might kill your dog[1]."? Come back to reality please.

1) Top result when I google "dog walker kills dog" https://nypost.com/2019/12/01/embattled-dog-walking-app-wag-...

> such effing rent-seeking

That was also my take-away regarding this idea.

20% cut? For match-making on private music lessons? Fuck that. Our local music store charges less than that for renting an hour in a soundproof room! And the discoverability/match-making "platform" is a free bulletin board next to the restroom...

Also love how that email described the radical act of checking the local music store's bulletin board (or going and asking a friend/coworker/etc. for a recommendation) as "under-the-table neighborhood activity".

Plus you can get a referral for private lessons from virtually any public or private school music teacher. For free.

> Plus you can get a referral for [snip] for free

Who actually thinks that a(ny) startup is going to be able to provide consistently higher-quality referrals than an actual word of mouth referral from someone you know?

And the existing competitors are apparently charging a 40% cut!

Even more depressingly, good music teachers, once they get established, will typically have as many students as they want through word of mouth alone. So the teachers you'll find who are desperate for students and willing to give up a 20-40% cut are probably mediocre.

I mean, there ARE services to provide in this space that aren't rent seeking and are worthy of being paid for... things like financial service (taking payments), escrow service (making sure each side provides the agreed upon service, and arbitrating any disputes), vetting services (could be ratings, background checks, etc), guarantees and insurance... i am sure we could think of many others

I am totally fine with a company providing those services to independent contractors... I just wish they would do more of those things and less of the attempting to force you to keep using their service

As others have said on this thread, this model is not very good for frequent, repeated services that are provided by the same person each time. When that happens, it becomes really easy to cut the middle man out and pay the service provider directly.

These services are matchmakers, and the matchmaker can only really be expected to be paid to make the match... not in perpetuity. This is why services like Uber, where you only are matched with a particular driver for a single ride and are unlikely to be driven by the same driver again (at least not intentionally) work so well for a matchmaking service.

If you hire a music teacher, you aren't going to have a different one every week. You don't need the match making services after that first lesson. So why are you going to keep paying the matchmaker?

For one on one personal service yes.

For something where people get some value out of using the intermediary over and over, even if it's the same teacher, see Peloton.

I could see that idea working for remote music lessons if the teachers were great.

A wall of text intended as smart content marketing to get some awareness for a rather unknown investment firm or incubator--Crunchbase didn't show any raised funds, so I don't call them VC. The long, over-polite text, the fake landing, the ad test weren't required in the first place.

Why:

The immediate answer every experienced VC would give is a simple 'no, this isn't a VC case' without all this fuzz and waste of time.

This market is useless for VCs because it's prone to disintermediation. Once people form a long-term business relationship, it's easy and reasonable to kick-out the middlemen (eg Homejoy). Marketplaces without long-term relationships don't face this problem (eg Airbnb, Uber).

Disintermediation is a hard problem nobody solved. 101 of investing.

Edit: Just saw another user posted the same. What is interesting, PSL didn't answer to that user's thread which could be interpreted as approval. So, PSL's post shows well that most investors are not per se smarter because they invest money. They're just humans like all of us trying to get free reach for a day with 'random' blog posts.

Counterpoint: Wag raised $300m for an almost identical idea.

Maybe because Wag targets a market where workers are hoping in and off more than cleaners ? Or because people are emotional toward their pets' spendings ?

Maybe because investors throw money to many stupid things.

("we never know" "you miss 100% of the shots you don't take").

And how's that working out for them?

Ansel from PSL here. PSL Studio has raised $28.5M. PSL Ventures has raised $85M. We have spun out 20 companies over 4 years. 6 of which are in the 50 Seattle companies to watch in 2020 (https://www.builtinseattle.com/2020/01/21/50-seattle-startup...) and of the early-stage venture backed companies who raised money in 2019, over 12% where PSL companies.

Can someone explain to me how this is legal? False advertising laws exist and this is literally marketing a non-existent product.

And even if it’s not illegal, I don’t understand how you can just be okay with lying to people.

Everyone in this thread seems to just be totally okay with this and I’m super confused about it.

I get the impression people here simply didn't catch that there was no actual product, since the authors were so nonchalant about it.

I went to the site myself and filled in some checkboxes. Added my email and it simply says i'm on a list and nothing else happens.

This is pretty disgusting.

If someone wanted them "brought to justice", they'd need to:

1. Sue them

2. Prove they were harmed in some way

3. Either hire a lawyer or do a bunch of paperwork (not to mention learn what paperwork they need to do)

That's why it's unlikely they'd face any repercussions from running the test.

I was scrolling the comments here to see if this was being discussed. It's my biggest takeaway from the article. We had a discussion many many years ago on HN (maybe more than one) and the feeling among the startup devs (not all, mind you, but the majority) was that such behavior was immoral. I think it came up once in a conversation about Steve Blanck (sp?) IIRC.

One thing I learned a few years back, however: If you're in a game and you're the only one playing by the rules, you're going to lose.

It probably is illegal under the GDPR and perhaps a future US personal data law. Take a look at what WilliamEdward wrote in a child comment:

> I went to the site myself and filled in some checkboxes. Added my email and it simply says i'm on a list and nothing else happens.

Under GDPR, your email is personal data (PII) and taking it from you under false pretences, to use for a purpose you were not informed about and did not consent to, is against the law.

Of course nobody is likely to sue over it.

Yeah, all I could think about was how crap this seems from a customer's perspective. If I'm trying to find lessons, and they promise lessons, why can't I get lessons? Not very customer-friendly, and since they're going to kill it, the people who signed up are just left scratching their heads that nothing happened.

I love this, I follow a similar method. One thing that immediately stood out is the Instagram ad creative is poor. The copy is overly technical and there’s no call to action. The image is a basic stock photo.

Spending 5 minutes in Biteable [1] or a similar service would kick out much better ad creative that would likely have yielded substantially better results.

Another point of contention is around the revenue model. Sure you could pay 4x to drive initial growth, but if you keep those customers longer than a month, the LTV could outpace the CAC.

Finally, all those efforts work together. Spend $20K on google ads and referrals will go up, you can test conversion funnels since you have traffic, back links will appear that will help your organic SEO in the long term, audiences for retargeting on Facebook will fill up, etc.

However, the conclusion of the author is one I agree with in the long term. Tech will solve for this. Hololens and Quest are close, the next generation will probably solve it across multiple platforms.

[1] I work for Biteable.

> that would likely have yielded substantially better results

No. It would yield higher click through, but not necessarily a higher conversion rate. It would actually yield a lower conversion rate because a lot of creative ads get curiosity clicks rather than people dying for the product on offer.

Let’s just pretend though in fantasy land that double the clicks meant double the conversion— the business would still be operating at a huge loss.

Do you have any concrete proof animated or video ads perform any better? I got pitched a similar service a few years ago and the whole thing came off as super obnoxious for the end user.

This startup also faces a fatal problem when users find a great music teacher, and then they just cut out the startup for future dealings with the teacher. They lose both the parents and the best teachers. Similar to those startups that pair you with a maid.

... or a soul mate, aka dating platforms.

Pairing people romantically is so hard to do that dating platforms can count on a long string of revenue from clients.

It's going to take many meetups before I find the woman of my dreams, but chances are pretty good that the first maid I get paired with will meet my needs (worst case, maybe I'll need to try one or two more), after a few successful visits, then I can offer to pay direct and cut out the middleman.

Don't people usually cut out the middleman on dating apps straight away? From my experience, after the initial connection, and before the first meet, you usually exchange numbers, and continue through anything else but the dating app.

In that aspect, I don't see how it's different than pairing with a maid.

You would only need to be paired with one maid though.

Correct, and if that maid is a match, you proceed. How is that different from a dating site? In theory they only have to match you with one other person.

But my main point was; on dating sites the middle man is almost always cut out instantly. Usually you try to immediately exchange numbers, and then proceed from there.

You probably need to go back to the dating site multiple times until you finally found "the one". Exchanging numbers and using the dating site service, again and again, multiple times.

Whereas when you're looking for a maid, you probably only need to exchange number once, and that maid will probably suit your need just fine for a really long time.

I don't know if you're tried online dating, but in my experience, it hasn't been "sign up for a month, find your wife and you're done".

I spent several years off and on with online dating (went on a lot of first dates, fewer several week relationships, and a few multi-month relationships) -- dating sites tend to give a steep discount for longer term (6mo, 12mo) subscriptions, so I was a subscriber for most of that time.

Ending up finding my wife offline though.

On the other hand, I found a housekeeper on Craigslist, called some of references (one was a neighbor), hired her, used her for a year, then when she moved out of the area, she referred me to someone else.

Turns out it's a lot easier to find a housekeeper than a romantic partner.

That's a lot of assumptions tho.

My experience with online dating has been the same, however mostly it's; match with a person, exchange a few words, exchange numbers, move on to WhatsApp. Yes, I do need to go back, but the middleman is almost always cut out instantly.

That said, my experience with finding a housekeeper has been the opposite of yours. I've had 4 different ones, and I ended up hiring none of them. This is not necessarily due to their professional skills, but mostly due to availability and trust. Just like finding a romantic partner. It's easier, but definitely not like you described for me. Hence why I saw them both as the same thing.

But don't they already have your money by then? They don't care if you stop chatting within their platform since you've already paid

> But don't they already have your money by then? They don't care if you stop chatting within their platform since you've already paid

I suppose so, if you pay for it, most of them are free. But that's besides the point, I just had a different experience from most people here, after 4-5 different housekeepers, I gave up and ended doing the work myself. So in that aspect, my experience has been the same as online dating.

Or is it not, and dating platforms make it difficult on purpose for people to return?

There was a comment some days ago along the lines of somebody being ordered to switch off a matching algorithm because it was too good.

at face value, this seems plausible. However, the dating dynamic is unique among other marketplace ideas. Case in point, tinder is profitable (whenever people find their soulmate through it, tinder loses customers)

or goes back after the heartbreak.

If they find a soulmate. Tinder is for casual sex, so people are returning.

Dating platforms subsist on ads and the sale of user data. Subscription revenues are a tiny proportion of their revenue base.

My major takeaway was how cheaply they tested this idea. Waiting for someone to sell this as a provider-as-a-service to VCs, so they can quickly get a real world estimate of TAM.

For $1k in spend and a day's work, they had a nice data point that tested TAM assumptions.

* Landing page is a template, figure 4-8 hours to put together the couple hundred words of content and configuring assets

* FB Spend = $509.23

* Estimate Google spend around the same (numbers not specified)

This is a very dangerous idea. It seems logical until you go and try it. The potential market is too small so you pass. Only to see the market blow up into billions of dollars that gets harvested by someone else. Do you think electric cars were that big an idea in July 2003 when Martin Eberhard and Marc Tarpenning started Tesla?

It also doesn't factor when you commit to an idea you become an expert. Often that knowledge forces you to pivot and become successful. Would the founders of Justin.tv ever have founded Twitch if they'd never started?

Ignorance, motivated learning and money fantasies are a proven method of exploring the business problem space. Analysis and wisdom have limited utility when it comes to uncovering opportunities.

They're not inventing any brand new technology. They're evaluating the business case for being a middle man in an oversaturated industry where both the customer and service provider want to cut out said middleman as soon as possible.

I wouldn't use Keyword Planner for keyword research data. The keywords that appear there are only being actively bid on. So keywords that have more search volume and no ads won't be included.

This is a great point. Can you please confirm with a source (or other commenters confirm)? Thank you.

You can use the Traffic Estimator Service

https://developers.google.com/adwords/api/docs/guides/traffi...

Disclosure: I work in Google DevRel on this API

What are some good alternatives ?

Neat article. It's definitely a good idea to get actual numbers to estimate total addressable market, CTR, and CPC. However I want to call attention to this bit:

> First off, there are a ton of enabled small businesses competing in this space and I'm not sure there needs to be a middle man. For example, I get 4 ads from businesses in Seattle offering lessons for guitar on Google. Thus, they know how to market and get customers, they are offering free first lessons, and have availability. So I am unsure/doubt there is any real consumer problem.

If your idea doesn't solve real consumer problems, kill it. It doesn't matter if the total market, CPC, and LTV are all amazing, this is sufficient. Even if you succeed, all you'll have done is shuffle some money around. Go make things that actually help people.

I agree but would add the caveat that many market niches have a surprising amount of depth that may warrant further discovery. For example TAKE one of these guitar lessons and if some creepy person comes up with a windowless van and gives an awful lesson there may be demand, just the demand is more hidden.

I think this part is really key to their analysis:

"First off, there are a ton of enabled small businesses competing in this space and I'm not sure there needs to be a middle man. For example, I get 4 ads from businesses in Seattle offering lessons for guitar on Google. Thus, they know how to market and get customers, they are offering free first lessons, and have availability. So I am unsure/doubt there is any real consumer problem."

The realization that this is not a big enough problem consumers are having.

I would guess that the musicians who want to sell their time would gladdly sign up to this platform. But the customers looking for these services, would most likely be very costly to reach.

The musicians already have a platform on which to find customers. It's Google.

I read one of the conclusion: "Lastly, I can't imagine that tech won't win here with non-human based teaching. It's affordable, convenient, and scalable to any genre, language, right from home on many platforms."

I strongly disagree with that one. The pie is big enough for different type of teaching, including real-life classes.

If what he predicts turn out true, and we all get plugged behind screens to learn new things... what a sad world it would be...

A "startup studio" and a "venture capital fund" ???

no conflict there !!

Like, great, please send us your detailed startup plan and then come pitch and answer our questions, and then.... well, its not a good match and we pass 99% of the time, but hey, look at this back office studio we got going here, trying out new ideas we have never heard of before.

I think you misunderstand how VCs work. If you come to them with a strong idea and evidence that you have built up that it will work and a plan for how to move forward and the right people and you try to get them to take equity they have really no incentive to to say no to you and try to build the thing themselves without you, when it would be so much easier just to help you build it.

A late reply.

Your logic is true for an 8-figure VC fund. Its more nuanced for a technically-proficient group looking at 5-6-digit (pre)seed rounds.

Sounded to me like they were paid for testing the idea for both the startup founders and investors because that's a specific service they offer.

Doesn't promoting a single page template breaking Google Adwords ToS? A few years ago when I did something like this Google disabled my ads.

It's crazy how markets saturate quickly online. You really need a large marketing budget and/or network effect to triumph nowadays, product quality is not differentiating enough. I am scared for the future, this will inevitably lead in harder-to-break monopolies.

I worked for an entire year on a next-gen comparison engine (picked.cc if you want to check it out), and even with exciting user feedback and crazy conversion rates, it probably will never scale because of those reasons.

> even with exciting user feedback and crazy conversion rates, it probably will never scale

How so? With crazy conversion rates, all you need is capital. Capital is readily available nowadays, if you need that marketing budget you can get it. Unless you do something inherently stupid like selling 1$ for 50 cents...

I was a student in debt at the time. Fortunately this project caught the eye of a fellow HNer and helped me get my first job in the bay area.

You need initial capital or connections. Right now in Canada the well connected are flushed with free money from the government. Just say you are doing AI or add AI to the company name (nope, elasticsearch document scoring recommendation is not AI!) and you get millions from the government to push your way into the market.

I’d love to see some of that money. I’m Canadian and you’re right, the stuff that gets put out as AI is quite sad. They tried to pitch me as well to get my feedback on their platform.

Neat idea, I created something similar years ago but didn't know how to market it.

What's your strategy to get users?

I tried every single trick in the book : SEA is too expensive. Word-to-mouth needs more critical mass. Website is not sexy enough to get press and backlinks. Blackhat SEO is out of question.

My only hope is to rebrand the website as a carbon comparison engine, which will make it easier to sell it to the mainstream media. Surfing trends is often the best solution.

As a musician myself, who has also been taught by music instructors over the years, I find this idea strange. Musicians are artists. They are also not beholden to market trends. Music instructors tend to be picky about who they accept when they choose their students.

Music is also a person to person thing. Those who are truly serious about learning will choose a person to person connection.

This is an interesting case of trying to quantify a subjective subject. Music is an art. It's hard to quantify. Musicians (mostly instructors in this case, then a bit later the students) will most likely shy away from this form of making a commodity of something that is inherently personal and expressive.

This model is mostly likely going to fail not because of the numbers, but because of the subject at hand.

This is an interesting case study. And it makes sense, the numbers prove that it's not viable. The fact that it's not viable isn't the numbers, however.

It's something to consider.

This makes some strong assumptions about the validity of the google and Facebook data that underpins this theory. As the case for the prevalence of ad fraud and questionable practices by these platforms mounts, it’s hard to believe any data around CTR and purported conversion vs actual legit human traffic and cash in hand conversions.

One project I've been working on is a competing (or now no longer competing :P) platform called https://classalog.org (marketplace of in-person classes). Found the same thing: buying paid ads does not make sense for a platform – you need to do this with free traffic and referrals.

I think the kind of math above is useful, but it doesn't replace making something people want. Maybe to get a rough idea of the metrics, sure – but airbnb etc wouldn't have started either if they'd done the math above.

This may strike some as silly, but before the iPhone was invented, there were ZERO searches for the term "iPhone". (I assume, ha!) By all means, do market analysis, but remember that demand is not static—it can be shaped by what you offer. I co-founded a firm 13 years ago that now employees 50 people that most certainly would have been killed by this type of analysis.

> before the iPhone was invented, there were ZERO searches for the term "iPhone".

Before the Apple iPhone was invented, there were probably some searches for “iPhone” because infogear had a product with that name in 1998, and Cisco (who has purchased infogear) started using it for Linksys products not long before the Apple iPhone launch.

Rumors of an Apple iPhone have been around since at least 2002 (5 years before its release).[1]

Apple trademarked the name in Singapore and the UK that year.

Here's an NYT article from the same year using the name "iPhone"[2]

1. https://www.fiercewireless.com/wireless/timeline-apple-iphon...

2. https://www.nytimes.com/2002/08/19/business/apple-s-chief-in...

there was also the Moto ROKR, which was a Jobs-blessed product. But in any case, keyword research isn't just about punching the exact term into a search engine. It's about looking at all the different combinations of terms that would identify user interest in your product.

For the iPhone that would have included search terms for "camera phone quality", "phone with web browsing", "phone with headphone jack" etc. At the time, the vast majority of phones came with a non-standard 2.5mm jack for a wired headset. The concept of listening to music on your phone was alien to all except maybe a couple of Blackberry and Nokia models.

> At the time, the vast majority of phones came with a non-standard 2.5mm jack for a wired headset.

The 2.5mm jack used on pre-iPhone wireless phones wasn't non-standard, it was a pre-existing, widely adopted standard for hands-free phone headsets established initially for cordless phones (the short-range ones with landline base stations) that was adopted by wireless phones because it was what every wired phone headsets on the market used.

of course there would be no searches for a new product name. However there are tons of people searching for “new phone” or “best phones”.

>I co-founded a firm 13 years ago that now employees 50 people that most certainly would have been killed by this type of analysis.

Which is the goal, because these guys aren't looking to build 50-man companies.

It seems like the authors made the unspoken assumption that the product would not significantly embiggen the market. That seems very reasonable in this case.

I think they missed the opportunity to add some value to the market, by doing some kind of background checks to ensure the teachers are safe to let into your kids home.

Trust is an element people always look for above all else.

It’s a value add for the first contact. But after you hire the teacher, why would you need the middleman? For continuing background checks?

It would be a one-off fee, but you would charge more.

You could also charge a fee for doing a credit check on the other side.

There are no music teachers. This product doesn't exist and giving them your info simply puts you on a mailing list waiting for an instructor that isn't real.

FWIW, my company, TrueFire, is one of the industry leaders in music education. We were bootstrapped, focused on a niche of the market first (intermediate to advanced video lessons for blues, jazz, etc = older men with expendable income), and have grown steadily since then. We were just acquired by a private equity firm and merged with another big competitor in the space (JamPlay) to form a new company dedicated to dominating this space. So... it is possible to not go the VC route and be (very) successful, albeit not $Bs.

I’ve always been annoyed with the fail fast model. The only group that benefits from fail fast are VCs looking for a big score. They push entrepreneurs to invalidate instead of teaching/helping them how to build a business. It’s all ass backwards.

But for the VC, if they see a hundred good ideas invalidated by bootstrapping founders, they save time and money.

The founders are the suckers if they listen to this bull crap. You still need to validate. But do it by actually trying to make the vision work, not by giving up at the first sign of trouble.

The only group that benefits from fail fast are VCs

Doesn't the entrepreneur also benefit if it keeps him from pursuing a business that will never work? Otherwise they may become a victim of the Sunk Cost Fallacy -- after investing so much time and money into their business, they don't want to give up.

The key, of course, is knowing if you really have a failing strategy, or if you just having devoted enough time/money into making it a success. And the VC can look at it more dispassionately than the inventor.

Ansel from PSL here. You're exactly right. The entrepreneurs we work with want the best shot at a winning idea. Just because we decide to kill an idea doesn't mean we don't end up finding something else in the same area to work on together. It also doesn't mean they can't go do that idea by themselves if they disagree. PSL does not hang onto any rights whatsoever after killing an idea.

It's also worth noting that only half the companies we end up spinning out were generated by entrepreneurs. The rest (including XYLO) are generated and validating in-house before we even seek an entrepreneur to partner with.

@treblig any advice on usually how much percent increase you see once you move up funnel with fb advertising (optimize)? I'm guessing the cpc you got are based on bottom funnel and highly targeted hence the high cpc. I find really difficult to estimate costs with fb, because it takes me a fairly long time (1-month+) to build the content and ad strategy to lower costs. But I'm wondering at what point do you know if its a bad business idea vs bad ad strategy?

I really liked their methodology. How can I learn more of it?

One Million Followers by Brendan Kane‎ although not directly related shows similar usage of ad platforms for content testing.

I wrote a blog post (https://blog.nichetester.com/the-assholes-guide-to-product-v...) and eventually made a web app (https://nichetester.com) to step people through a pretty similar process for new ideas.

Seems like some of this analysis could have been done before building anything. I also wouldn’t model it as ongoing CPC costs indefinitely. If the product is good it should develop word of mouth and use other marketing channels.

I would love to see a similar analysis that was greenlit because CPC looked so profitable. (Maybe it exists, I don’t know.)

This reminds me of the book "Will It Fly?" His approach wasn't targeted to SV startups, but the concepts are similar.

https://www.amazon.com/Will-Fly-Business-Waste-Money-ebook/d...

> Lastly, I can't imagine that tech won't win here with non-human based teaching. It's affordable, convenient, and scalable to any genre, language, right from home on many platforms.

That's just sad, even if probably true I'd still hope teachers remain. I'd rather have a human mentor when it comes to art.

They're assuming that because justin.tv exists and is popular, and that e-learning is now common in a number of fields, that this will apply to music as well.

Not so in my experience. I quickly learned that I can't follow along with my virtual guitar tutor if I broke the strings while attempting to tune it and have no idea how to re-string it.

Just a dumb question, but in the testing phase, what was actually happening here? Are the adverts for an effectively 'fake company'? What happens to users who sign up, do they get an e-mail saying 'thanks for your interest, but at this stage we are only gauging interest'?

This model seems to ignore ideas that could be big if you could find a cheaper distribution strategy other than ads, ie social, virality, etc. It also ignores ideas that don't have demand yet.

By this analysis you guys would have likely killed AirBnb and Dropbox.

I think this is a solid assessment and a great data point for founders assessing their idea.

Wow, didn't realize adwords and FB ads are so expensive. How do small businesses compete? are there any alternatives for those with smaller ad budgets?

From my experience with adwords/fb ads, this is not normal. The reason the cost is so high is because they're targeting ads to things like "bass" instead of "learn bass from scratch" in Sacramento, CA. The more specific you are, the less competition and the easier to target and win on keywords. The tough part is that being a strong force in the Sacramento, CA Bass market for new players from scratch is not a billion dollar idea.

Sometimes, more specific actually means more cost - in B2B, I've seen roughly $15 and $7 CPC for heavily contested keyword phrases, and know of a competitor that bids over $30 CPC for a particular phrase.

A slight aside, but although it's definitely market forces at work, I'm certain Google inflates CPC prices before those market forces take hold, forcing the hand of smaller players.

You kill the idea because it is so expensive, if you cannot easily reach consumers for a small amount of money then the demand side is low. It's cheaper to reach consumers who are willing to click links because they actually need a service.

Ansel from PSL here.

That's not quite right. Ad pricing is a marketplace like any other. The "demand" in the marketplace is demand by advertisers to get in front of any particular set of eyeballs. The price will be high if many advertisers are interested in the same group of people (and low if not). Price is an important metric for determining customer acquisition cost but you can't use it to infer customer demand. You instead use impressions, click through rates, and conversions rates to do that.

Generally correct, but do watch out for some recent changes by Google, where you pay what you bid instead of being charged the lowest next bid even if you set your bid higher.

They used extremely broad keywords and phrases that are by definition expensive to bid on. For example, one of the terms in the screenshot was "bass lessons".

What is that? Bass guitar or bass fishing? If this were a real business, not only would you be overpaying due to poorly phrased, non-specific ad title, you'd also be hit with Quality Score penalties by Google if you had too many users clicking through in search of bass fishing lessons and exited the page when they realized it was bass guitar.

$60/hr is actually pretty good, even in the bay area. Currently paying $40/0.5hr for 1:1 violin and $20/hr for group piano for my kid

Fantastic. Thanks for sharing. Two questions:

1. How much data did you gather before extrapolating to the numbers in your post?

2. What tooling did you use to estimate traffic?

There is no mention of Yelp or Craigslist as a means to find local help? And the analyst spent a week on this? Waste of time. Nothing here.

What tool did they use to come up with that demand generation spreadsheet? Is that just keyword planner or is there something else?

How was the data in the first table estimated (clicks, impressions, cost, etc)?

Google ads keywords planner. So many variables with deploying a real campaign and not optimizing end to end, so actual results can vastly vary with CPC, CTR, and conversions. If you build great ad campaigns and landing pages, you should be beating these estimates nearly every time.

Even with these keyword estimates you would never go full speed with ad spend until you optimized every step in the funnel on-going, then ramp up ad spend to scale in the following months. On top of that where are the display and youtube campaigns? Blaming these click estimates and ad platforms is not the reason of a failure, rather the value add, demand, and audience targeting. So many techniques to optimizing ad campaigns, I didn't get that was done.

Reply on news.ycombinator.com