Mistakes You Should Never Make (2014) | Hacker News Reader

Mistakes You Should Never Make (2014)

sethbannon.com

108 pointsgautamsomani73 comments

> Just as important are frequent, honest checkins to see how everyone is feeling on an interpersonal level.

That only becomes meaningful when there is already 100% trust. People learn very quickly (usually the hard way) that honest answers are not always in their best interest.

Yep, every time I've ever worked for a manager who wants regular 1:1 meetings "just to touch base" I pretty quicky just start feeding them platitiudes that I know they want to hear. They are without exception holding these meetings because of their own self-confidence issues, if they were concerned about me getting work done they wouldn't keep pulling me away from it.

> Yep, every time I've ever worked for a manager who wants regular 1:1 meetings "just to touch base"

Huh? Not even slightly.

I scheduled one-on-ones with my reports generally every week and no later than every other week. No exceptions for me.

My reports could always opt out of them if they chose (maybe they were super busy--maybe nothing needed to be covered), but I could do so only very rarely and I forced myself to reschedule.

It is vitally important to make sure that people reporting to me could get my personal attention even when I was super busy.

While I'm sure some of them "fed me platitudes", they also sometimes jumped up and down about something going wrong that they didn't feel comfortable bringing up in public contexts. That's really important.

Same with myself. And I can say my best manager of all time was someone I wanted to have those talks with, because I learned so much from him. Half the time we would talk about work, the other half we would talk together on various philosophical subjects.

Because of those chats, he learned that I was a knowledge sponge and loved trying new things, and as a result he challenged me. I got to implement some really cool middle ware that became the bedrock of the middle office trading systems.

That said, I have had bad managers where the 1:1 is a forced chore. My experience isn’t that 1:1 are bad, but that they will quickly reveal the character of the manager. I do agree that if my manager is not great, it is often best if we interact little and I can go my own way. And yes, I am lucky to have had jobs like that where I can own my path.

How often? Do they catch on to the platitudes?

Weekly meetings were the norm when this happened to me. I guess maybe "platitudes" isn't quite the right word but there just often wasn't much to say from one week to the next and saying "same as last week" was never enough, it was like we were obligated to spend 30 minutes talking about work because that is what was on the calendar. So a lot of it was repetition of stuff that my manager already knew (or should have) but for some reason wanted to hear it again week after week.

Ten years later this company, its product, the money that went into it, and most evidence of its existence barring this post and another article covering it, seems to have vanished into the aether.

The author of course seems to be doing fine at some VC firm. Just so you know what kind of behaviors Silicon Valley rewards.

> What kind of behaviours Silicon Valley rewards

Introspection? Honesty? Reflecting on one’s mistakes? Personal growth?

The author is not answering questions anymore so here's a link to the 2014 expose https://newrepublic.com/article/119350/amicus-app-how-tech-i....

> In plain English, this roughly amounted to saying: I, [Seth Bannon], have lost most of your money and my employees due to a combination of incompetence and evasiveness. I have no prospect of raising more capital, and my only chance of generating new revenue is a product that’s still unproven. Nonetheless, I’d like your blessing to keep going. It was preposterous pitch. But something strange happened after he made it: The investors more or less acceded. They weren’t going to kick in more money, but neither would they force Bannon out.

> ... No surprise, then, that at least one investor, Dave McClure of a Y Combinator-like accelerator called 500 Startups, was practically blasé in response to Bannon’s email. “[I]f you need an office in NYC to work out of, we probably have some extra space,” McClure wrote, cc’ing the group. McClure said he hoped Bannon had “enough trust / bandwidth / support to keep rolling.”

My take is that the apology in the OP is performative. The lessons are weak, both in an ethical sense and a business sense. But the words are a good shroud for their failures; it's a "learning experience" that they used to pursue a career in venture capital.

This is useful additional context, thanks.

> My take is that the apology in the OP is performative. The lessons are weak, both in an ethical sense and a business sense.

I am naturally inclined to agree, which is why I have to remind myself to keep trying to be optimistic about such things.

Move fast, break things, pretend to be sorry about breaking things, move fast away from the broken things, keep at it as long as there's someone else left holding the bag.

- never promise something that has not already been built.

- never put a timeline on something that has not already been built (#)

- sell what you already have

Software is incredible. you can sell one copy and instantly have another ready to sell. It's the best business in the world - don't make your life more complicated

(#) Oh ok. but give timelines with "cones of confidence" and ensure everyone else upstream and downstream does the same and can adjust their cones. It's kind of like how projects are managed today (hit this date or we panic) but with less wild swings - I am pretty sure there is some feedback-control logic thing there. Percentage confidence of delivering depenandcies enables the people expecting multiple dependencies to manage their panic levels and their resource allocation

(ie project mama gets need to be in every detail or just don't have project managers and have systemic feedback loops)

I'm the author of the post. Happy to answer any questions anyone has! So many lessons learned the painful way.

Thanks for writing this up, I found it very useful and I promptly went to go check that we're paying our payroll taxes haha.

fyi you misspelt startup as starup at one point

Thanks for helping me catch a 10 year old typo!

Did your cofounders forgive you?

Another mistake/very expensive hard lesson I've learned is to never bring people into leadership on potential. Not unless there are exceptional circumstances where you can confidently say the person, even lacking experience, will perform.

In "the hard thing about hard things" they echo this advice as well - usually your executive team should have previous experience with the exception of maybe one person.

Thanks for the book recommendation!

Do you mean high-level leadership? Everyone has to start somewhere after all.

There is much simpler solution. Just tell the customer you can have the feature in a short time, cost free. Tell them you value your relationship with them and you are always interested in making the product better with the feedback of the customer.

Knowing they are able to develop a feature quickly is valuable. I know it just becomes a discussion about pricing and priorities. Knowing they can be honest is also valuable. It reduces the risk I have stumbled on another dishonest company that just looks perfect until you sign the contract.

I worked with a lot of companies. I have dropped a bunch of them for lying to us because I am sick and tired of dealing with untrustworthy service providers. And if you keep lying about your features, guess what, you are bound to slip up be caught at some point.

“Entrepreneurs often write about what’s going right, but too rarely write about what’s gone wrong“

They do in retrospectives like this after they achieved later success.I found that a lot of rich and successful people in different domains love talking about pass failures after they eventually succeed.

You very very rarely see a post on here from someone who failed, and they do not know how to fix the issue or their next move in life.

Fwiw I published this post when my startup was falling apart because of my mistakes and I had no idea how to save it (and in the end couldn't).

Have you had any follow-up advice or epiphanies that could have reasonably helped save it, after the time of this post?

That’s honestly amazing. I apologize if my comment seemed specifically targeted to you.

It's weird how he wrings his hands over the damning half-truth of telling people he dropped out of Harvard instead of the Extension School at Harvard, but seems tolerantly baffled that his coworkers didn't like him outright lying to clients about their software capabilities. He's still defending that as a "common practice."

Isn’t the IRS generally willing to work out plans for paying back taxes?

Post author here. We did work out a repayment plan for some taxes. Payroll taxes are generally "if you have the money in the account, you have to pay immediately" though, so we wrote a large check for those as soon as we discovered the error.

What didn't make sense to me about that story is that even if their payroll system wasn't submitting the tax payments to the IRS, it seemed it was witholding the tax on the payroll. So the money was sitting in an account somewhere, they should have had it available to pay all the back taxes. They would only really be out the penalties, which could probably have been negotiated down on the basis of an "honest mistake." Though he did admit to sloppy bookkeeping overall, so maybe they saw all this money in the bank account and spent it on other stuff.

> I would often let potential customers think we already had a feature they wanted and, if they signed, would come back to the team and say “we’ve got to build this before they launch!”

And the takeaway is to make sure your cofounders are OK with this first. Maybe just don't commit fraud? If something is truly a dealbreaker then just be honest and promise to build it out with contract terms giving them an out if you fail?

Do these people use deceptively attractive pictures in their dating profiles? Then expect they'll just hit the gym and/or plastic surgeon if they can convince someone to connect with them?

> And the takeaway is to make sure your cofounders are OK with this first.

Let's not forget the employees who will actually plan and build the features. I've seen not one of these deliverables fail because executives forgot to tell their team in time. Or, in fact, tell at all.

It's easy to get overwhelmed with all the white lies and forget some, I expect. It will damage relationships and employee morale. The white lies are not free.

True. And you're likely to lose your team's trust to. If employees see a leader telling white lies to potential customers (or to anyone) they're likely to ask "I wonder what white lies they're telling me?"

It's also "what else have we promised that I'll have to deliver but don't know about?" and "what else will the executive promise before this ships?"

Those are the real morale killers - knowing that you might need to start stressing and crunching at any moment, and that the roadmaps you have aren't real. Nor are your commitments to anyone working in the company if you get pulled off your planned work to service a white lie. They just complicate everything.

An executive lying to customers is a huge liability. I won't take an ethical stance on it, but in a business sense, the lies incur a debt that must be paid.

A coworker said he worked for a person who did this routinely, inventing whole offerings in the moment. Then demand the team crunch to deliver. Supposedly this was their idea of proving the market.

Post author here. If a customer asked "and I'll be able to upload my own data?" and the response was "absolutely" even though that functionality didn't exist yet because we knew we could build it before they launched, that's nothing approaching fraud. I don't believe we ever once didn't have a feature like this that a customer wanted before they actually launched. The disagreement is whether it's ok to just say "absolutely" or whether it's important to say the full "just to be clear, the product can't do that today but we can build it before you launch." 10 years later I'd take the latter approach.

I've been on the other side of this. We talked to this startup about their product and they claimed it had exactly the features we needed. We said fantastic and were ready and willing to hand over $10-20k to them for this, and they just kept stalling and making up excuses and demoing things that didn't actually do what we actually needed. Eventually they had to admit that they couldn't deliver the feature we needed. We were rather pissed off and cut all future dealings with them. They could have ended up with at least $50k from us over 2-3 years if they'd been honest and showed willingness to work with us, and instead they got zero.

Great example of how sacrificing integrity to potentially cut corners is long term self-defeating.

'Absolutely' is strongly affirmative wording. I'd say it's fraudulent even from an engineer deeply familiar with the level of effort. Language matters and executives should know better.

I know from personal experience it's hard to avoid speaking too quickly, yet it's also important to self correct in the moment. Otherwise why have such people on a call if they cannot communicate accurately and won't even admit when they misspeak?

Who is the “we” here? Because it sounds a lot like _you’re_ assuming it can be be built for the people who are going to be tasked with doing so.

My experience being a leader in similar situations is that when I say "we will have that in the next release", I mean "the engineers who are tasked with that will have completed it OR I will personally do it."

It absolutely is. What is wrong with you? It is really normal to lie to people on their face? I mean, you can perfectly say “The feature is not ready, but will be at X time” Ready, no need to be dishonest.

On the other hand, Bill Gates famously did this to IBM regarding DOS. When it works, it works. Sometimes pressure makes a diamond, as many of us with ADHD procrastinstion know

I don't think this is an endorsement of this practice. Consider that Gates then spent the next couple decades engaging in anti-competitive behavior that IMO did a lot of harm to the personal computing industry.

If anything, this tale about Gates just tells us that you can "win" in unethical ways by continuing to do unethical things throughout your career.

As far as business ethics, I feel like a little overpromising with legitimate confidence you'll actually deliver isn't the worst. It's high-risk high-reward.

>I would often let potential customers think we already had a feature they wanted and, if they signed, would come back to the team and say “we’ve got to build this before they launch!”

Or as most B2B companies call this, "a typical weekday".

Also make sure, that the people who need to develop it are on board. It is a bad feeling, when there is time pressure, just because management thought they could promise something that does not yet even exist.

Company did not survive. Appears he is still citing Harvard on his LinkedIn https://www.linkedin.com/in/sethbannon? Even though he calls this a mistake in this ten year old blog post

It says Harvard university extension school on his profile. That’s what he said he would start calling it in the blog.

He is citing Harvard University Extension School on his LinkedIn.

He says he regretted implying that he dropped out of Harvard College.

I'm not sure I agree that just saying "I dropped out of Harvard" is a bad thing. After all, Harvard is putting their name on the "Extension School." If they didn't want to trade on their reputation, they could have called it something else.

I got my Master's at Harvard Extension School. It doesn't feel right to say "Harvard", though I sometimes say "Harvard" if I don't want to get into the usual conversation about what the Extension School is, but only if it's not important to the discussion and there's no harm done. Sometimes random online forms ask what college I attended for no good reason, so I'll just type "Harvard"; if Mark Cuban would ask, I'd always say "Harvard University Extension School" to avoid misleading him.

What's an extension school? That part of the article confused me. Is it not a part of Harvard?

It's usually a satellite (or even virtual) location (or locations) that the parent university administers. Extension schools usually cater to past-college-age adults for continuing education, are often part time, and are usually commuter schools (and don't really have a live-in campus culture). I can't speak to any differences in quality of instruction, but I wouldn't be surprised if the admissions requirements at most extensions schools weren't as strict as for the parent university (which makes sense, because people attend for very different reasons, often at different stages of their lives, with different backgrounds than you'd expect for a high school grad going to an undergraduate university).

I think, though, that the disdain for saying "I went to Harvard" when you've attended the extension school is rooted more in emotion/ego/pretension than anything else. A "real" Harvard attendee might not actively look down on someone who attended HES, but would probably scoff at the idea that they "went to Harvard". Higher education has a lot of weird social signaling and other crap that's hard to unpack.

But I do think that the accepted meaning of "I went to X" is that you physically attended classes full-time at the university's main campus. Saying that, when you attended their extension school, is at best misleading.

This pretty much sums it up. As an immigrant, I had to work full time and study late at night. In terms of differences, at the time, half the courses were the same as Harvard Engineering School, and those were the best courses in the program with the best professors. The other half felt like they were just filling up the curriculum. Overall I'm a much much better developer after having attended Harvard Extension School, though I went for the hardest courses. It's certainly possible to skate by and get your degree by only taking easy courses.

> Company did not survive.

He's now "Founding Partner, Fifty Years [Seed fund], 2015 - 2023"

> Appears he is still citing Harvard

Yes, because he did study at Harvard's "Extension School", which is properly cited on the page.

Due Diligence.

Don’t forget that the amount of scrutiny that an accepted YC startup gets amounts to their YC application being read a few times, a ten minute interview with three partners, a background check and another ten minutes for a decision.

This is a lot less than a regular job interview. Consequently, anyone who can sing a song that YC wants to hear has a chance of passing.

A single bad investment, in a market where most firms fail but a tiny minority deliver outsized returns, does not ruin the entire portfolio. Conversely, a single bad hire can ruin an entire team, department, or company.

Once you get accepted they do further diligence (that can take several weeks if you’re missing things or have been sloppy) to get it all together before they give you the money and let you do demo day.

> In the early days I would often let potential customers think we already had a feature they wanted and, if they signed, would come back to the team and say “we’ve got to build this before they launch!” No harm, no foul, I thought, so long as we knew we were able to build the feature before they started using the product. This is a tactic commonly suggested by lean practitioners. My co-founders, though, would often frown on this behavior, worrying it was unethical, causing a huge amount of tension to grow beneath the surface.

I'm not a lawyer, but I'm pretty sure this is the same kind of fraud that Holmes got nailed to the wall for committing.

It's entirely possible to close deals without lying, if you can still deliver the needed functionality in time. Why would anyone do this?

If you are competing against folks who always tell the truth, a willingness to lie will get you some short-term wins. He said it is “commonly suggested by lean practitioners” whatever that means.

Personally it makes me want to figure out what a “lean practitioner” is so I can avoid them like the plague.

From experience, this is quite common in enterprise sales. Having been on both sides of such transactions, I’m now quite sceptical of features/functionalities unless I can test drive myself. Some stuff I’ve seen:

- chatbot sales: “see how it recovers here because I misunderstood the question and gave a different answer than expected?”. (Demo was a linear path with this ‘recovery’ hard coded) - Buy side: “yes absolutely we already integrate with X”. Queue 12 months of delays because they had never integrated with that platform before.

But, I think I get it. The deals are huge and you only sell a few a year. Once an enterprise commits, it is unlikely to switch vendors unless it goes really really bad. So, the salesperson that convincingly lies gets the deal, and surprisingly the downside is relatively limited.

OTOH, not massaging the truth might mean losing that deal. Huge downside, very limited upside. (Yes the buyer might be a bit more happy, but that likely doesn’t outweigh sometimes missing a multi million deal, unless buyer talk a lot with each other).

Guess it all depends on how you value your company's reputation versus short-term profits. I've personally worked on an API integration with such a company who delivered less than they'd promised. We certainly won't be building anything new in the future with their platform. And even this existing project will probably migrate away when feasible (though that, admittedly, will be quite a while from now, given how busy we are).

"lean practitioner" probably means the lean startup types. A set of ideas ostensibly based on Lean Manufacturing (aka, Toyota Production System) that, like Lean Manufacturing, are often implemented by imitation rather than by comprehension and serious consideration.

There's a massive difference from saying "yeah, this thing has keyword search already, we just didn't turn it on yet" during a software from giving a diabetic saline while claiming it's insulin. Holmes did the latter. People actually expended additional bloodwork and treatment due to the false testing.

> I'm not a lawyer, but I'm pretty sure this is the same kind of fraud that Holmes got nailed to the wall for committing.

Interestingly, she was found guilty of defrauding investors, but acquitted with regard to defrauding patients: https://en.wikipedia.org/wiki/Elizabeth_Holmes

It's not the same as what Holmes did, but it is dishonest.

All good relationships, including business ones, are built on a foundation of trust. If you don't have that, it will fall.

There is a saying:

”The law, in its majestic equality, forbids the rich as well as the poor to sleep under bridges, to beg in the streets, and to steal bread.”

- Anatole France

Because even customers don’t always know what they want. A lot of customers want peace of mind when they’re buying a product, and you can’t always build up that trust in a single call, so a white lie that gets them as a customers as long as you can fully deliver on that promise seems optimal for both parties.

Telling a customer “we don’t have that yet” knowing full well they just care if you have it, and knowing full well you can deliver it as promised, just creates unnecessary doubt that could lead to a worse situation for everyone. One where a customer passes on you because they don’t know you well enough to trust you can deliver, and you don’t get the sale because you didn’t happen to create some very simple feature before a customer demanded it. That’s a lose-lose.

This complaint just reeks of “I’ve never worked in a company or role that didn’t already have all the uncertainty figured out”. Building features nobody wants as insurance in case somebody finally asks for it is a waste of human talent.

>I'm pretty sure this is the same kind of fraud that Holmes got nailed to the wall for

Yes but she got nailed for violating FDA laws. Saying your software can do something before it's even on the roadmap is what I've experienced in every org I've worked in... overpromise and if the oompa-loompas can't deliver it's their fault not mine

Lying to customers isn't good, but lying to investors is a crime against rich people & will certainly cause trouble for you.

ITT we’re all learning why people don’t honestly share their mistakes. Be better folks.

Url changed from https://web.archive.org/web/20140803073308/http://sethbannon....

Submitters: Please don't make archive links top-level submissions unless you've searched and there's really no other option.

https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

Reply on news.ycombinator.com