Hacker News new | past | comments | ask | show | jobs | submit login
Before Growth (samaltman.com)
364 points by firloop on Jan 15, 2016 | hide | past | favorite | 206 comments

This reminds me of pg's comments about "playing house". The appearance of a startup is more important than the reality of a startup, for founders whose core drive is to "be a startup founder", not "create this product the world desperately needs so maybe it will stop haunting my dreams".

I've met founders like this, recently. They're building some sort of social app that does something no one needs and no one would pay for. Maybe I'm being dense and they're actually geniuses, but I don't think so. Most startups fail, and I think a lot of that is because a lot of startups are really stupid. They're "solving" worthless non-problems because the founders aren't doing the hard work of finding real, valuable problems to solve.

Building a great product is really hard. And it can't all be done completely lean - at some point, you need to envision something so profound that people don't realize how much they need it. You don't just write Hello World, charge a buck, growth hack, and brag about how you got two bucks for it the next week. As Henry Ford said, if he'd asked customers what they wanted, they'd have said faster horses. Don't build a faster horse. You won't be great that way.

As an aside, reading posts like this makes me feel a little better.

I'm 31, have a great job etc., but have always wanted to start a company.

The problem is, I've never encountered a problem big enough that would want me to do this and rather than start a company for the hell of it, I've just stayed put.

What is the right move in this scenario? Start a shitty company, or keep the nice job and potentially regret never starting a company?

Viewed from this perspective, it's not as clear.

I wanted to post this as a child to @gist's comment, but since there are a lot of comments similar to yours, I'll put it as your child.

>"I've never encountered a problem big enough that would want me to do this and rather than start a company for the hell of it, I've just stayed put"

Fuck[1] those who make you feel bad that way. It is RIDICULOUS that working reasonable hours a week, doing a challenging (or not), but stable job, great benefits, and being rewarded for loyalty is now LOOKED DOWN UPON! Your entire life is not a challenge to see who can push the hardest, 'solve' the most problems, start many companies, etcetera. You should do what makes you happy. Honestly. And if that means starting a company, go for it. If you're not so inclined, don't fret over it.

I have realized that the problems will come to you if you look for them. Like, random ideas. For example, you will see in my comment history that I often suggest ridiculous business ideas as solutions to my problems/something I want. If you want something/aren't satisfied with some kind of stuff, maybe you should work on a solution, and try to sell it to people with problems you faced? You may not make a lot of money, or any profit, but might make yeah happy, and give you the satisfaction of having started something.

In a related note, in one of my previous comments I mention JuJu Dhau, this awesome/incredible yogurt from Bhaktapur, Nepal. If you/anyone wants to start a yogurt business to make that (I'm assuming U.S), I'd be a sleeping/reasonably active partner for a little bit of money. Just throwing ideas out there... : )

1. Does HN have a swearing policy? In either case, I mean in the figurative sense anyway.

Does HN have a swearing policy?

I don't think so. I think as long as you aren't cussing "at" someone, you are probably fine. Personal attacks are not okay. That is not what you are doing here.

A mod actually changed the title of a submission I made to include a swear word when I posted an asterisk-laden version of it [1]. So I think you are correct that there is not a swearing policy.

[1] https://news.ycombinator.com/item?id=8897764

I'd say that the 'original title' policy simply trumps the 'gratuitous swearing policy'.

Just don't attack people personally and you'll be fine, though a user that excessively swears without purpose would likely be chided (is there a keyboard version of Tourette's?)

Does HN have a swearing policy?

There is no official policy on this to my knowledge. But my feeling is that expletives indicate and incite strong emotions that can cloud insight into the complexities of an issue.

I think I'm not alone in that.

I'm in that "I need to build this so it stops haunting my dreams" category. And I'm a lot older than you. After 20 years in the industry, I kept having the same problem over and over, at job after job. One night, I thought "Man, I wish there was a product for this, but no one makes it." The I thought, "Hey, I could make this! And lots of people have this problem, and it's a really expensive problem, so I could make money at it!"

Three years, a lot of personal cost, false starts, finding co-founders, and other headaches later, and I'm finally getting to where this year, it should start making money. I can't even begin to describe what a difficult experience it's been. Someone once told me the only emotions founders experience are elation and terror. What they didn't tell me is that they are often at the same time, and often for the same reason, and sometimes are indistinguishable from one another.

I'm a founder because I couldn't live with myself if I wasn't. Doesn't mean it's easy, or it's going well all the time.

So that being said... if you want to be a founder, and you're a good software engineer (I'm assuming you're tech, that's why you're here), don't even bother coming up with an idea! Instead, go looking for experienced business people who have a real business idea and the knowledge to pull it off - except for the technical skill. Become their tech co-founder. Because there are so many of them, and so few of us, you can afford to be choosy.

If it's someone who has successfully started businesses before, all the better! There are lots of really smart, driven, reliable business people out there with great ideas, but without the technical skill to implement. Find one, and make something great.

Matchmaking here seems like a challenge—you're dependent on already having a big network, with the right kinds of people in it.

There are a lot of ways for it to go wrong, eg business competitions, hackathons, etc.

So I looked at your product. How do you deal with the firehose of data issue making it too hard to figure out what happened.

Hey, I just wanted to say that I checked out your company, and I think you have a very compelling product. Unfortunately, I'm not in a position to buy it (junior engineer at a massive corp). Instead I just wanted to wish you luck!

I have a very compelling idea, and a very compelling problem. And a product that is not yet sales-ready yet. But we're getting there. It's been in a fundamental rewrite mode since August (what I thought was the beta was actually the alpha), and I've since added two co-founders so I'm not doing it alone anymore - deeply experienced enterprise engineers like me who have felt the pain that we're trying to solve viscerally.

But yes, thanks for the confidence! :)

The best part of it is when I talk to the target users about it and their eyes light up and they get so excited. The problem is huge, and the current solutions are terrible. If we can't make money at this, it's an execution problem, not a market problem.

Had a look, and yes, that's not an easy problem to solve. Good luck with it.

How is your product different from etckeeper + flyway + source control?

First and foremost, it monitors the real world, not the version control system. It sees things as they are, not as they're supposed to be. It detects manual, unversioned changes. It detects changes from automation, or from intrusion. It detects things that can't be versioned, like reboots or cert expirations.

Second, it is minimally invasive. You don't have to change your process in order to use it.

Third, it shows relationships between systems. An application isn't just source code. It's compiled binaries, OS patches, user security, databases, firewalls, message queues, apis, and more. Here's a common failure pattern - System A changes, System B breaks. System A might actually be perfectly fine and correct, and B as well, but the relationship between A and B is broken. Change the database schema, and the app can no longer read it. Change the dns, and the UI can no longer find the API. For any decently complex application, the potential sources of failure are in the thousands, and potential interactions effectively infinite. Now scale this up to the enterprise, to organizations so large that your app can be broken by departments you didn't know existed...

Does it detect when you have a busted marketing site?


Thanks. It's a known problem... we're currently in product work and haven't been dealing with marketing at all. It's just a cert problem - http://congruence.io still works.

And yes, once it's working, it will be able to detect such issues.

And the certificate is now fixed. Try it, https://congruence.io is working again!

Part of finding great problems to solve is cultivating a sense of taste for a process/product/service/thing, so you can say, "[x] is a hack and could be done a lot better."

To develop that sense of taste you have to go deep (usually) into the thing.

Using the Henry Ford example, he went deep into engineering and gasoline powered engines.

Maybe people did want faster horses, maybe the limitations of horse-based transportation was a huge problem of the day... Either way Ford was able to say:

1) Horses are a hack and transportation could be done better 2) Cars are the answer but current manufacturing practices are a hack and could be done better 3) Employer-employee relationships are a hack -- turnover is way too high and must be done better. Profit sharing and high wages for factory workers are the way.

By the way, Henry Ford was an avid cyclist ((and lots of other car pioneers were, too). So comparisons to bikes might be more appropriate than horses.

"or keep the nice job and potentially regret never starting a company?"

What's unfortunate is that that regret comes almost 100% from the fact that you spend your time at places like HN or Tech Blogs or business press that are brainwashing breeding grounds for having that type of regret. You feel that you would have regret because you think it's on the dartboard to achieve the type of (gambling I might say) success that you read and hear about. (As opposed to playing in the Super Bowl). Because while the founders are smart they don't seem smart in such a way that you think you couldn't achieve the same success if given the chance, right? But what you have to realize is that there are so many things that go into this oversized success that it's a complete crapshoot even with an idea that is good. Bottom line is if you are going to spend your time reading about something like this you will inevitably wrestle with that regret. Which is really unfortunate.

You are still very early in your career and very far away from realization of your regret. Find someone you want to work with or a problem you want to solve, ideally both. You can accelerate the process by working on interesting side projects with different people. When you do start a company, people will think it's shitty. But do not intentionally start one you think is shitty too.

Similar age and situation but have started a couple of companies before. I spent some time thinking about ideas last week to keep the mind fresh. At first none of the ideas seemed worth anything, many "also rans".

Then I started to think about all the things that are broken at my current company or which are frustrating and why. Within a few hours I had a dozen solid ideas that could be more broadly applied.

If I wanted to start a company again at this stage (not ready for another couple of years at least) I would rank these ideas by ones I could start on the side at first, or which justified leaving my job. The ideas that could be started in my situation and met my passion + other factors I felt important would be the one I would actually work on.

In the meantime I'm saving my pennies, learning as much as possible and having a blast.

If you already regret not starting a company, then start one.

Generally you would want to give this "experiment" a realistic timeline (ideally, multi years). Set goals on when you think this is or is not working, so that you can bail. This is very similar to making bets on the stock market, except that you are investing in yourself.

Now, look at your worst case scenario (failed startup). Things you will MOST definitely learn:

- How to start a company?

- What does being your own boss mean?

- Independent thinking

- Make some new friends/contacts and meet other people.

Things you will lose:

- Time

- Money

- Balanced lifestyle, perhaps

If you think, you, as a person, are the same or better, then you've won half the battle.

You need to be sure that years down, knowing the worst case scenario you dont have a new set of regrets. Also, ideas are a dime-a-dozen, its all in the execution.

I'm 29, have no job, and have started five companies, all of which have failed.

So, if it helps at all, when you feel the pangs of regret at not starting a company, you can think, "Hey! I'm doing OK. I mean, I'm doing better than that one smart guy on HN (everyone on HN is smart)."

Grass is greener, etc.

Why did you start your 5th company with all that experience with failure? Did you have something that you thought was special?

If you got fired from four jobs in a row, would you give up working?

If you got fired from four jobs in a row, it's probably because you already gave up working five jobs back.

He now knows 4 things not to do!

5 failed companies is not an especially high number, in SV or the broader entrepreneur community.

Build your brand; become a The Guy. It's less obsessive.

In other words, truly master something, then write about it well, to where you have a reputation. That's far more than darn near anyone does and you don't have those long nights while your baby is in a potentially fatal fever.

Recently had an idea for a product that seems amazingly obvious (but I would have never had with different life circumstances), one of those "this has to exist" moments followed by "why the hell doesn't this exist what am I missing".

Still not sure if it's an amazing idea missed because most people who start startups are in a particular demographic or I'm completely insane.

I think I'm going to have to build this one to actually find out...

I certainly don't think staying in a great job is a bad decision at all, but I did want to just mention something that your comment brought to mind. A lot of people appear to think that in order to start a (good) startup, you need to solve a problem that hasn't been solved before. I would argue that it's often much easier, and potentially just as lucrative, to solve a problem better than it has been solved before. Thinking of a really meaningful product that no one else makes is hard. But imagining how to do something that's already done, but do it better? That's much more do-able, and could still have a significant impact.

So, I'm not pushing you to start a company at all. But if you do, don't limit yourself to "unsolved" problems!

to me, the compelling feature of solving a problme better than an existing company is that they already have a market tested and proven exists. finding a market is actually a non-trivial problem that often seems to be over looked!

>What is the right move in this scenario?

It depends on the details but maybe stay till you see an opportunity that seems a no brainer? I probably helps if you have a job where you interact with customers - then you may meet some saying we really need an x and would be willing to pay but can't find anything. Then jump ship and build x.

Slightly different but I know a couple of guys who were head of something at smallish companies and then when they were taken over / fell apart they spun off their own businesses. It's low risk because they know the market, have personal contacts with the customers and so on. Not very tech startup though - travel and marketing. But maybe try to get in to that sort of situation where you are effectively running a small business within a larger company.

When you decide to start a business, often times, you'll be doing lots of things you don't really want to be doing for that business. Like filing paperwork with the state or making sure your taxes are done on time. This stuff adds up and there is a lot of it.

To start a business, you really have to enjoy the product you are building enough to cover the work you don't want to do and enough to no longer do some things you enjoy doing in your life, as you may no longer have time for it.

Starting a real business is a challenge. I would suggest, if you do have an idea, try to build it. If you can build the idea, you might want to consider starting a business around it.

Starting by building is a great way to waste a huge amount of time on solving a problem no one has. Start by talking to potential customers and growing an email list, not building. Aside from being less risky, it's way more fun to work on something that you know people want and are waiting for.

It also conditions you for the fact that you'll need to spend at least 50% of your time on squishy marketing stuff to get anywhere. You can't just code your way to getting customers.

Building a product is easy. Especially if you are an engineer. Getting people to pay for a product now that's where a business sprouts from. Just because you built a product doesn't mean you have a company/business. You need users and those willing to pay for it. Most realize this when it comes to raising Series A.

Do something in between. If you have a great job that pays well expand your network and start investing excess cash into startups really early on. This way you dip your toes into those waters and get to see all the dirt and help them clean it up. Along the way you might find an interesting opportunity/problem worth solving yourself that could turn into a future business.

Look at this as "preparation". Far far better than following the herd coz it's the thing to do "to start a startup" and ending up with nothing after few failed attempts.

Why do you have to solve a big problem to start a company? People start companies that do run of the mill things all the time and there's nothing wrong with that. To a certain type of person, starting and growing a company is inherently more rewarding than working for someone else, regardless of whether the company is doing something sexy or novel. If you might be the type of person who enjoys autonomy, responsibility, owning/growing/managing something, wearing a lot of hats, and so on, you don't have to wait for a world-changing idea to go into business.

This really depends what is driving you to start your own business and what you are passionate about.

Are you happy with a $2million revenue a year business, or do you want to be the next unicorn?

You might find a large problem to solve, but is it some thing your are passionate about? Or perhaps you love selling antique nail clippers for example, which is probably a very small market, but you love doing it and supplementing your day job income.

Find a problem, decide if you are passionate about solving it, can you make money from it or not, then build a solution.

The default best move is to join a young company. I never had the luxury of joining good companies so I started one. Given the choice, I'd join a good company any day. The odds of success (as in actually making more money than working full time) is about 20%.

What I'd suggest is start working on side projects. Scripts that make something easier. A tool. Side projects like todo lists and note taking apps turn into billion dollar companies.

How to get startup ideas: http://paulgraham.com/startupideas.html

Find a problem. What would you pay to have solved? Now out of that list, what do you have the skills and experience to address?

I think the "what would you pay to have solved?" question is why business-types are often better than technical-types.

I meet very few technical-type hackers who aren't absolute cheapskates with any products and services that they can just clobber together with a shell script or two.

Therefore it may be often more helpful to ask "What would my friends pay to have solved?"

Laziness is useful here. Lazy developers are among the most inventive, since invention generally involves wanting to automate something that you formerly had to do manually.

Don't forget to ask "Am I going to get to $50 million in revenues before a big company sues me?" If not, recognize that room for your existence is limited.

So far I haven't personally seen this problem in the field. I know it exists to some extent but is it really that common? How many promising startups are killed by lawsuits?

If Travis Kalanick had followed that rule, we wouldn't have Uber today.

Uber came out of the gate with war chests from investors. Real startups that have to make actual money usually don't have investors that'll throw tens of millions into legal fees.

If your investors have indicated that they will bankroll the massive legal and regulatory issues that your company will provoke, and you believe them, then you can ignore that rule. If you're a normal person trying to start a business on your own, you shouldn't.

They didn't, though. They closed their Series A in Feb 2011. They got their first cease & desist in October 2010 [2], 4 days after closing angel financing. [3]

IMHO, the key success factor for Uber is their ability to spin every setback into an advantage. So, the fact that they were sued for operating like a taxi cab company means that the taxi cab industry views them as a threat. They could then take this to investors and say "We're about to disrupt a $16B market. The market opportunity is there, and the lawsuit against us is proof of that. The product is there. People are choosing our product over the competition. We need your help to roll back protectionist legislation that is hurting consumers. Are you in?"

And then many investors won't be, but they just keep trying new ones until they find folks who buy in.

[1] https://www.crunchbase.com/organization/uber/funding-rounds

[2] http://techcrunch.com/gallery/a-brief-history-of-uber/slide/...

[3] http://techcrunch.com/2010/10/15/ubercab-closes-uber-angel-r...

> Uber came out of the gate with war chests from investors

How the hell did Uber manage to convince investors to fund them so before they showed success?

There's nothing really technically special about Uber so I doubt that investors were lured by the strength of the proposal or anything like that. The founder of StumbleUpon was a co-founder, so it was most likely through his connections + track record, but I don't actually know the details (someone else here might). I would guess that it's just an accident of luck that Uber became Uber -- right time, right place, and right connections. That's what it takes to get a business off the ground, especially a business that raised $1.5 million in funding before it was released (per Wikipedia's timeline).

Kalanick's previous company Red Swoosh sold for $19m. They were probably backing him as much as anything.

you could start a consulting company

I've always believed that some of this "playing house" is fueled by the fact that a lot of startup founders are young and lack years of direct experience with the various problems industries face (I say this as a 28 year old).

Of course a bunch of twenty somethings gravitate repeatedly to social apps; one of the core focuses of their lives to that point, whether they know it or not, is to learn how to be a social creature.

As you get older and more experienced, you start to unearth the "real" problems in the world oftentimes in some niche you could have never predicted you'd be involved in your younger years.

I could be completely wrong, just something I've always pondered when seeing so many of these ridiculous startups that keep churning out of VC land.

The problems faced by businesses aren't actually any more "real" than social apps, they're more differentiated. What tends to happen as you get older is you start digging deeper into specific fields. Instead of looking at everything from the glossy overview perspective, you end up looking at the details, and finding specific tasks that can be done more efficiently.

I suspect this is why younger entrepreneurs tend to build bigger companies, but have a lower chance of success. The size of an eventual company comes from the size of the market it's serving. Basic needs - food, socializing, sex, information, transportation - are shared by nearly everyone, and so the markets are huge. But everyone can see any problems that arise in those areas, and so the competition to win the market is intense.

Building something like software for petroleum geologists working in undersea basins is a much tinier market, but many fewer people have the requisite domain knowledge to even understand what the problems are. So your chances of success are higher, but you aren't going to strike it big like Facebook.

Yep. I'm currently solving problems experienced by senior/lead engineers on large enterprise systems. I know about the problem because I've been that engineer, and know a whole bunch of my peers have the same problems. If you haven't been a lead engineer on a big system, it's not part of your experience - it's not something some fresh grad would work on, period.

Even better startups come from deep business knowledge. Not necessarily unicorns, but great lifestyle and niche businesses that make the founders rich. And really, do you need to buy two Lear jets?

Personality type might also be a factor. Not all young founders "play house".

There's always been silly startup ideas, but that's mainly because a lot of people are silly. The "playing house" thing seems to be a different phenomenon: over the last few years I came across people who were very smart, very driven, but completely focused on other people's opinions. They like winning contests and climbing ladders. Now that startups are becoming a respectable career path, more and more of these people are trying to start startups.

In the corporate world, you can always focus on the next presentation, the next interview, the next hurdle, and gradually iterate upwards. To build a company you need to think in terms of fundamentals, to think long-term, and to think independently: members of this type can be very good at producing polished and sensible-sounding market analyses but distinctly bad at those three.

I've used the phrase cargo cult startup to describe this phenomenon, (after Feynman's phrase "cargo cult science"): going through the motions in the hopes that your intended result will happen, without understand cause and effect.

Argh, typos. For the record, here's what I meant to write:

I've used the phrase cargo cult startup to describe this phenomenon (after Feynman's phrase "cargo cult science"): going through the motions in the hopes that your intended result will happen, without understanding cause and effect.

Excellent insight and description; thanks. Going to use that as needed. Now that's an idea worth spreading!

Yeah if you read https://en.wikipedia.org/wiki/Cargo_cult#Post-war and swap "investor money" for "cargo" it works quite amusingly.

Did it, agreed :). Re. "cargo cult", interestingly, the way I got to know of the term, was via a National Geographic article from years ago. [1] They actually sent a team to those islands in the Pacific (Melanesia, maybe) and documented the cults live, with text, maybe interviews and definitely photos - airstrip, trucks, people ... I still remember some of those things they showed and they match what is mentioned in the Wikipedia article. Think I even remember the term / cult John Frum (and just guessed, correctly, that Frum means from).

[1] This is from long back, when the mag was extremely good, with those stunning photos and great write-ups of places and things and people of all over the world. Later it became thinner, with shorter articles, less photos, etc.

Slightly off topic but it seems Ford never said if he'd asked customers what they wanted, they'd have said faster horses.

In fact when he started experimenting trying to building cars in 1893 it was eight years after Benz had started producing his Motorwagens in Germany. If he'd asked his customers what they wanted they probably would have said a cheaper and better Benz which is kind of what he produced with the Model T fifteen years later.

The Model T was a great hit, costing about $22k in modern money, much less than his competition. He continued driving the price down and by 1916 it was about $7k in modern money and took 2/3 of the market. However in the 20s General Motors started offering more luxurious cars with deluxe features like a roof and it turned out customers were happy to pay a bit more for that. It wrong footed Ford and their share dropped to 15% by the time they reacted.

So he took the market by building what customers wanted at the time - a reasonably priced car and lost a lot by ignoring what they wanted later - a reasonably priced and comfortable car.


I agree with everything you said, but want to offer an addendum.

There is still a ton of potential out there for faster and more efficient horses. You won't be a billionaire or on the cover of Time, but you can become insanely rich and seriously contribute to a better world.

The last job I left was at a medical diagnostics company focused in a niche oncology market, it started out trying to invent a crazy new diagnostics tool, failed, and pivoted to a somewhat traditional medical lab. We didn't invent new technology, we didn't have some insane research advantage, we heavily focused on customer service. We bent over backwards for our clients and remade traditional lab reports so they were much easier to read and 10x more visually appealing (our CEO's approach was "we are Neiman Marcus, Lab Corp and Quest are Walmart). The cofounder and CEO made close to 9 figures between IPO and later selling to a large multinational.

> for founders whose core drive is to "be a startup founder"

Yap saw that first hand. Someone read too much HN, wanted to run a startup. They didn't even have a good idea, but that didn't stop them. However, the was a minor problem and that was they needed investors.

Well as it turns out, if there are investors who are even more stupid than the founders and are willing to bankroll that, it's not all bad. Just take their money. Convince them you could be the next Facebook, look at the billions Facebook made!

So the startup is just a proxy to take investors' money, and play CEO for a while, and then when it fails it becomes a resume line "CEO at Uber For Dogs inc."

And yeah they didn't really have an idea, they just sort of came up with on, basically what they thought investors would like. For example, say, investors liked dogs, so the idea he came up with was Uber for dogs.

Hello. I am an investor. I'd like to contact you privately about this "Uber For Dogs" idea.

Well hello! My business plan is to become a unicorn by Jan next year. There is a hot market for dogs wanting rides to the park while their owners are too busy. For the seed round of just $50M you get 29% of the company.

First off I'll need to move to San Fran then rent an open floor office, preferably in a warehouse-like building with red brick walls. Then I'll hire 20 young people, preferably hipsters, with lots of energy. Experience not imporant, as long as they look good in hoodies and jeans, and know how to look busy typing and pretend to collaborate with each other all the time (because it looks good in pictures). Also must know how to correctly use latest memes. The tech stack will be all Go with reactive interfaces. Mostly microservices, nested inside lightweight Docker containers.

So there, welcome to the future (I'll send a link soon where to wire the funds).

Funniest damn thing I've seen all day.

Well played, sir.

In a world where "Color" received $41M in VC funding, I'd argue that solving worthless non-problems has no bearing on cashing out.

In a world where people like Dixon advocate having ideas that "look stupid"[1], it's no surprise that even the ones who don't want to cash out but genuinely want to build something useful follow the stupid path.

Because, let's face it, for any startup with VC money plans, the primary goal is not building a good product, but convincing VCs they're building a good product. The appearance of a good product is more important than the reality of a good product.

[1] http://www.businessinsider.com/the-secret-is-to-have-a-stupi...

Well, to be fair, Color did not make anyone any money. The company sold in a fire-sale for less than the amount raised; and the only thing that founders / employees likely got out of it was a mediocre retention bonus.

Not to mention, the idea wasn't necessarily worthless. It wasn't any worse than, say, Instagram. We only remember it that way because Instagram succeeded and Color failed.

What was Color's idea? I had heard of it at the time, including the talk about the high funding for what may have been a simple idea, but did not check (partly due to reading about it being overhyped, probably). Something to with color and photos?

Color was an "elastic" social photo network based on shared locations. Elastic meaning that instead of having binary "friends" and "non-friends," your friendship with people is was a algorithmic strength value based on how often you shared photos from the same location. (Which would map to how often you were together or at least at the same events, presumably mapping to how close you are in the present.)

Interesting, thanks.

>for any startup with VC money plans, the primary goal is not building a good product, but convincing VCs they're building a good product

I'm not sure that's true. For angel investors maybe but come VC time most companies already have traction and the VC money is to go from $10m value to $1bn. If the product's bad the founders get wiped out so it's not so helpful.

The idea is that great ideas look stupid; but they aren't stupid.

Good ideas may look good, but good-looking ideas capable of creating huge companies (thus great) are almost certainly tested already, so you are either wrong or late to the game.

Of course, YMMV. Nobody ever agrees on what an idea looks like.

Depends what your goal is as well, I've no interest in creating a startup with the potential to be a huge company, 1) those ideas are really hard to come by 2) they will inevitably attract competition 3) unless they are not trivial well funded competition can and probably will catch you rapidly 4) first mover advantage works on a product that has network effects but not much else.

The product I want to build next is a niche product for a very specific (but quite large in absolute) terms demographic, profit per user (after tax/VAT/Hosting minus staff) is approx $6 per month, at 500 users I'd be doubling my current income, at a 1000 I'd be able to work on it full time, at 2000 I'd be very comfortable, anymore than that would just be gravy.

It's not even a difficult build (I've built much more complex projects for clients) but the market doesn't exist at all (which means it's either a really good idea (unlikely) or a terrible idea (more likely)).

In either case I'm gonna give it a shot since it's a product I'd have bought if it existed immediately and happily.

Why not start by building a landing page and seeing if you can get at least a few hundred people to sign up? Seems like a much easier way to determine if the market's there than diving in head first.

Honestly? Because if I can't sell it, I'll give it away, its something I badly wish existed for my mum over the last few years so if I can't make a living on it I'll host it for free and fund through my current business as a donation.

True, lots of great ideas look like bad ideas at first (again, pg). But there's things that look bad because they're bizarre (hey, let's rent out bedrooms!), and things that look bad because they're trivial (hey, let's help people find the nearest bar!).

The good bizarre ones could wipe out an existing industry. That's what makes them interesting.

The appearance of a startup is more important than the reality of that startup period. The fact is that most people aren't capable of an objective critical analysis of the options before them, so they just go with whichever thing invokes more of their positive biases. People are simply incapable of differentiating substance from non-substance, or objective value from non-value. It's all about the signals that they pick up from your appearance and presentation (including your representation by third parties, like friends or press outlets).

You can be a massive chunk of crap that only works 10-20% of the time and still get 90% market penetration based solely on the robustness of your marketing/image management teams. The key, therefore, is to look substantive and important, even if you're not, and trip the correct biases in your target audience.

This acts against good founders and good products, because they usually spend a lot of time thinking about how to fix something and improve their product instead of worrying about their PR.

They're "solving" worthless non-problems because the founders aren't doing the hard work of finding real, valuable problems to solve.

Define real valuable problems.

Curing Malaria is a real valuable problem. Sending your friends pictures (Instagram, Snapchat) is inane crap comparatively.

I'm not defending or attacking these by the way, I am just saying that "value" is totally subjective.

The majority of Startups aren't solving anything that the rest of the world would consider "real valuable problems" but that doesn't mean they wont use them.

By "valuable", I mean $$$ kind of value.

I completely agree that value doesn't necessarily translate into money, or may do so in such a diffuse way that it can't be turned into profit. Large, expensive projects that do good for society but can't be done "profitably" are why we should have nice, modern governments.

> Most startups fail, and I think a lot of that is because a lot of startups are really stupid.

Nailed it. Part of the issue is, starting without enough exposure to worthwhile problems. When you hear about success stories of Tinder / Facebook / Snapchat, all you want to make is the next "social" app which is almost in all cases is a lottery but try working with people outside your circle / industry, and you will so tons of unsolved problems with people ready to pay a lot for even a little solution that adds business value.

Hi there ! Honest question, how do I find actual real world problem that people would pay for ? I posted question yesterday where to meet and know about SMB problems ? Do I build something and then go around showing , which is not correct imo or should I interview people first and then build something?

Where to meet people with problems that can be scaled and paid for solutions?

You might want to start by reading _The Lean Startup_, which will teach you a method for discovering and refining valuable problems, and avoid wasting effort on unproductive ideas.

Must say i am iffy about the Ford quote. while it is clearly aimed at being condecending, i find that often the public know what they want but not how to describe it to the engineers. This because they do not know the engineer vocabulary, and thus instead fall back to allegory or similar.

I find it quite overdone, but very important as a principle. I am working for a startup that is moving way to slow. They are doing tons of users testing and making decisions based on the wisdom of the crowd, which is not much.

Again, this quote at least to me, means that you listen to people and give them what they really want. People wanted to get places quickly, at the time, they thought that meant faster horses.

You listen to what your users are saying to you, not doing what they say. People propose solutions before porblems.

The other piece, is that you have a concrete idea and you develop it, and think about it enough, that you then validate by asking people. In a fragmented world like this many people want different things than other people, even from themselves day to day.

So, on second read, if I am hearing you correctly, I think we are all saying the same thing. I don't think many people would suggest google ask a bunch of users how they should implement a distributed network architecture, but ask you every time you log in what you are looking for.

Quick side note - you should get a decent SEO to look at the technical side of your site sometime, but as a free tip, pick www or non-www and canonicalise one to the other. You gots indexation issues.

I don't really understand why they need to grow quickly. You only need to grow quickly if you take millions of dollars and need to distribute a profit to your investors before too long.

Whatever happened to growing slowly, proving the revenue stream before you throw millions of dollars at something? Why is that such a bad thing?

I understand from a VC's point of view why it's a bad thing, and that's the point of view which YC and other incubators are coming from. But why is it a bad thing for the entrepeneurs?

Taking time to consider what you're building is important. Taking time to grow is also important.

They don't need to grow quickly. But this community has set their own definition of 'startup', which includes rapid growth, and this is a member of YC talking about what they tell their startups, so it is falling within that context.

Outside of the 2-letter acronym world (YC/VC/SV/HN), absolutely entrepreneurs should set their own speed -- they only have to answer to themselves.

> But this community has set their own definition of 'startup',

This is reasonable though, right? We already have words for such ventures and venturers when they don't necessarily grow fast: "small business" and "entrepreneur". The reason "startup" has entered the lexicon is because there seems to be something categorically different about a large fraction of tech small businesses these days, and rapid growth seems to be a the correct differentiator.

This community seems to have its own definition for a lot of words, startup included. Nothing wrong with that, so long as one recognizes that. The trouble for me tends to be when the wider press gets involved, and it isn't clear if the "two-letter acronym world" (nice expression!) definition is intended, or the definition used by others is intended.

What does one call a new company formed to develop a pharmaceutical candidate? I would call it a startup, but I don't know that such a company has much at all in common with what people around these parts call a startup.

A startup is a company with a new type of product or business model. That experimental aspect is the salient difference from a vanilla small business.

Having a brand new thing to offer the world often entails rapid growth (and the VC funding model demands it), but the idea that growth is synonymous with startups is peculiar to this here community.

I object to the idea that we should even concede "new type of product or business model".

A startup is a new business.

There was a brief push to qualify "startup" with "scalable", but the only people that seem invested in this (bogus) controversy are on HN, and HN just assumes "startup" means "scalable tech startup".

(Of course this is just semantics, but let's indulge...)

New businesses started up all the time over the past century, and there was never a need for additional terminology. People just said "I joined a new business" or "Bob started a small company". Why, then, did a new word "startup" develop? I'd say it's because people intuitively realized there was something about these businesses, which were largely tech and especially software, that set them apart from the existing idea of a new business. When someone starts a landscaping company today, no one ever calls it "a startup", and there's a reason.

The word "startup" predates the kinds of companies you're talking about. It's been co-opted.

Doesn't matter. The word "drone" meant different things in the past too (and was used much less often, like "startup"). What matters is actual usage, and whether that usage is sensible, i.e. "carves reality at its joints".

There are two different groups of people using the word "startup", and only one of them seems intent on telling the other to stop.

Objection sustained.

Wikipedia also currently has "the essence of startups is generally related to the concepts of ambition, innovation, scalability, and growth." So I don't think it's just YC/VC/SV/HN.

Of course you can have slow growing or non growing profitable businesses too. It's a bit of a terminology thing.

What is the cited source on Wikipedia for that? A link to a pg essay?

I think about this question a lot because and will eventually write something longer-form on it, but here's a short summary that builds on what others have said:

It's not that startups "need to grow quickly" -- when VCs say that, they're coming at it from the wrong perspective, and it's a source of a lot of confusion. It's that a good idea WILL grow quickly if you let it, and what you NEED to do is not get in the way of the natural growth of an explosively successful idea.

The perfect model that (most) VCs are trying to match is Facebook, Instagram, or WhatsApp. A new innovative product that takes the world by storm and relatively quickly monetizes into something worth $billions.

VCs want the next Facebook/Instagram/WhatsApp. That's why they're giving you $millions, and they know 9 out of 10 shots aren't going to hit. Because of the way VCs frame their desire for success, it's often confusing to folks on the frontlines and they hear it as "your mission is growth" vs. "your mission is to build something everyone on earth wants and then give it to them quickly."

Rapid growth is almost more a RESULT of a successful startup than it is a pathway. It's a symptom, not a task.

So what investors mean when they say "you need to grow quickly!" is more like: "We want to see you growing quickly, because that's a sign to us that you've hit the nail on the head."

Facebook/Instagram/WhatsApp didn't have growth hacking teams. They didn't sit around saying "We need to stop working on the product and figure out how to grow! grow! grow!" They nailed it, they saw that everyone on earth wanted their product, and they enabled it to happen by not getting in their own way.

>I don't really understand why they need to grow quickly. You only need to grow quickly if you take millions of dollars and need to distribute a profit to your investors before too long.

You need to grow quickly because the world is a big place and technology concepts unfold more or less sequentially, with each iteration building on the past.

As a result, the world is full of people who are going to have similar ideas to you and similar opportunities to take advantage of them, around the time you do, and whoever gets established first will get to take most, if not almost all, of the pie.

I don't think it's really about that, especially because getting most of the pie doesn't really matter if your business collapses a year later, or users move on because they got bored with the product or service.

I think it's really just the nature of so many startups acting in the consumer space, where you need to have tons of users otherwise your business has no purpose.

If you customers 'get bored' with your product or service, it won't matter whether you go fast or slow: you aren't chasing a real opportunity.

I guess I get that but the marketplace is also a big place and often times there is room for more than one player. Apple and Samsung, Gillette and Schick, Pepsi and Coke. You don't need to totally wipe out competition in order to be a successful company. Furthermore, when there are totalizing effects, it's because the technology benefitted from massive built-in network effects (i.e. Facebook, Uber.) Many (possibly even most) start-ups don't rely on those.

> Apple and Samsung, Gillette and Schick, Pepsi and Coke.

None of which got started during the current technology cycle.

>Many (possibly even most) start-ups don't rely on those.

I disagree. In fact, IMHO, the overwhelming majority of startups benefit from organic methods of growth that come from word-of-mouth and reputation. Any startup with more than one customer, pretty much. Owning the market means owning that, in your respective field of business.

> None of which got started during the current technology cycle.

Then I need to understand why the current technology cycle is fundamentally, essentially different from previous ones. A business needed to bring in more money than it spent 30 years ago just the same as it does today.

> I disagree. In fact, IMHO, the overwhelming majority of startups benefit from organic methods of growth that come from word-of-mouth and reputation. Any startup with more than one customer, pretty much. Owning the market means owning that, in your respective field of business.

Start-ups rely on networking but by network effects I'm referring to a very specific phenemenon whereby the value of the product itself increases with each user / car driver / payee / node, etc. This in turn leads to this kind of run-away virality that the Facebooks and Twitters and Ubers of the world experienced. Plenty of start-ups achieve virality in other ways, merely by selling the product over and over to people, and having people sell it to each other through as you point out word-of-mouth. But in the case of network effects that process is compounded exponentially, making it very difficult to scale (without capital) because often times revenue lags behind user growth.

> Then I need to understand why the current technology cycle is fundamentally, essentially different from previous ones. A business needed to bring in more money than it spent 30 years ago just the same as it does today.

The barriers to entry and growth are radically different. As a result, there is more competition and faster growth at all company scales. So businesses have to scale their velocity accordingly.

>Start-ups rely on networking but by network effects I'm referring to a very specific phenemenon

Which is moot: the competitive effects of brand awareness and economies of scale don't disappear because one chooses not to consider them.

except the world and san francisco/bay area are about the same thing around here. if you're coming from a normal place, your first ride through SF will be an eye opener: instead of ads for shoes, tvs and toothpaste you've got iphones, tech conferences and cloud services. hiking around national parks in the area you can listen in to people chatting about service stacks and investors instead of movies and books. it's a world within a world.

There are obviously lots of businesses that benefit from VC investments, and for lots of reasons.

For example, at some point, hopefully, you're going to have a product that lots people are willing to pay for, but those people don't really know about it. You could reinvest revenues from existing customers in order to fund growth, but that could take many multiples longer than if you got an influx of cash to do that.

There are lots of first-mover advantages as well. Or if you're taking on a big competitor, that faster you grow, the less time they'll have to adapt. There are also a lot of businesses that eventually become profitable that simply cannot be profitable early on.

I personally don't like these types of businesses and would much rather bootstrap a lifestyle company or companies, but that's a personal preference and I know it limits the types of businesses I could start.

If you want to grow slowly, that's fine. It's just a different game.

With technology, there's so much leverage, the winner can achieve massive spoils and that typically requires growing rapidly.

Because the point of starting startups isn't to create a successful business. It's to trick VCs into giving you money to burn! /s

I think this comment still stands without the "/s". The extension is that VCs pawn the company off to the public markets, take their money and are long gone when reality catches up (Twitter, Square, GoPro, Zynga, Groupon...).

definitely... MSFT, Facebook, Google, Apple, Salesforce...

I'll give you Facebook and Salesforce, they're the only ones comparable

MSFT was profitable from almost day 1. I don't think its a good comparison.

Except Microsoft and Apple are both in the business of making things and then selling them to people for more than they cost to make.

If that's what you think, why are you even here? Maybe assume positive intent and you'll have better luck understanding all this.

I don't think assuming "positive intent" is the problem here. Read pg; he's talked about exactly this problem. A lot of would-be founders think that you "win" when you land that investment and get some publicity. It's easy and alluring to think, "Hey, we got a Series A! We're millionaires!" when the truth is nothing of the sort. It's actually, "Crap... that's all the money we have to get us to that next goalpost that soooo far away."

Taking investment is, at best, a necessary evil for startups. It's not a victory. It's never a victory. It's a compromise. It's "I just sold a hundred million dollars for a million dollars" - with the awareness that if you hadn't sold it, you wouldn't get the other hundreds of millions either.

If you are a VC backed startup, growth solves all of your problems (at this particular point in history). If you are a non-VC backed startup, sales/profit solve all of your problems.

Either way, growth is a wonderful thing and I think you should grow as fast as you can.

I find profits to be way more fulfilling than growth/sales. Growth does not increase my happiness. Pressure from third-party investors does not. Money in my pocket, free time, decision freedom does.

>"If you are a VC backed startup, growth solves all of your problems (at this particular point in history)."

What I think you're alluding to here is that growth solves all your problems as long as interest rates are still low so money is pouring into VC without a requirement of seeing revenue, and as long as the company is not in the process of taking the company public since the public markets have wisened up to the "all growth, no profit" game.

Is that accurate?


Right on. Makes sense for YC and Sam to say "you need to grow fast" because they're investors, and understandably want return.

But for an entrepreneur, Sam's rationale is not yours. How much $$ do you need to make a happy living per year? Put that number down. The number of opportunities to achieve that goal is much much higher than unicorn opportunities.

TAKE NO INVESTMENT, and pursue that opportunity. Once you make that number/year, you will just think about working less, rather than making more money. And you'll be happier than most unicorn-founder who need to grow X% more next quarter because of their investors.

Yes: easy to say. But it is is much easier to do than building Facebook or Uber. A bit sad for Sam and YC, but much better for you.

Have been doing that for a while. It's working. No big press articles, no fame, but I am spending weeks skiing with family right now instead of growing faster. And making really good money.

Like many SMB founders, it's a throwaway account, not wanting to give away too much... Just a bit tired of reading of the "growth bible" as being the only way to be an entrepreneur.

Startup as PG defines it in his essays is a company that's experiencing rapid growth - often to support growing rapidly you need a lot of money.

I think growing rapidly can have certain advantages - in Facebook's case it allowed them to take over the social environment for all colleges and then eventually everything else. Slow growth in trying to do this may have not allowed them to win.

Since VCs are betting on the exponential growth wins (AirBNB, Facebook) they're only interested in hypergrowth companies which YC narrowly defines as startups. Going slow doesn't mean you can't make a great company, it's just not a 'startup'.

Of course you need to be 'making things people want' and iterating with users before you can grow, which is what I think sama is talking about.

> Startup as PG defines it in his essays is a company that's experiencing rapid growth - often to support growing rapidly you need a lot of money.

I see this response a lot, but that just begs the original question. Why is a startup with VC money a good model for a business? Why not a bootstrapped, "lifestyle" company? What empirical data do we have that effective long-term wins for investors and employees and the world (and not for VCs, who are simply artifacts of the startup ecosystem) are more effectively or efficiently accomplished by startups and not by slow-growth business—or by nonprofits, or bored hobbyists working on things "just for fun", or anonymous pseudonyms who post research papers and then disappear, or anything else?

Facebook was a slow growth case in the early days: you had to be a member of certain colleges, and eventually all colleges, and only later the general public. Unlike Airbnb or Uber or another business with a good reason to start in new cities slowly, Facebook's closed registration prevented them from getting any of their users' complete social network for little obvious benefit. There were a good handful of systems at the time with open registration and a similar target user base that are nowhere near as successful (like MySpace, LiveJournal, Xanga, etc.). It might even be possible to make the argument that Facebook's slow growth was important to getting the product right, but in any case, it was certainly the nature of the product more than the ability to grow that helped it.

I think you're arguing against a straw man.

Nobody in this community (including pg) thinks there's something wrong with small businesses. They likely patronize many small businesses every week.

The point is that's not what they're doing. Startups are categorically different from small businesses and there's a lot of advice for my tech startup which isn't applicable to my mom's landscaping business (and vice versa).

I agree that startups are categorically different from small businesses, but why do we believe that startups are a better approach for / more suitable for tech businesses than the small business approach?

(I'm deliberately leaving "better" underdefined, but I do want a definition that focuses on it being better for founders, early employees, late employees, users, or the world, more than better for VCs. I know it's better for VCs; that's not interesting.)

Technology and technology-based products and services have the intrinsic ability to scale, especially digital ones. Physical technology can be manufactured and mass distributed; digital products and services can be copied or spun up.

WhatsApp was first released in 2010. Five years later, it had 900 million users with ~50 employees. That kind of expansion was only possible because of the intrinsic scalability of technology: the ability to build an app or a website or a web service once and sell it to "N" users with low marginal effort and unit cost.

There's no way for a home gardening service to scale to 900 million users with ~50 employees, so if you're looking for investments with the potential for very high return, you might be inclined to look toward technology businesses.

Tech businesses have this natural ability to make a product/service once and sell it to vast numbers of users, while traditional businesses do not (restaurants, laundromats, gardening services, consulting, etc.). Some chain businesses like McDonald's and WalMart have developed technology to allow them to scale above competitors, but no restaurant chain can match the growth potential (per unit of effort) of a digital product (viz. WhatsApp).

A final reason for the investor and founder preference for digital tech-based businesses is the low capital requirements and operating costs. A business could not compete with WalMart or McDonald's with a shoestring budget, because of the unavoidable reality of the expense of real estate and physical infrastructure. Digital infrastructure has relatively low costs per unit - consider the cost of a restaurant chain serving 100 million people daily, vs. the cost of a digital appp/website/service at the same scale. The average McDonalds serves about 2000 customers per day, so you'd need 50,000 McDonalds to serve 100m people daily, and the capital to open that many stores would amount to tens of billions of dollars. If you've developed a game like Clash of Clans, then you can sell it to 100m people and earn revenue daily with relatively low unit and operating costs compared to a McDonalds. Clash of Clans is estimated to earn $1.5m in revenue per day, which is comparable to about 220 McDonalds stores; and Clash of Clans has far lower operating expense than 220 McDonalds stores, and has a negligible cost of goods sold.

Digital infrastructure is also far more fungible. If Clash of Clans ceases to be popular, the servers that power the game can be retasked to the next big game far more easily than McDonalds stores can be remodeled into a new restaurant. Cloud hosting providers make it possible to rent a large amount of compute power without investing capital (taking it as an operating expense instead), and at short notice, whereas it'd be difficult to open a large restaurant chain without purchasing real estate up front, requiring a large up front investment in the business; and it would take considerable time to acquire the necessary property and assets (refrigerators, stoves). If the digital product/service proves to be successfu, the business can be scaled up rapidly, while if it is unsuccessful its operating expense can be slashed relatively easily and redirected elsewhere (compare the ease of ceasing to purchase cloud hosting, vs. closing a restaurant and remodeling or selling the property).

All of these attributes together mean that: digital technology businesses have lower capital requirements and are easier to start; they are easier to grow rapidly if they prove to be successful; The businesses and their assets and employees are easier to pivot and repurpose; the ceiling of maximum possible success can be very high (the whole world buys it)

Where did Sam say that startups are "better." He might personally find them more interesting to work on and better suited to his skills, but that doesn't mean they are in general superior.

I wouldn't advise my mom to quit her business to make a "startup."

Do you disagree that Y Combinator encourages people who wish to work on interesting technical problems to do startups?

It is certainly true that nowhere in this post is an explicit statement "You should do a startup." But the post is of the form "Given that you are doing a startup, you might run into such-and-such problem", and I would like to dispute the implicit assumption that you should be doing a startup in the first place.

>I think the right initial metric is “do any users love our product so much they spontaneously tell other people to use it?” Until that’s a “yes”, founders are generally better off focusing on this instead of a growth target.

I have a warning for future, young developers: do NOT join a startup that hasn't found product/market fit, and isn't trying, with all of its might, to find that product/market fit. And when I say, by all of its might, I mean companies that product/market fit isn't the first and foremost goal of the startup.

This is really hard to determine from the outside.

For example, a friend of mine is working in a well-funded, post series A startup that just declared internally its biggest target was a billion dollars of products shipped by the end of the fiscal year. The team is smart, the founders are passionate, and they've been judicious with their funding.

Problem is that the CEO herself admitted that they don't yet have product/market fit. Which is to say their "guiding light" metric is a lagging indicator, i.e. it measures that tail-end of their efforts, and certainly not how much customers love them (and are willing to pay). It's the equivalent of early Facebook using advertising revenue or page views as their growth target instead of monthly active users.

As Sam Altman points out, everything they do will be "hacking" or, as PG puts it, doing tricks that are not sustainable for a real, billion dollar company, in attempts to achieve that misleading number.

At least from what she told me, the most senior engineer said that this isn't bad: the engineers will focus on product/market fit and the sales/marketing teams will focus on growth. Of course, she didn't say that, in a conflict between the fake growth target and product/market fit, the fake growth target always wins. Even now, I'm trying to pull my friend out, before the mental gymnastics set in.

Don't be fooled. A startup needs a real, sometimes small, star in the sky to navigate churning waters. A ship that chases the moon is a ship that will soon be sinking.

This is absolutely true. I work for a YC-backed startup and sometimes felt like that "star in the sky" wasn't there, and worried that my lack of faith was disturbing (to quote movies). Thankfully, the founders here are open enough that I was able to voice my concerns, and it turns out that I had a totally incorrect idea of the product/market fit that they were/are aiming for. As a core member of the dev team, this could have been a potentially catastrophic misunderstanding. So yeah, everyone being on the same page and seeing that same goal is really crucial.

> I think the right initial metric is “do any users love our product so much they spontaneously tell other people to use it?” Until that’s a “yes”, founders are generally better off focusing on this instead of a growth target.

It all comes back to this. I was listening Aaron Harris' interview with Digital Ocean's Mitch Wainer on Startup School Radio the other day (phenomenal podcast). Wainer mentioned how people actually love Digital Ocean, and how rare it is for that word in particular to be paired with the name of a company. I think it's said so often that it can easily be glossed over, but that word is not chosen lightly. Having a product users actually love is extremely difficult, and extremely rewarding.

I've wondered what about Digital Ocean users love so much (I'm a customer)? They are not the first to offer cloud housting at cheap prices (I've used slicehost, EY before they became pricey in the past).

Is it price? Is it ease of setup? Something else?

What appealed to me initially was that they were charging the same as the shady ultra cheap VPS providers but had an actual brand I could trust and were actively developing software to make managing their boxes easier. When I first saw them I was in college, so very price sensitive. They also have had a really effective content strategy from the start [0]. The tutorials are mostly on open source stuff you can do on any infrastructure provider, but the quality is pretty good and reinforces the trustworthiness of the brand.

[0]: https://www.digitalocean.com/community/get-paid-to-write

For me, it's the wealth of tutorials and set-up information, as well as the straight-forward nature of their interface (both at a web and shell level). The pricing model is also fantastic.

I think the right initial metric is “do any users love our product so much they spontaneously tell other people to use it?”

Many larger companies use "Net Promoter Score", as a way of telling this:


Of course, like Agile, there are Consultants, a methodology and books, but there is some truth in asking your customers if they would promote you to another person.

Perhaps I'm the only one that sees this, but this essay appears to be a significant departure from the YC rhetoric I've observed in the past few years. Every time I've heard retention discussed (at YC) in the context of growth, it's cast as something you address when your growth has stalled. Not, as this essay would lead you to believe, a prerequisite of growth itself.

This essay, while it discusses a really important idea, comes across as saying: "Foolish, fashion-focused founders! Clearly retention comes first! We would never imply that you should focus on growth before your product had adequate retention!"

But that's exactly what I've seen YC partners do. Recommend that companies focus on growth, because that's what YC's definition of a startup is: a company that grows very quickly. Having seen both sides of the curtain, the essay leaves me with a greasy, queasy feeling.

Am I missing something here?

Sam, are you changing your standing advice from "focus on growth" to something else? If that's what this post signals, please take ownership of the change and spell it out.

It's poor form to imply that misguided founders (and their devotion to a fad) are driving the growth zealot craze when you've had a hand on the wheel for years.

Our motto is "Make something people want!" If that's not a statement about retention, I'm not sure what would be.

And a startup needs to grow quickly. It's not an either/or--at most, it's a sequencing issue.

Most founders fail because they don't make a product product some users love, or they never make a product at all. Others fail because they don't figure out how to grow fast.

But the best way to grow fast is to have a product that's so good people tell their friends about it.

All I'm saying is that I think things have shifted--perhaps because we've given bad advice, but also probably for lots of other reasons--and founders should course correct.

It's definitely a sequencing issue.

Glad you mention the "make something people want" motto :). I've recently realized that it's actually wrong. Well, backwards.

A more instructive formulation is: "find something people want and make it." Better to acknowledge the value of the search (and confidence in the finding) before encouraging the making.

I have seen many founders struggle to understand when the time to grow is. Pressure to raise deepens the struggle. On the topic of growth vs retention, of the hundreds of founders I have met or read about, I have been most inspired by the actions of Ooshma Garg (of Gobble). The quality of her service is second to none. She's the only founder I've ever met that has completely shut down her product because it wasn't good enough, only to reboot it to better quality, retention, sales, and growth than ever before.

In my experience (flawed and narrow as it is), the best way to course correct is to issue a recall. Treat bad advice like bad goods. Post the criteria where those affected will recognize themselves and make needed reparations.

As a community, we need to better celebrate founders that are doing it right. Not mock those of us still figuring it out.

If YC wants to get the word out about prioritizing product market fit above growth, it should elevate Ooshma's perspective, along with other founders that have made similar gut-wrenching decisions to invest heavily in retention while sacrificing weekly growth targets.

Overall I'm glad you wrote and posted the essay. And I appreciate any humility you may have felt while drafting it.

In general I don't agree. Maybe in a B2B setting you can do this. Do you have an example where this has worked? I can only think of many examples where it wouldn't have worked. Do I want a website where I can make friends with my real friends? No, what for, we'll just hang out in real life? Do I want a service where I can post 140 character messages to the world? Only 140 characters? What's the point of that? Do I want to be able to stay in other people's houses while they are away? No way, that's crazy, nobody would go for that over a hotel... (I happily use all these services now, of course).

There's always going to be a perspective that makes a business sound crazy before it's mainstream. That said, I used many of those products when they were very, very young and I did want what they were building. In many cases, those businesses came after others that lit the way but somehow came up short.

My point isn't that people should be asking for what you're building. My point is that you should think about how you'd know if people want it before you worry about making it.

That said, I regret adding that comment here, as it detracts from my larger point about encouraging YC to use more constructive ways to help their founders course correct.

It seems like there are a million different ways to say the same thing: early on focus on retention.

You can get this from your data at sufficient scale. You can get it from surveys like NPS. You can get it from talking to customers.

One problem I have with the word "growth" is that it implies top level acquisition. A better definition would be hitting your goals. If you want to grow your retained userbase, then getting top level acquisition isn't actually hitting your goal.

Retention and NPS seem orthogonal.

Nope. NPS is a measure of satisfaction. You care about the word of mouth component, but it isn't measuring something different than retention.

Nope. They're literally measuring two different things, return rates vs "would tell a friend" rates.

Whether those two happen to coincide is quite product-dependent. Finding examples where they significantly diverge is trivial.

A recent example of this is Peach.

Did people want to tell their friends? Yes, but not because they loved the product. Because they wanted to be seen as cutting edge.

Most of the product hunt/tech crunch circle jerk is stuff like this. Lots of fluff with short-term spikes but little long term retention because the product doesn't solve a real need.

I'd say they're different, but definitely not orthogonal. NPS asks if you'd recommend the product, retention is how long you stick around after first use. Both metrics embed the user's affinity for the product (which is the underlying thing sama's saying to measure and improve).

This is closely related to Aristotle's concept of potentiality[1] and the active intellect. More succinctly, a startup must build up the potential energy of its product prior to transforming it into the kinetic energy of growth.

[1]: https://en.wikipedia.org/wiki/Potentiality_and_actuality

I like this. I am also reminded of something I read about agriculture and food forests. The idea was that all land "wants" to return to its forest state, and that traditional agriculture requires constant energy inputs to hold the land in a raw, tilled state that can be quickly planted. The alternative is planting multiple compatible canopies—an edible forest—so it can stably sustain itself without inputs.

I think something similar happens socially... Whenever you diverge from the standard model of doing things (as startups do) it requires a massive amount of energy to stay in that raw state. Markets, employees, everything "wants" to pull you back to the typical way of doing things. The forest of the status quo fights like hell to grow back.

So ideally what you want to do is clear a little land and ASAP plant a complete ecology that can fill that little space. I think this is where I get to the potential energy.... If you are going to have to wait a long time to plant all the different parts of the ecology, you need more potential energy to hold the land open. If you are going to plant a balanced ecology right away, you can get away with less energy.

Anyway. Mixed metaphors but I appreciate the opportunity to mix in your idea to things I've been thinking about.

> A startup that prematurely targets a growth goal often ends up making a nebulous product that some users sort of like and papering over this with ‘growth hacking’. That sort of works—at least, it will fool investors for awhile until they start digging into retention numbers—but eventually the music stops.

But by then, the startup has already won. They already got the funding thanks to their sleazy marketing, and the social validation/connections that comes with it.

The essay doesn't contradict the "it is easier to ask for forgiveness than permission" idiom, unfortunately.

>But by then, the startup has already won. They already got the funding thanks to their sleazy marketing, and the social validation/connections that comes with it.

Only if getting seed money and failing in 18 months is their idea of winning. I have a hard time seeing how this is an ideal outcome in either the short or long term for anyone.

> But by then, the startup has already won.

Has it, though? You will run out funding eventually, and if your revenue isn't cutting it, your users are leaving, the few employes you have quit, etc... Getting funded doesn't solve everything magically.

They've maybe won the game they thought they were playing - the one where you get an investment and maybe mentioned on TechCrunch and stuff. They lose the game they're actually playing, which is to build a sustainable business.

This is a great post. This is the by far most common miss-understanding I get from new startups. People are far too confident their product have reached product-market-fit before that is actually true - often making this claim with very little data.

I've seen many VC's who are not data-literate enough to be able to understand the difference between measuring product value and growth without substance.

Isn't Y-Combinator's whole thing "Make something people want."? That's a rather succinct and meaningful statement on its own. A very powerful one that should be held above perhaps all others.

I'm sure that whatever your business practices are, after boiling everything away, you should be doing this above all things. An article like this suggests perhaps some sizable amount, or many organizations are losing sight of the way.

The alternative to this is focus on getting your customers to pay you in advance for your product. Not only will this tell you if your customers really love your product, but it will provide you with all the cash you need to grow as fast as you want.

The downside is very few products are that loved that customers will pay for 12 months in advance.

Sam is talking about the importance of watching leading indicators when there is a good bit of time between your present actions and future performance.

If you are relying on your users to drive growth, an easy-to-measure proxy of "do any users love our product so much they spontaneously tell other people to use it?" is Net Promoter Score[1], or the difference between the % of users who would recommend you, and those who would not.

There are other (perhaps non SV startup) businesses that dont rely on users to drive growth. For example, if you rely heavily on distribution arrangements or retail placement. In that case you might want to look at other metrics that are leading indicators of growth.


If you are bothering the user with a modal, popup, post-install screen, etc., then why not ask your customers TO promote the product providing Email and Social sharing options? Then you can measure how many people ACTUALLY promoted your product, instead of how many said they would (but mostly probably didn't).

Declared action is a notoriously unreliable signal.

As Warren Buffet has said, the best strategies are “simple but hard.” This is YC advice at it’s best.

(I'm paraphrasing Buffet)

Great post where great product -> growth, but I think it's not stating the real problem. We need to ask why founders are starting to have this mentality. From my perspective this is because YC tends to give too many generic advices and some sticks more than others. The big one that caused this is PG's startup = growth post, and yes he is right and it specifically states you really need to concentrate on growth.

If you have to take away one thing, then it will be growth but this generalization is very dangerous because as Sam's post states you need a product before growth. The better advice is to not concentrate on one thing but to bluntly say you need to have everything and everything will all come together. You need great team, great product, and great growth. If one is weak, you need to fix it.

I'm curious what happened to HomeJoy? They talked about having good cohort analysis numbers(retention) and then they blew up?

They ran out of money, basically.

This growth belief reminds me of the trap of focusing on 'what' instead of 'why' as in https://www.ted.com/talks/simon_sinek_how_great_leaders_insp...

It seems stupid to do something just because you think you're expected to do it rather than think for yourself and do what you want to do and think would be a good idea.

NAIL ENGAGEMENT FIRST. THEN GROWTH. Otherwise you are just bruning through users. Your best hope is to attract them back later via transactional notifications.

Yep. Growth almost never happens by continually attracting more users than you lose. It happens by not losing your current users and continually adding more.

This is interesting in how un-interesting it is.

Find product-market fit before taking investment. Because the only reason to take VC investment (for the etype of startups we are talking about) is as a steroid injection towards growth.

Why is this the only time it makes sense? Because that's the only way you and your investor's incentives are aligned.

Now, whether you need to/should grow fast to survive in the long term is another question.


It's fine to raise a seed round (< 1M) to experiment with some ideas, but raising money after that without product market fit is a recipe for failure.

If you don't have P/M fit and you raise a bigger round, you're going to hire people and create an organization around something that doesn't work. With those extra people and processes, things are going to slow down to the point where finding that P/M fit becomes much harder. Moreover, you are now accountable for expectations and a reasonable return for your investors, which is not a bad thing, but it is a distraction when it comes to P/M fit.

For B2B businesses, raising a Series A means being able to hire your first sales/marketing people.

For consumer businesses, raising a Series A means being able to pay for the growth that's happening without worrying about revenue right away.

Two months ago: https://news.ycombinator.com/item?id=10495402 (Altman's previous blog article)

How much or little has changed?

So his advice is to build a great product?

Or "If you don't build a great product, nothing else you do matters".

I'm running into the growth problem with my website: http://muxme.com. Lots of people say the love the website, but it seems like no one is telling their friends. I was thinking of maybe adding user created sweepstakes, because I think the only way people would spontaneously tell their friend is if they win something from the site. More winners = more people telling their friends, but I can only go so far on my budget. What do you guys think?

I don't think anyone I know would be interested in this website. So even if one person likes, it doesn't mean any of their friends will. As a web savvy person, I assume services like this are not worth trying because

1) it's just as likely to be a scam as not and

2) even if it's worth it right now, it will soon not because because of people gaming/spamming the system. The odds of winning an item for free are not displayed, and therefore are assumed to be astronomically high. It's assumed it's not a good return on my time to register and enter.

That said, it's a cool site, and I like your intentions. I do not think there's going to be a feasible way to prevent gaming and spamming unfortunately.

1) I post a receipt of the item, have a video drawing of the sweepstake, and post the tracking number of the package after it's sent out. Short of revealing the winners name, I don't know what other proof I can provide (although googling the user's usernames reveals quite a bit about them, I encourage them to change their usernames). I understand what you mean, on first glance I can see how people think the site is a scam.

2) I show how many people enter, so the odds are 1/(total number of entries). I was thinking of requiring a valid/unique phone number on registration and require the users to input the code I text them. Think of Craigslist / Chatroulette / etc.

I don't envision any people from Hacker News using the site on a daily basis. In fact, I posted it here as a "Show HN" and a whooping <10 people visited. There are people where an extra $25 means getting to take your daughter out to dinner when she announces she's engaged or an extra $100 means you get to pay your bills on time for the month (real winner stories so far). What really pushes me to work on the site every day is that even if the site fails to grow, at the end of the day, I've helped people without asking for anything in return.

I appreciate the comments.

So are you paying for these prizes out of pocket? What's your goal?

Yeah, the prizes are paid for out of pocket by me. It's not the most expensive project in the world, let's say I average $50/day ($1,500/month) in gift card costs and $10/month on a server. I live in a 42 square foot room and share an apartment with 5 people, so I'd say that the expenses are taken care of from my choice of a frugal lifestyle (we're talking Manhattan, New York rent here). Those are my only expenses to run and maintain the site. The risk of $1,500/month for a couple of months to possibly achieve my dream is a no-brainer.

My goal is growth. Craigslist is a great example of a site that I want to be like. Offer an awesome service for free. I'd like to have companies fund the giveaways or maybe even have companies pay to have their giveaways featured as a top result. I don't really know yet. I'm sure there's some win-win situation that I haven't thought of where the users get awesome free stuff and there's a way for me to work on the website full time.

Thanks for the responses. Have you considered reaching out to businesses to offer free stuff? Sort of like Groupon that doesn't cost them anything other than the discount/free item they provide?

Yeah I've definitely considered it. It's on my list of things to do, but I'm also sort of experimenting with the website and trying to optimize the user experience and make the site fun.

At first glance, I thought that your website was listing random free items on the internet and that you'd make money with affiliate links. I looked at a few items and they were all given by "MuxMe" so I was really confused as to WHY you were giving free items. I then looked at the about page trying to understand a bit more and there it is, you seem like a genuine nice person giving free stuff away, hoping one day to become a non-profit.

I'm still a bit confused. In a way, it seems "Too good to be true". I would craft a nice paragraph and put it at the top of the page so that first-time visitor understands what you're trying to do. You cannot blame people for thinking that this is scam (Hey I'm receiving emails every day telling me I won $1M).

Also, you realize that if I tell my friends about your website I have less chance to win.. why would I do that?

What's your long-term goal with the non-profit?

That's very valuable feedback. Thank you for telling me that you thought I was listing random free items and made money with affiliate links. I like the idea of making an extremely visible "about" at the top of the page so that a first time visitor reads what the site is about before making an assumption.

You're absolutely right that the more people you tell, the less of a chance you have to win. The model is flawed right now. The winners are telling their friends (or even friends are asking the winners how they are taking them out to dinner), so if I could somehow get more winners with a low budget, then the site would be a success.

My goal is to work on the website full time. It makes me very happy running the site and I'd say it's the only thing I think or care about. Also, I thought that maybe some of the users are interested in learning programming and I could create Youtube videos to teach them how to make websites/program. I've always wanted to be a teacher.

Neat concept you have going! You come off as a genuinely nice person in your videos, which is great for building trust. My one small piece of feedback would be to consider changing the 'About' link to 'How it Works' instead. I went to your site and immediately started scanning for those words, because usually a sweepstakes site has a catch of some sort.

Good luck!

Changing it to "How It Works" was a really good idea. Thanks.

Thanks for the feedback, I'll take it into account and possibly change that.

when you talk to your users, ask them if they've shown this to their friends. If they haven't ask them why they haven't. If they have ask them why their friends didn't like it. I would also recommend sending out a NPS survey to all your users

Its amazing you've doing this. Definitely keep it up. My advice would be to find a business guy who could do a solid pitch of this for companies. I'm sure there are some out there who would be interested.

Thanks, I've had my ups and downs the past month but contacting the winners and hearing their responses is what keeps me going. I appreciate the comment.

I agree that if I had a business guy that could do a solid pitch that I could get the donations from companies. The website is also very experimental right now, so I think I'm going to keep it up and figure out a good model before introducing people and companies into the mix. The great thing is that the website is currently a hobby project, so I can do whatever I want to it without being forced to do something I disagree with.

Why would I decrease my odds of winning by telling my friends?

The more people that enter the giveaways, the less of a chance you have to win. There's a guaranteed winner every time and your odds are 1/(total number of entries).

how would your website ever make money? Why would someone post a sweepstakes?

I'm not worried about money, I'm worried about growth. There are services like RaffleCopter / Gleam / KingSumo / Woobox / etc. which let you create Sweepstakes to promote viral growth for your website or business.

This definitely rings true to me. The company I work for has been around for ~6 years and while we have made money and grown in that time, it wasn't until the last year really that we hit on exactly what our clients love about our product and continued to nail that. It can take way, way longer than some entrepreneurial people are comfortable waiting for. But now that we've found it, and are continuing to expand our product around what we've found, growth is starting to come much more easily.

'I think the right initial metric is “do any users love our product so much they spontaneously tell other people to use it?'

Question - is this a valid metric when talking about B2B companies? I would imagine there are plenty of B2B solutions which became huge, but which don't exactly have companies rushing out to tell everyone to use them.

I tend to think sama's question is only relevant for B2C, but I'd love to be proven wrong.

I would imagine it also applied for B2B. Think about apps like Coda, or services like GitHub. I remember using a task tracking system that I completely swore by and told everyone I knew about. All it did was track tasks and is very much B2B, but my obsession with its UI made me more productive and those became a talking point, where I'd always highly recommend it. Slack is another great example of a B2B co that passes that test.

I've been explaining this perspective to most clients for years, almost word for word. Specifically the question/focus: "do any users love our product so much they spontaneously tell other people to use it?"

I'm excited to now have an "official" source I can easily cite for it!

I think there are some businesses where this might be less true, but it's pretty rare.

Didn't Alex Schultz, VP of Growth at Facebook, already say this...


Sometimes it's best to grow more slowly at first, build your brand, your culture, figure out your process, don't burn through mountains of cash (if you're lucky enough to have that problem) and wait for the market to come around.

So, basically, get the seeds right, then plant, water, and tend. Don't get crappy seed ideas and try to overcome their deficiencies with some kind of fertilizer. (fertilizer AKA "manure"/bs.)

> The other thing that these companies have, and that also usually gets figured out early, is some sort of a monopoly.

I wonder how this would apply to, say, Uber and AirBnB. What kind of monopoly do they have?

Also, good unit economics or a plausible path to get to good unit economics for your product (i.e., technology pipeline will dramatically reduce some key cost for your company).

small typo: "and that also usually gets figure out early" => figured

Great essay, Sam! I want to expand slightly from a founder's perspective:

You always have two constituencies to think about: (A) your users / customers, and (B) your not-yet-users/customers.

"Before growth", you should really pay very very little attention to constituency B. If you're doing anything interesting, there will always be a torrent of skeptics, doubters, and haters. (That sucks, but it's very human, and you just have to ignore it.)

Focus entirely on constituency A. There are three "R" metrics to look for: (1) Retention. (2) Revenue. (3) Referrals. (Note: these are the three "RRR" of the classic "AARRR" model.)

Initially, constituency A might just be your friends or extended network who you can convince give what you've built a try. You may be able to get them to try your new product once or twice, but that's about it.

However, if they keep using it again and again voluntarily, that's a really good sign (Retention). If they voluntarily and happily pay for it, that's great (Revenue). And if they voluntarily tell other people they know about it (Referrals), that's amazing!

Until you have those nailed -- or at least retention and referrals if you're deferring monetization -- don't worry at all about the "B" group (i.e. don't worry about acquiring new leads at the top of your funnel), because you don't have product-market fit yet.

Once you do have a product that the constituency A people love, only then can you start thinking about constituency B and how to turn them into new happy users / customers. (While continuing to make your constituency A happy!)

I think Sam is going a step further to place extra emphasis specifically on Referrals out of all three of these metrics. This is a subtle but powerful insight. Whether your product is B2C or B2B, most purchasing decisions have a huge emotional component. If your product is so good that your users want to tell other people about it, that's a huge step above being merely satisfactory, and goes a huge way toward powering your growth.

I don't care if you're buying a burrito, a car, CDN bandwidth, or an analytics SaaS. You as a buyer want a reasonable degree of certainty that you're going to feel good about that purchase after you've made it. Hearing that referral from your peer, whether a friend who tried that food truck before, or another engineer who used that SaaS provider before, goes a huge way toward giving you that pre-purchase confidence.

If your users aren't excited enough about it to be talking about it with their peers, you may have to adjust your product or your segmentation of the market until you get there.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact