Hacker News new | past | comments | ask | show | jobs | submit login
Startups in 13 Sentences (2009) (paulgraham.com)
306 points by gasull on Aug 15, 2016 | hide | past | web | favorite | 78 comments

I like all of these except number 10: avoid distractions.

I can't support the claim myself, but I have seen a startup succeed as a professor's side project with his brother (lasted 10+ years, IPO).

Going all-in on your startup from day 1 might seem efficient, but it may not be most effective.

For instance, if I have an idea for an EduTech startup, I might e-mail some players/professors in the field and chat with them about the idea, and try to see why most startups in the field fail.

You have to wait until they respond, because your entire assumptions might fail! Keeping the youthful spirit is good, but talking to people who dedicate their lives to these fields is often invaluable. Don't build on top of unsteady assumptions.

"Launching fast" might not be the only way to test your assumptions.

I don't think he's referring to making the startup your full focus vs side project at the early stage; after all PG et al still advocate burning slow. Sometimes a side project doesn't benefit by having you work on it 4–8x as many hours per week if you have the right prioritization in place in the first place.

I think this is more concerned with things like media coverage, founder interviews, attending/panelling conferences, and things like that.

I like how Sam Altman described it in Focus & Intensity from Startup Playbook [0]:

> Don’t get caught up in early success—you didn’t get off to a promising start by going to lots of networking events and speaking on lots of panels. Startup founders who start to have initial success have a choice of two paths: either they keep doing what they’re doing, or they start spending a lot of time thinking about their “personal brand” and enjoying the status of being a founder.

> It’s hard to turn down the conferences and the press profiles—they feel good, and it’s especially hard to watch other founders in your space get the attention. But this won’t last long. Eventually the press figures out who is actually winning, and if your company is a real success, you’ll have more attention than you’ll ever want. The extreme cases—early-stage founders with their own publicists—that one would think only exist in TV shows actually exist in real life, and they almost always fail.

[0]: http://playbook.samaltman.com

I think you've missed a key distinction: the time before you accept investor money is very different from the time after. Prior to that point, sure, pursue whatever leads seem interesting at whatever level of effort they seem to merit. After that point, though, you need to go all-in.

I went to a class on startups once where the guy there said you should bounce your idea off like "100" people, (basically, to pivot early). Before you go "great guns" on your idea. FWIW, not sure how that ties into "pre investor" vs. post however :)

Yes but there is only so much energy in the brain. A good startup requires extraordinary thought and work. If you are working on other things that are also demanding, i.e. a a job, then you won't have your best energy left-over.

I agree with PG on this one, though it is definitely a luxury to avoid distractions.

If you are listening to your customers, trying to get the product "right", sometimes its hard to know what the distractions are, and what the great idea's are. As CTO I know I find myself constantly try to keep out CEO on track to delivering one product, when he wants to try 3 other variations.

Heh, that sounds backwards (between CEO/CTO).

The early stage CEOs I know, especially non-technical ones, are always pitching multiple variations of their product (and sometimes multiple products) to see what sticks. I think the CTO brings the technical realism of "here's what we can build this week".

>Yes but there is only so much energy in the brain. A good startup requires extraordinary thought and work.

Or that's what the old wives tales say.

which part?

That a "good startup requires extraordinary thought and work".

I know several people who had started and sold companies (admittedly to the tune of a few million dollars, not billions), without devoting "extraordinary thought and work" to them.

For some it was just a side project where university was their main work.

This is one of my favorite posts from Paul Graham. I would add one more point to it based on my experience.

- Make something that has value from day one. Many tech based startups have significant value for users only when many other users are also using it. Tech based political startups, for example, suffer from this problem. These kind of startups are harder to build because getting and retaining the seed users can drain you easily. It is easier if you build something that adds value to the life of the very first user, even if he is the only one using it.

It's a really great list.

I've worked closely with M&A for Fortune 50 in a previous job:

Here are some dark, somewhat cynical rules:

14) Get the CEO of a big company to Angel in your company, hype you up, and use cash from his company to buy you for a ridiculous amount. You just need to build something that works, you don't need many customers. The conflict of interest for the CEO is positively crazy, and yet, somehow how they get away with spending company $ to make themselves a massive return. This 'dark skill' involves leveraging the relative powerlessness of public investors in a big company with a powerful CEO. Like Salesforce. Or Oracle.

15) Make your product 'cool to nerds'. Like HN readers. Techies are as trendy as any other social group, whether they have the self awareness to realize it or not. Sometimes I think the Valley is a bag of religious cults :) (Kidding.. almost). Being 'nerd hip' can differentiate you from the dozen or other equally good products that are made in far off countries in Poland that will never be able to generate any real buzz.

16) Work with influencer circles. Some people have 10000x more influence than others. Have you ever seen the first rev of Twitter? It's positively laughable. It's your teenage brothers first web-page bad. Twitter would have lasted 10 days if the founders didn't have access to 'influencer networks' and eventually the press, and since it was a press tool ... Twitter is one of those companies, were it not to have started in SV, LA, or NYC - it would never been.

17) The 'force it to happen' plan. Take any established online business model + $500 M in investment, and just make an unprofitable business based on a couple 'interesting key ingredients' or 'tweaks'. Operate reasonably effectively, distribute a lot of koolaid, and get your well connected investors to sell for a small multiple. Jet -> Wallmart.

I could go on :)

Except you make those ideas sound easy to execute. How am I supposed to get a CEO's attention and trust to invest in my startup with just email?

I'm not saying it's at all easy. I'm saying those are paths for those who have the power to pursue them.

There are very few successful startups that came out of nowhere through grind and effort - at least not the fast impact kind.

Being in the Valley and networking hard is one way to get yourself some 'zen' if you don't have Ivy League 'cred' and an influential network. A lot of those folks spent years at Google et. al. which gave them 'cred' and network.

A lot of kids who get into Ivy League have been working pretty hard for a long time. Some of them are learning to be 'players' by the time the are 13, which is amazing and sad at the same time. So - they've been working at it for quite some time.

For most of us, we have to play the regular long game.

Agreed, which is why getting into YC or something similar is very important.. that's how you'd get intro'ed to those CEOs for a round

Two more I'd add:

- Actually doing something is a necessary condition. An obvious statement but I've heard people bemoaning not having a startup success who didn't try. (Also, those regretting not having a bestseller book, didn't try writing a book, etc.)

- Failure is very likely (regardless of Paul Graham's inspiring and encouraging essays). This doesn't mean you shouldn't try. Many of the factors that determine success are things that you don't control, can't predict, can't understand no matter how smart you are, and can't even explain after you're become rich or broke. There's a lot of randomness in outcome.

Whenever I design and implement software, I expect it to fail. I wonder, "where's the catch?" I almost obsess over finding it. This design will fail. Why?

And like any good software, most of them eventually do. But with that attitude, I had obsessed over understanding the limitations and problems rather than successes, which made the next version much more clear to design.

Failure is part of the process!

"If I had ever learnt, I should have been a great proficient."

- Lady Catherine de Bourgh https://en.wikipedia.org/wiki/Lady_Catherine_de_Bourgh

Short version (the tips sans descriptions):

    1. Pick good cofounders.
    2. Launch fast.
    3. Let your idea evolve.
    4. Understand your users.
    5. Better to make a few users love you than a lot ambivalent.
    6. Offer surprisingly good customer service.
    7. You make what you measure.
    8. Spend little.
    9. Get ramen profitable.
    10. Avoid distractions.
    11. Don't get demoralized.
    12. Don't give up.
    13. Deals fall through.

I've been re-reading "The Inmates are Running the Asylum" and am no longer convinced that "2. Launch fast" is true.

Launching and iterating a product quickly can easily lead to false negatives, where an early unpolished product that does not quite meet the needs of the end user might turn off users and harm the brand for good, but by involving user research before launch, product fit for users can be evaluated and iterated on without annoying potential future customers.

I agree with this. The obsession with building something fast means that the market is flooded with shallow products that are little more than a website plus a months worth of coding. Hey a new food delivery startup! Hey a new messaging startup! Hey a new job website! No wonder they are all the same.

I've been building something for 18 months now and whilst I could be proven completely wrong, I feel like I'm working on something people will want and I feel like getting it built right is the correct thing to do.

The thing I am building needs alot of bits and pieces to be in place and working in order to realise the original concept. The MVP and "release fast" line of thinking says "too long, get it out there", but this would result only in releasing software that doesn't do what I envisioned it doing - what would be the point of that?

I really think the opportunity going forward is to build products that are hard to build and require lots of work to implement the actual vision, but ultimately to come out with a product with deep value.

I like the idea of building something deep and technical and functional.

We shall know soon enough if I am right or wrong as I plan to release soon....

I totally get where you are coming from. I really do. Some software is just complex, and it feels impossible to ship anything without all those parts in place.

Regardless. You must find a way to ship early and often. You will go under if you do not.

It helps to make a mental shift, and understand that the problem you are trying to solve is not related to features. You're trying to prove whether enough people want what you are building, and if so, what exactly it is they want (you will get it wrong the first time, and probably the second time too... this is why iterating fast with as many customers as possible is key).

Will the product be sort of crappy and missing all sorts of "essentials"? You bet. But, crappy code that does what people want and they will pay money for always, always beats an untested fully-engineered solution.

>I like the idea of building something deep and technical and functional. Here here!

However its easy to get lost in such products. That why there is still a lot to be said for knowing your MVP. I fight this battle everyday.

> I feel like I'm working on something people will want

Read Four Steps to the Epiphany and apply the process described therein. You'll be able to turn your feeling into an actual conviction.

You can start with a shallow product and slowly build it up with user feedback to create a defensible and deep product. The key is to continuously improve it.

No, you often can't. If you could, it would be already done by one of those shallow initiatives the GP is talking about.

I'd be very interested on any ideas about how to make deep innovation less risky. For shallow ones is "launch fast", but what to do when this one is not available?

>No, you often can't. If you could, it would be already done by one of those shallow initiatives the GP is talking about.

Well, a lot of them have. Besides "you often can't" is not very helpful.

The question is whether it's more difficult because you started as shallow or not -- not whether it's not a 100% surefire way to make it.

Maybe you can start with "just one feature" and then work forward... [?]

It looks like you haven't tricked anyone into asking you what your project is yet. Sorry man, but it was a great attempt to get your project some exposure. :(

Don't worry, I'll be entirely unafraid to overtly ask people to check out my product when its ready to go. I'm not ready that's all.

I think it also depends on the type of startup you have, and how confident and knowledgeable you are of your target market.

I'm currently reaching the end of the ideation phase, and although I'm familiar enough with my target users, I plan to spend a few months in Asia (where they are) to test my assumptions - hidden and conscious - and see if I'm missing out anything else. No matter how careful and smart your assessments are, ultimately it's all in your head! So yes you're right, we need to be wary of false positives and make sure that we really understand our users. But launching fast is a pretty valid idea too; personally it would be a wasted opportunity if I go on the trip without anything to show! A simple prototype can be more powerful and a much better conversation starter, plus the further advantage of lighting the fire under your bum. That would close the catch 22 loop, I hope.

I agree in general -- one should have utility for users to launch. But I suspect PG is seeing (many) more startups that do not release when they could ("we will do one more iteration" to mean "we do not need to deal with users, ops, etc.") than those that release too early.

I agree with this. Lean Startup got really popular in the last years and it is a great tool but it only works if you are in the ballpark. Peter Thiel said that it is a way to reach a local maximum.

If you are not in the area, you won't get enough feedback to iterate, you'll need to pivot and play again. It's like a game of bocce ball in that sense


I think this is a circumstantial point.

I suggest that if you have a much more intimate understanding of your customers - AND - enough self-awareness to be objective in determining what they need vs. what is nice-to-have ... then you can go further.

Maybe the rule should be: launch as soon as you have something that will provide value to someone, i.e. MVP in the technical sense of the term.

I think he is missing one of the most important one, 'Don't neglect your soul and health'.

But this list deals with mostly obvious things about how your startup may generate wealth. It has nothing to do with the people involved. Otherwise, the list in this guy's redundant writing style might be

1. Don't die. Many people fail to achieve their dreams because they die too early. Try not to die as it is hard to keep on living in that case.

2. Be happy. One of the most often stated reasons for being unhappy is sadness. Try to be happy and you will achieve happiness with ease.

I know you wrote this with tongue firmly planted in cheek, but I agree this list has nothing to do with the people involved. It seems possible that if pg wrote a piece on maximizing your odds of being a healthy adult both of those could be included, probably along with not doing a startup.

Trouble is, like with all advice, your mileage may vary. Some people sacrifice themselves and the people around them completely for years in the hope of startup success and then bask in glory and wealth, others do the same thing and end up broken and alone at the end. Some people take the safe path, have loving friends and family and enough to live on, others take the safe path and end up unfulfilled, not producing a tremendous amount of value and next in line for layoffs.

It's probably easier to make a list on what startups need than what people need at a fundamental level.

Many people don't even get to the point where they ask themselves "What makes me happy?" As you said, they see other people (or read lists), think they're happy, want to have the same, work for it, and end up at the same place, but completely unfulfilled.

Being honest to yourself, admitting you are different, admitting you have desires noone else shares, accepting that people may look down on you when you pursue your dreams of happiness, is the hardest thing to do.

I don't talk about running away from anything, I talk about confronting everything that makes you who you are. Inspecting and taking apart everything you ever felt. Bringing it out into the open (not posting on Twitter but bringing it out in front of yourself) and accepting whatever it is that you are. And for that matter, accepting that other people might have it even harder than you and -- I don't get why you wouldn't -- being nice to strangers since their lives might have just fallen apart. The strongest indication that someone is an emotional teenager is when you see that they treat people, who they're not dependent on, badly.

I regularly see people who can't even see when they're figuratively slitting their own wrists. People who project their own unhappiness onto others. People who hurt others without a second thought, who are so incredibly selfish that they can't see that it's like a brick wall standing in between them and their own happiness.

Especially in the tech community and with scientists it is so common that they suppress their own emotions by rationalizing them away. So here's a list of one item that is probably true advice for happiness (except for psychopaths)

1. Make sure you feel and understand your own emotions. Rationality is not the opposite of emotionality (that would be irrationality). So use your emotions as a premise for your own existence and act rationally starting from that premise.

I have to second this. I am a huge skeptic of the 80-120 hour work week. In my experience it results in a lot of sound and fury but the results are of lower quality with little creativity or insight. It also leads to emotional problems and burnout.

I try to stick to 40-50 hours unless something is on fire.

My first two startups failed simply because we were not able to "launch fast".

It's so very important.

Would love to hear more details about what happened.

Startup #1 We spent two years continuously improving a product that wasn't validated yet, as opposed to getting something out the door and pivoting our way into PMF. Our MVP was too planned and too precise. The codebase was quality (Angular/Rails) and the design was tight, but we didn't have something people wanted. Our biggest mistake was trying to do product/customer development while we were engineering in parallel. We failed because we over-engineered a simple idea without truly knowing our customers or the industry.

Startup #2 We started with product and customer development to verify the need. It was there. We had no funding so I (the only BE engineer) had to find a job to support my family while trying to finish the product. That was the mistake. Startups need 100% focus. I ran out of time, energy, and steam while the rest of the team waited on development. I ended up having to leave due to the new time commitment, and they weren't able to find a capable replacement or assistance to hit the finish line.

Startup #3 We verified the need by launching a mediocre product (using rapid development and open source tools instead of rolling our own) and then quickly improving/iterating based on customer response. Three months later our revenue is ~100k/month.

Did you get beat by others to the finish line?

This paper and the "Do things that don't scale" one are regular "must read" from Paul Graham.

To make a startup successful, this is necessary but definitely not sufficient.

From my experience, location is also key; may be to t attract early adopters or customers but to meet interesting and engage with valuable investors. Who you know is as important as what you know.

First and third are very hard. I always have the idea of creating a startup myself, but I lack of network to find a good co-founder, specifically building the application with me (I can't do front-end and mobile). A lot of amazing ideas came out with some nice demo. Being really picky about frontend, I am not thrilled to present a shitty-looking application without some sane user-interface. I get that idea is more important because once I have a solid idea and funding I can hire professionals to do the job. I also get that a shitty-looking application with functional feature is better than a nice webpage with nothing usable.

That leads to the third, which is to iterate fast. I spend a lot of time thinking how to evolve my idea because I have to convince VC for "future", not the present, but often my idea just die with "someone is already doing it", or "it can't really sustain in the long run or can't really work because of X reason." I always seem to have some really solid examples to defeat my own idea.


1. It's very difficult to make a sane user interface without letting customers try whatever you come up with. Tweaks are always needed. If your app is like most, a small portion is the most used, and you simply focus your energy improving that section. Also, your first customers are early adopters and are very capable of using bad interfaces. You get this, but you aren't doing it! This is why he wrote this, heed the advice :)

Also, don't ever stop a startup because someone else is doing it. Name an industry with only one company... With only 10!! If anything it validates the market. At my first startup we consciously decided not to build a new product that was going to face many other entrants, despite the fact that we were experts in that domain. Guess what - the next year, 3 of the top 10 best sellers were in that category! Expensive lesson learned.

And in case you think "well why should I follow pg's advice," then go read a copy of Rules for Revolutionaries by Guy Kawasaki. Same basic ideas, only about 20 years before!

Good luck.

Try something different. Instead of starting with an idea, building some of it, and than trying to iterate. Why not just go straight to "step 4".

Learn some marketing, learn how to network, go out and talk to people. Instead of starting with an idea, start with a customer. Go to the customer, figure out what they need, than go to step 1.

#13 should probably say 'Most deals fall through.'

surviving the sales roller coaster requires a particular (and peculiar) mindset.

I've been trying to keep this "launch fast" advice in mind as I work on a product (pretty much the first serious personally-created app I'm putting out there.)

Another piece I looked up to re-read is this from the founder of WordPress: "Usage is like oxygen for ideas. You can never fully anticipate how an audience is going to react to something you’ve created until it’s out there." https://ma.tt/2010/11/one-point-oh/

They forgot the most important one. "Make sure use the word disruptive". Every single start-up has to have that word.

One thing that could be added to the point about getting demoralized:

Success is "just" 1000 iterations away.

> 8. Spend little.

I think, no one gets this right. What's missing here: for software startups, hardware (servers) should be the main cost of running the business. Increasing load due to the growing user base is a very nice to have problem.

Salaries (for engineers) is and should be the main cost of a startup.

Yes, for software startups that do not have any employees, including engineers, their biggest cost should be hardware.

> The worst type are those that pay money: [...] profitable side-projects

This is the worst for me and it's not really about the money. My startup generates way more money. It's just that I prefer working on my side-projects.

I would add, make something YOU would actually use. Don't make something with the intent only on getting rich. That's how you're successful.

But, for god's sake, don't make the nth "skateboarding startup", "surfing startup", "t-shirt startup", "music startup", "Startup for enthusiast photographers", etc -- in other words, things that young developers, in the microcosm, believe everybody cares about, when not even a large percentage of the younger demographics gives a damn about such things...

I wasn't aware of the huge skateboarding/surfing startup culture!

There is -- just not on the unicorn level.

But startups are not only those designed to be unicorns and be sold for $10 billion -- "lifestyle businesses" are also startups.

If it makes money, then it makes money. Let the market decide if it needs another one.

I don't believe that money is the ultimate justification for everything -- including running a business.

Actually that is the definition of a business.

- the purchase and sale of goods in an attempt to make a profit.

>Actually that is the definition of a business.

The definition of something is not its justification.

There's a definition for murder too (including what is an effective murder, e.g. not getting caught, etc), doesn't mean it's ok to do it.

I don't even know how to respond to someone comparing murder to running a business. So we'll end this discussion here.

Probably it's for the best, because I also don't know how to respond to someone feigning a moral outage because of a casual comparison meant to illustrate a point, and not to say that business = murder.

While agree that there's some truth to this statement, a counter argument would be that this is exactly the reason there are so many trivial lifestyle related platforms and services that solve a problem that really only 20-30 something well off Bay Area software developer types tend to have.

This may be one of those cases where living outside the Bay Area (and being in a different demographic) actually provides a competitive advantage.

So basically Twitter, Facebook, Instagram, SnapChat? Not one of those interest me in my thirties. Obviously there has to be some substance to what you're creating, but you actually wanting to use it, is a big driving factor.

I write software that helps big companies solve issues with printing. My small software business doesn't have these issues because we print almost nothing. Does that mean I should ignore this market?


1. Launch with cryptocurrency built-in


and Pokemon.

This is hilarious—I just added support for Pokémon Go users to our app (launching next month) yesterday.

Step 3: Profit!


  Array.from(document.querySelectorAll('b')).slice(1, -1).map(b => b.innerHTML.replace(/^.*?\.\s*/, '')).join(' ')

All your code proves is you don't get my point. If those 13 sentences could stand alone, they wouldn't need all the explanatory paragraphs.

Here's my Startups in 1 Sentence: Know what's important and execute. I'm being half facetious, but also half serious. My sentence is more universally applicable than the 13.

See how silly this click-baity title game gets?

:D Nope it's not, yet I made one based on the article and tried to keep it to 13: https://www.youtube.com/watch?v=aGcwmJvtWVs

"just as you'd be careful to bend at the knees when picking up a heavy box"

Wait, have I been doing this wrong?

I thought you were supposed to bend at the knees.

Careful to bend at the knees, not careful not to bend. You and the post agree that bending at the knees is correct.

Registration is open for Startup School 2019. Classes start July 22nd.

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