How to Get Startup Ideas

http://paulgraham.com/startupideas.html

November 2012

The way to get startup ideas is not to try to think of startupideas. It's to look for problems, preferably problems you haveyourself.

The very best startup ideas tend to have three things in common:they're something the founders themselves want, that they themselvescan build, and that few others realize are worth doing. Microsoft,Apple, Yahoo, Google, and Facebook all began this way.

Problems

Why is it so important to work on a problem you have? Among otherthings, it ensures the problem really exists. It sounds obviousto say you should only work on problems that exist. And yet by farthe most common mistake startups make is to solve problems no onehas.

I made it myself. In 1995 I started a company to put art galleriesonline. But galleries didn't want to be online. It's not how theart business works. So why did I spend 6 months working on thisstupid idea? Because I didn't pay attention to users. I inventeda model of the world that didn't correspond to reality, and workedfrom that. I didn't notice my model was wrong until I triedto convince users to pay for what we'd built. Even then I tookembarrassingly long to catch on. I was attached to my model of theworld, and I'd spent a lot of time on the software. They had towant it!

Why do so many founders build things no one wants? Because theybegin by trying to think of startup ideas. That m.o. is doublydangerous: it doesn't merely yield few good ideas; it yields badideas that sound plausible enough to fool you into working on them.

At YC we call these "made-up" or "sitcom" startup ideas. Imagineone of the characters on a TV show was starting a startup. Thewriters would have to invent something for it to do. But comingup with good startup ideas is hard. It's not something you can dofor the asking. So (unless they got amazingly lucky) the writerswould come up with an idea that sounded plausible, but was actuallybad.

For example, a social network for pet owners. It doesn't soundobviously mistaken. Millions of people have pets. Often they carea lot about their pets and spend a lot of money on them. Surelymany of these people would like a site where they could talk toother pet owners. Not all of them perhaps, but if just 2 or 3percent were regular visitors, you could have millions of users.You could serve them targeted offers, and maybe charge for premiumfeatures. [1]

The danger of an idea like this is that when you run it by yourfriends with pets, they don't say "I would never use this." Theysay "Yeah, maybe I could see using something like that." Even whenthe startup launches, it will sound plausible to a lot of people.They don't want to use it themselves, at least not right now, butthey could imagine other people wanting it. Sum that reactionacross the entire population, and you have zero users. [2]

Well

When a startup launches, there have to be at least some users whoreally need what they're making—not just people who could seethemselves using it one day, but who want it urgently. Usuallythis initial group of users is small, for the simple reason thatif there were something that large numbers of people urgently neededand that could be built with the amount of effort a startup usuallyputs into a version one, it would probably already exist. Whichmeans you have to compromise on one dimension: you can either buildsomething a large number of people want a small amount, or somethinga small number of people want a large amount. Choose the latter.Not all ideas of that type are good startup ideas, but nearly allgood startup ideas are of that type.

Imagine a graph whose x axis represents all the people who mightwant what you're making and whose y axis represents how much theywant it. If you invert the scale on the y axis, you can envisioncompanies as holes. Google is an immense crater: hundreds ofmillions of people use it, and they need it a lot. A startup juststarting out can't expect to excavate that much volume. So youhave two choices about the shape of hole you start with. You caneither dig a hole that's broad but shallow, or one that's narrowand deep, like a well.

Made-up startup ideas are usually of the first type. Lots of peopleare mildly interested in a social network for pet owners.

Nearly all good startup ideas are of the second type. Microsoftwas a well when they made Altair Basic. There were only a couplethousand Altair owners, but without this software they were programmingin machine language. Thirty years later Facebook had the sameshape. Their first site was exclusively for Harvard students, ofwhich there are only a few thousand, but those few thousand userswanted it a lot.

When you have an idea for a startup, ask yourself: who wants thisright now? Who wants this so much that they'll use it even whenit's a crappy version one made by a two-person startup they've neverheard of? If you can't answer that, the idea is probably bad. [3]

You don't need the narrowness of the well per se. It's depth youneed; you get narrowness as a byproduct of optimizing for depth(and speed). But you almost always do get it. In practice thelink between depth and narrowness is so strong that it's a goodsign when you know that an idea will appeal strongly to a specificgroup or type of user.

But while demand shaped like a well is almost a necessary conditionfor a good startup idea, it's not a sufficient one. If MarkZuckerberg had built something that could only ever have appealedto Harvard students, it would not have been a good startup idea.Facebook was a good idea because it started with a small marketthere was a fast path out of. Colleges are similar enough that ifyou build a facebook that works at Harvard, it will work at anycollege. So you spread rapidly through all the colleges. Once youhave all the college students, you get everyone else simply byletting them in.

Similarly for Microsoft: Basic for the Altair; Basic for othermachines; other languages besides Basic; operating systems;applications; IPO.

Self

How do you tell whether there's a path out of an idea? How do youtell whether something is the germ of a giant company, or just aniche product? Often you can't. The founders of Airbnb didn'trealize at first how big a market they were tapping. Initiallythey had a much narrower idea. They were going to let hosts rentout space on their floors during conventions. They didn't foreseethe expansion of this idea; it forced itself upon them gradually.All they knew at first is that they were onto something. That'sprobably as much as Bill Gates or Mark Zuckerberg knew at first.

Occasionally it's obvious from the beginning when there's a pathout of the initial niche. And sometimes I can see a path that'snot immediately obvious; that's one of our specialties at YC. Butthere are limits to how well this can be done, no matter how muchexperience you have. The most important thing to understand aboutpaths out of the initial idea is the meta-fact that these are hardto see.

So if you can't predict whether there's a path out of an idea, howdo you choose between ideas? The truth is disappointing butinteresting: if you're the right sort of person, you have the rightsort of hunches. If you're at the leading edge of a field that'schanging fast, when you have a hunch that something is worth doing,you're more likely to be right.

In Zen and the Art of Motorcycle Maintenance, Robert Pirsig says:

You want to know how to paint a perfect painting? It's easy. Make yourself perfect and then just paint naturally.

I've wondered about that passage since I read it in high school.I'm not sure how useful his advice is for painting specifically,but it fits this situation well. Empirically, the way to have goodstartup ideas is to become the sort of person who has them.

Being at the leading edge of a field doesn't mean you have to beone of the people pushing it forward. You can also be at the leadingedge as a user. It was not so much because he was a programmerthat Facebook seemed a good idea to Mark Zuckerberg as because heused computers so much. If you'd asked most 40 year olds in 2004whether they'd like to publish their lives semi-publicly on theInternet, they'd have been horrified at the idea. But Mark alreadylived online; to him it seemed natural.

Paul Buchheit says that people at the leading edge of a rapidlychanging field "live in the future." Combine that with Pirsig andyou get:

Live in the future, then build what's missing.

That describes the way many if not most of the biggest startups gotstarted. Neither Apple nor Yahoo nor Google nor Facebook were evensupposed to be companies at first. They grew out of things theirfounders built because there seemed a gap in the world.

If you look at the way successful founders have had their ideas,it's generally the result of some external stimulus hitting aprepared mind. Bill Gates and Paul Allen hear about the Altair andthink "I bet we could write a Basic interpreter for it." Drew Houstonrealizes he's forgotten his USB stick and thinks "I really need tomake my files live online." Lots of people heard about the Altair.Lots forgot USB sticks. The reason those stimuli caused thosefounders to start companies was that their experiences had preparedthem to notice the opportunities they represented.

The verb you want to be using with respect to startup ideas is not"think up" but "notice." At YC we call ideas that grow naturallyout of the founders' own experiences "organic" startup ideas. Themost successful startups almost all begin this way.

That may not have been what you wanted to hear. You may haveexpected recipes for coming up with startup ideas, and instead I'mtelling you that the key is to have a mind that's prepared in theright way. But disappointing though it may be, this is the truth.And it is a recipe of a sort, just one that in the worst case takesa year rather than a weekend.

If you're not at the leading edge of some rapidly changing field,you can get to one. For example, anyone reasonably smart canprobably get to an edge of programming (e.g. building mobile apps)in a year. Since a successful startup will consume at least 3-5years of your life, a year's preparation would be a reasonableinvestment. Especially if you're also looking for a cofounder. [4]

You don't have to learn programming to be at the leading edge of adomain that's changing fast. Other domains change fast. But whilelearning to hack is not necessary, it is for the forseeable futuresufficient. As Marc Andreessen put it, software is eating the world,and this trend has decades left to run.

Knowing how to hack also means that when you have ideas, you'll beable to implement them. That's not absolutely necessary (Jeff Bezoscouldn't) but it's an advantage. It's a big advantage, when you'reconsidering an idea like putting a college facebook online, ifinstead of merely thinking "That's an interesting idea," you canthink instead "That's an interesting idea. I'll try building aninitial version tonight." It's even better when you're both aprogrammer and the target user, because then the cycle of generatingnew versions and testing them on users can happen inside one head.

Noticing

Once you're living in the future in some respect, the way to noticestartup ideas is to look for things that seem to be missing. Ifyou're really at the leading edge of a rapidly changing field, therewill be things that are obviously missing. What won't be obviousis that they're startup ideas. So if you want to find startupideas, don't merely turn on the filter "What's missing?" Also turnoff every other filter, particularly "Could this be a big company?"There's plenty of time to apply that test later. But if you'rethinking about that initially, it may not only filter out lotsof good ideas, but also cause you to focus on bad ones.

Most things that are missing will take some time to see. You almosthave to trick yourself into seeing the ideas around you.

But you know the ideas are out there. This is not one of thoseproblems where there might not be an answer. It's impossiblyunlikely that this is the exact moment when technological progressstops. You can be sure people are going to build things in thenext few years that will make you think "What did I do before x?"

And when these problems get solved, they will probably seem flaminglyobvious in retrospect. What you need to do is turn off the filtersthat usually prevent you from seeing them. The most powerful issimply taking the current state of the world for granted. Even themost radically open-minded of us mostly do that. You couldn't getfrom your bed to the front door if you stopped to question everything.

But if you're looking for startup ideas you can sacrifice some ofthe efficiency of taking the status quo for granted and start toquestion things. Why is your inbox overflowing? Because you geta lot of email, or because it's hard to get email out of your inbox?Why do you get so much email? What problems are people trying tosolve by sending you email? Are there better ways to solve them?And why is it hard to get emails out of your inbox? Why do youkeep emails around after you've read them? Is an inbox the optimaltool for that?

Pay particular attention to things that chafe you. The advantageof taking the status quo for granted is not just that it makes life(locally) more efficient, but also that it makes life more tolerable.If you knew about all the things we'll get in the next 50 years butdon't have yet, you'd find present day life pretty constraining,just as someone from the present would if they were sent back 50years in a time machine. When something annoys you, it could bebecause you're living in the future.

When you find the right sort of problem, you should probably beable to describe it as obvious, at least to you. When we startedViaweb, all the online stores were built by hand, by web designersmaking individual HTML pages. It was obvious to us as programmersthat these sites would have to be generated by software. [5]

Which means, strangely enough, that coming up with startup ideasis a question of seeing the obvious. That suggests how weird thisprocess is: you're trying to see things that are obvious, and yetthat you hadn't seen.

Since what you need to do here is loosen up your own mind, it maybe best not to make too much of a direct frontal attack on theproblem—i.e. to sit down and try to think of ideas. The bestplan may be just to keep a background process running, looking forthings that seem to be missing. Work on hard problems, drivenmainly by curiousity, but have a second self watching over yourshoulder, taking note of gaps and anomalies. [6]

Give yourself some time. You have a lot of control over the rateat which you turn yours into a prepared mind, but you have lesscontrol over the stimuli that spark ideas when they hit it. IfBill Gates and Paul Allen had constrained themselves to come upwith a startup idea in one month, what if they'd chosen a monthbefore the Altair appeared? They probably would have worked on aless promising idea. Drew Houston did work on a less promisingidea before Dropbox: an SAT prep startup. But Dropbox was a muchbetter idea, both in the absolute sense and also as a match for hisskills. [7]

A good way to trick yourself into noticing ideas is to work onprojects that seem like they'd be cool. If you do that, you'llnaturally tend to build things that are missing. It wouldn't seemas interesting to build something that already existed.

Just as trying to think up startup ideas tends to produce bad ones,working on things that could be dismissed as "toys" often producesgood ones. When something is described as a toy, that means it haseverything an idea needs except being important. It's cool; userslove it; it just doesn't matter. But if you're living in the futureand you build something cool that users love, it may matter morethan outsiders think. Microcomputers seemed like toys when Appleand Microsoft started working on them. I'm old enough to rememberthat era; the usual term for people with their own microcomputerswas "hobbyists." BackRub seemed like an inconsequential scienceproject. The Facebook was just a way for undergrads to stalk oneanother.

At YC we're excited when we meet startups working on things thatwe could imagine know-it-alls on forums dismissing as toys. To usthat's positive evidence an idea is good.

If you can afford to take a long view (and arguably you can't affordnot to), you can turn "Live in the future and build what's missing"into something even better:

Live in the future and build what seems interesting.

School

That's what I'd advise college students to do, rather than tryingto learn about "entrepreneurship." "Entrepreneurship" is somethingyou learn best by doing it. The examples of the most successfulfounders make that clear. What you should be spending your timeon in college is ratcheting yourself into the future. College isan incomparable opportunity to do that. What a waste to sacrificean opportunity to solve the hard part of starting a startup—becoming the sort of person who can have organic startup ideas—by spending time learning about the easy part. Especially sinceyou won't even really learn about it, any more than you'd learnabout sex in a class. All you'll learn is the words for things.

The clash of domains is a particularly fruitful source of ideas.If you know a lot about programming and you start learning aboutsome other field, you'll probably see problems that software couldsolve. In fact, you're doubly likely to find good problems inanother domain: (a) the inhabitants of that domain are not as likelyas software people to have already solved their problems withsoftware, and (b) since you come into the new domain totally ignorant,you don't even know what the status quo is to take it for granted.

So if you're a CS major and you want to start a startup, insteadof taking a class on entrepreneurship you're better off taking aclass on, say, genetics. Or better still, go work for a biotechcompany. CS majors normally get summer jobs at computer hardwareor software companies. But if you want to find startup ideas, youmight do better to get a summer job in some unrelated field. [8]

Or don't take any extra classes, and just build things. It's nocoincidence that Microsoft and Facebook both got started in January.At Harvard that is (or was) Reading Period, when students have noclasses to attend because they're supposed to be studying for finals. [9]

But don't feel like you have to build things that will become startups. That'spremature optimization. Just build things. Preferably with otherstudents. It's not just the classes that make a university such agood place to crank oneself into the future. You're also surroundedby other people trying to do the same thing. If you work togetherwith them on projects, you'll end up producing not just organicideas, but organic ideas with organic founding teams—and that,empirically, is the best combination.

Beware of research. If an undergrad writes something all his friendsstart using, it's quite likely to represent a good startup idea.Whereas a PhD dissertation is extremely unlikely to. For somereason, the more a project has to count as research, the less likelyit is to be something that could be turned into a startup. [10]I think the reason is that the subset of ideas that count as researchis so narrow that it's unlikely that a project that satisfied thatconstraint would also satisfy the orthogonal constraint of solvingusers' problems. Whereas when students (or professors) buildsomething as a side-project, they automatically gravitate towardsolving users' problems—perhaps even with an additional energythat comes from being freed from the constraints of research.

Competition

Because a good idea should seem obvious, when you have one you'lltend to feel that you're late. Don't let that deter you. Worryingthat you're late is one of the signs of a good idea. Ten minutesof searching the web will usually settle the question. Even if youfind someone else working on the same thing, you're probably nottoo late. It's exceptionally rare for startups to be killed bycompetitors—so rare that you can almost discount the possibility.So unless you discover a competitor with the sort of lock-in thatwould prevent users from choosing you, don't discard the idea.

If you're uncertain, ask users. The question of whether you're toolate is subsumed by the question of whether anyone urgently needswhat you plan to make. If you have something that no competitordoes and that some subset of users urgently need, you have abeachhead. [11]

The question then is whether that beachhead is big enough. Or moreimportantly, who's in it: if the beachhead consists of people doingsomething lots more people will be doing in the future, then it'sprobably big enough no matter how small it is. For example, ifyou're building something differentiated from competitors by thefact that it works on phones, but it only works on the newest phones,that's probably a big enough beachhead.

Err on the side of doing things where you'll face competitors.Inexperienced founders usually give competitors more credit thanthey deserve. Whether you succeed depends far more on you than onyour competitors. So better a good idea with competitors than abad one without.

You don't need to worry about entering a "crowded market" so longas you have a thesis about what everyone else in it is overlooking.In fact that's a very promising starting point. Google was thattype of idea. Your thesis has to be more precise than "we're goingto make an x that doesn't suck" though. You have to be able tophrase it in terms of something the incumbents are overlooking.Best of all is when you can say that they didn't have the courageof their convictions, and that your plan is what they'd have doneif they'd followed through on their own insights. Google was thattype of idea too. The search engines that preceded them shied awayfrom the most radical implications of what they were doing—particularly that the better a job they did, the faster users wouldleave.

A crowded market is actually a good sign, because it means boththat there's demand and that none of the existing solutions aregood enough. A startup can't hope to enter a market that's obviouslybig and yet in which they have no competitors. So any startup thatsucceeds is either going to be entering a market with existingcompetitors, but armed with some secret weapon that will get themall the users (like Google), or entering a market that looks smallbut which will turn out to be big (like Microsoft). [12]

Filters

There are two more filters you'll need to turn off if you want tonotice startup ideas: the unsexy filter and the schlep filter.

Most programmers wish they could start a startup by just writingsome brilliant code, pushing it to a server, and having users paythem lots of money. They'd prefer not to deal with tedious problemsor get involved in messy ways with the real world. Which is areasonable preference, because such things slow you down. But thispreference is so widespread that the space of convenient startupideas has been stripped pretty clean. If you let your mind wandera few blocks down the street to the messy, tedious ideas, you'llfind valuable ones just sitting there waiting to be implemented.

The schlep filter is so dangerous that I wrote a separate essayabout the condition it induces, which I called schlep blindness.I gave Stripe as an example of a startup that benefited from turningoff this filter, and a pretty striking example it is. Thousandsof programmers were in a position to see this idea; thousands ofprogrammers knew how painful it was to process payments beforeStripe. But when they looked for startup ideas they didn't seethis one, because unconsciously they shrank from having to dealwith payments. And dealing with payments is a schlep for Stripe,but not an intolerable one. In fact they might have had net lesspain; because the fear of dealing with payments kept most peopleaway from this idea, Stripe has had comparatively smooth sailingin other areas that are sometimes painful, like user acquisition.They didn't have to try very hard to make themselves heard by users,because users were desperately waiting for what they were building.

The unsexy filter is similar to the schlep filter, except it keepsyou from working on problems you despise rather than ones you fear.We overcame this one to work on Viaweb. There were interestingthings about the architecture of our software, but we weren'tinterested in ecommerce per se. We could see the problem was onethat needed to be solved though.

Turning off the schlep filter is more important than turning offthe unsexy filter, because the schlep filter is more likely to bean illusion. And even to the degree it isn't, it's a worse formof self-indulgence. Starting a successful startup is going to befairly laborious no matter what. Even if the product doesn't entaila lot of schleps, you'll still have plenty dealing with investors,hiring and firing people, and so on. So if there's some idea youthink would be cool but you're kept away from by fear of the schlepsinvolved, don't worry: any sufficiently good idea will have as many.

The unsexy filter, while still a source of error, is not as entirelyuseless as the schlep filter. If you're at the leading edge of afield that's changing rapidly, your ideas about what's sexy willbe somewhat correlated with what's valuable in practice. Particularlyas you get older and more experienced. Plus if you find an ideasexy, you'll work on it more enthusiastically. [13]

Recipes

While the best way to discover startup ideas is to become the sortof person who has them and then build whatever interests you,sometimes you don't have that luxury. Sometimes you need an ideanow. For example, if you're working on a startup and your initialidea turns out to be bad.

For the rest of this essay I'll talk about tricks for coming upwith startup ideas on demand. Although empirically you're betteroff using the organic strategy, you could succeed this way. Youjust have to be more disciplined. When you use the organic method,you don't even notice an idea unless it's evidence that somethingis truly missing. But when you make a conscious effort to thinkof startup ideas, you have to replace this natural constraint withself-discipline. You'll see a lot more ideas, most of them bad,so you need to be able to filter them.

One of the biggest dangers of not using the organic method is theexample of the organic method. Organic ideas feel like inspirations.There are a lot of stories about successful startups that beganwhen the founders had what seemed a crazy idea but "just knew" itwas promising. When you feel that about an idea you've had whiletrying to come up with startup ideas, you're probably mistaken.

When searching for ideas, look in areas where you have some expertise.If you're a database expert, don't build a chat app for teenagers(unless you're also a teenager). Maybe it's a good idea, but youcan't trust your judgment about that, so ignore it. There have tobe other ideas that involve databases, and whose quality you canjudge. Do you find it hard to come up with good ideas involvingdatabases? That's because your expertise raises your standards.Your ideas about chat apps are just as bad, but you're givingyourself a Dunning-Kruger pass in that domain.

The place to start looking for ideas is things you need. There must be things you need. [14]

One good trick is to ask yourself whether in your previous job youever found yourself saying "Why doesn't someone make x? If someonemade x we'd buy it in a second." If you can think of any x peoplesaid that about, you probably have an idea. You know there's demand,and people don't say that about things that are impossible to build.

More generally, try asking yourself whether there's something unusualabout you that makes your needs different from most other people's.You're probably not the only one. It's especially good if you'redifferent in a way people will increasingly be.

If you're changing ideas, one unusual thing about you is the ideayou'd previously been working on. Did you discover any needs whileworking on it? Several well-known startups began this way. Hotmailbegan as something its founders wrote to talk about their previousstartup idea while they were working at their day jobs. [15]

A particularly promising way to be unusual is to be young. Someof the most valuable new ideas take root first among people in theirteens and early twenties. And while young founders are at adisadvantage in some respects, they're the only ones who reallyunderstand their peers. It would have been very hard for someonewho wasn't a college student to start Facebook. So if you're ayoung founder (under 23 say), are there things you and your friendswould like to do that current technology won't let you?

The next best thing to an unmet need of your own is an unmet needof someone else. Try talking to everyone you can about the gapsthey find in the world. What's missing? What would they like todo that they can't? What's tedious or annoying, particularly intheir work? Let the conversation get general; don't be trying toohard to find startup ideas. You're just looking for something tospark a thought. Maybe you'll notice a problem they didn't consciouslyrealize they had, because you know how to solve it.

When you find an unmet need that isn't your own, it may be somewhatblurry at first. The person who needs something may not know exactlywhat they need. In that case I often recommend that founders actlike consultants—that they do what they'd do if they'd beenretained to solve the problems of this one user. People's problemsare similar enough that nearly all the code you write this way willbe reusable, and whatever isn't will be a small price to start outcertain that you've reached the bottom of the well. [16]

One way to ensure you do a good job solving other people's problemsis to make them your own. When Rajat Suri of E la Carte decidedto write software for restaurants, he got a job as a waiter to learnhow restaurants worked. That may seem like taking things to extremes,but startups are extreme. We love it when founders do such things.

In fact, one strategy I recommend to people who need a new idea isnot merely to turn off their schlep and unsexy filters, but to seekout ideas that are unsexy or involve schleps. Don't try to startTwitter. Those ideas are so rare that you can't find them by lookingfor them. Make something unsexy that people will pay you for.

A good trick for bypassing the schlep and to some extent the unsexyfilter is to ask what you wish someone else would build, so thatyou could use it. What would you pay for right now?

Since startups often garbage-collect broken companies and industries,it can be a good trick to look for those that are dying, or deserveto, and try to imagine what kind of company would profit from theirdemise. For example, journalism is in free fall at the moment.But there may still be money to be made from something like journalism.What sort of company might cause people in the future to say "thisreplaced journalism" on some axis?

But imagine asking that in the future, not now. When one companyor industry replaces another, it usually comes in from the side.So don't look for a replacement for x; look for something thatpeople will later say turned out to be a replacement for x. Andbe imaginative about the axis along which the replacement occurs.Traditional journalism, for example, is a way for readers to getinformation and to kill time, a way for writers to make money andto get attention, and a vehicle for several different types ofadvertising. It could be replaced on any of these axes (it hasalready started to be on most).

When startups consume incumbents, they usually start by servingsome small but important market that the big players ignore. It'sparticularly good if there's an admixture of disdain in the bigplayers' attitude, because that often misleads them. For example,after Steve Wozniak built the computer that became the Apple I, hefelt obliged to give his then-employer Hewlett-Packard the optionto produce it. Fortunately for him, they turned it down, and oneof the reasons they did was that it used a TV for a monitor, whichseemed intolerably d閏lass� to a high-end hardware company like HPwas at the time. [17]

Are there groups of scruffy but sophisticated users like the earlymicrocomputer "hobbyists" that are currently being ignored by thebig players? A startup with its sights set on bigger things canoften capture a small market easily by expending an effort thatwouldn't be justified by that market alone.

Similarly, since the most successful startups generally ride somewave bigger than themselves, it could be a good trick to look forwaves and ask how one could benefit from them. The prices of genesequencing and 3D printing are both experiencing Moore's Law-likedeclines. What new things will we be able to do in the new worldwe'll have in a few years? What are we unconsciously ruling outas impossible that will soon be possible?

Organic

But talking about looking explicitly for waves makes it clear thatsuch recipes are plan B for getting startup ideas. Looking forwaves is essentially a way to simulate the organic method. Ifyou're at the leading edge of some rapidly changing field, you don'thave to look for waves; you are the wave.

Finding startup ideas is a subtle business, and that's why mostpeople who try fail so miserably. It doesn't work well simply totry to think of startup ideas. If you do that, you get bad onesthat sound dangerously plausible. The best approach is more indirect:if you have the right sort of background, good startup ideas willseem obvious to you. But even then, not immediately. It takestime to come across situations where you notice something missing.And often these gaps won't seem to be ideas for companies, justthings that would be interesting to build. Which is why it's goodto have the time and the inclination to build things just becausethey're interesting.

Live in the future and build what seems interesting. Strange asit sounds, that's the real recipe.







Notes

[ 1]This form of bad idea has been around as long as the web. Itwas common in the 1990s, except then people who had it used to saythey were going to create a portal for x instead of a social networkfor x. Structurally the idea is stone soup: you post a sign saying"this is the place for people interested in x," and all those peopleshow up and you make money from them. What lures founders intothis sort of idea are statistics about the millions of people whomight be interested in each type of x. What they forget is thatany given person might have 20 affinities by this standard, and noone is going to visit 20 different communities regularly.

[ 2]I'm not saying, incidentally, that I know for sure a socialnetwork for pet owners is a bad idea. I know it's a bad idea theway I know randomly generated DNA would not produce a viable organism.The set of plausible sounding startup ideas is many times largerthan the set of good ones, and many of the good ones don't evensound that plausible. So if all you know about a startup idea isthat it sounds plausible, you have to assume it's bad.

[ 3]More precisely, the users' need has to give them sufficientactivation energy to start using whatever you make, which can varya lot. For example, the activation energy for enterprise softwaresold through traditional channels is very high, so you'd have tobe a lot better to get users to switch. Whereas the activationenergy required to switch to a new search engine is low. Which inturn is why search engines are so much better than enterprisesoftware.

[ 4]This gets harder as you get older. While the space of ideasdoesn't have dangerous local maxima, the space of careers does.There are fairly high walls between most of the paths people takethrough life, and the older you get, the higher the walls become.

[ 5]It was also obvious to us that the web was going to be a bigdeal. Few non-programmers grasped that in 1995, but the programmershad seen what GUIs had done for desktop computers.

[ 6]Maybe it would work to have this second self keep a journal,and each night to make a brief entry listing the gaps and anomaliesyou'd noticed that day. Not startup ideas, just the raw gaps andanomalies.

[ 7]Sam Altman points out that taking time to come up with anidea is not merely a better strategy in an absolute sense, but alsolike an undervalued stock in that so few founders do it.

There's comparatively little competition for the best ideas, becausefew founders are willing to put in the time required to notice them.Whereas there is a great deal of competition for mediocre ideas,because when people make up startup ideas, they tend to make up thesame ones.

[ 8]For the computer hardware and software companies, summer jobsare the first phase of the recruiting funnel. But if you're goodyou can skip the first phase. If you're good you'll have no troublegetting hired by these companies when you graduate, regardless ofhow you spent your summers.

[ 9]The empirical evidence suggests that if colleges want to helptheir students start startups, the best thing they can do is leavethem alone in the right way.

[ 10]I'm speaking here of IT startups; in biotech things are different.

[ 11]This is an instance of a more general rule: focus on users,not competitors. The most important information about competitorsis what you learn via users anyway.

[ 12]In practice most successful startups have elements of both.And you can describe each strategy in terms of the other by adjustingthe boundaries of what you call the market. But it's useful toconsider these two ideas separately.

[ 13]I almost hesitate to raise that point though. Startups arebusinesses; the point of a business is to make money; and with thatadditional constraint, you can't expect you'll be able to spend allyour time working on what interests you most.

[ 14]The need has to be a strong one. You can retroactivelydescribe any made-up idea as something you need. But do you reallyneed that recipe site or local event aggregator as much as DrewHouston needed Dropbox, or Brian Chesky and Joe Gebbia needed Airbnb?

Quite often at YC I find myself asking founders "Would you use thisthing yourself, if you hadn't written it?" and you'd be surprisedhow often the answer is no.

[ 15]Paul Buchheit points out that trying to sell something badcan be a source of better ideas:

"The best technique I've found for dealing with YC companies thathave bad ideas is to tell them to go sell the product ASAP (beforewasting time building it). Not only do they learn that nobodywants what they are building, they very often come back with areal idea that they discovered in the process of trying to sellthe bad idea."

[ 16]Here's a recipe that might produce the next Facebook, ifyou're college students. If you have a connection to one of themore powerful sororities at your school, approach the queen beesthereof and offer to be their personal IT consultants, buildinganything they could imagine needing in their social lives thatdidn't already exist. Anything that got built this way would bevery promising, because such users are not just the most demandingbut also the perfect point to spread from.

I have no idea whether this would work.

[ 17]And the reason it used a TV for a monitor is that Steve Wozniakstarted out by solving his own problems. He, like most of hispeers, couldn't afford a monitor.



Thanks to Sam Altman, Mike Arrington, Paul Buchheit, John Collison,Patrick Collison, Garry Tan, and Harj Taggar for reading drafts ofthis, and Marc Andreessen, Joe Gebbia, Reid Hoffman, Shel Kaphan,Mike Moritz and Kevin Systrom for answering my questions aboutstartup history.