Innovative startups: the 18 mistakes to avoid

We love startups 

With this intervention by Paul Graham, founder of the incubator Y Combinator developer, author and art lover, we begin a series of articles dedicated to startups innovations in the technological sector. Becoming a nation of startups and excelling in technology is one of the most shared aspirations, and one of the few on which there is a common convergence of all the governments of the world. Even in Italy there is a lot of talk about it. 

The Italian government has recently earmarked an interesting package of resources and subsidies to startups innovative. The details of the project are not yet known, which it is hoped will be emasculated by the bureaucracy that generally accompanies such public initiatives, but the step is undoubtedly commendable, because of money and aid startups they are in great need.  

Yet this commendable initiative does not fail to raise existential questions. Is it really possible to create a vibrant startup community with taxpayers' money or are other conditions, even immaterial, that a government could put in place to achieve this more important? One would have to answer that both things are needed. But the answer is not so trivial. In fact, many people believe that the government would do well to keep a safe distance from intervening directly in entrepreneurial projects of this type which pertain more to subjects operating in the free market than to those institutional.  

However, the esteemed scholar of Italian origin, Mariana Mazzucato, professor at theUniversity College London (UCL). In a 2011 book, The entrepreneurial State (also translated into Italian, Lo stato impresario, Laterza 2014) showed, in a rather convincing way that also struck Martin Wolf of the “Financial Times”, such as the start of hub innovative entrepreneurial areas, where startups are born and thrive, are generated precisely as a result of investments and public projects that go to incubate and develop the entrepreneurial energies of the place. The same   Valley, according to Mazzucato, owes a huge debt to the US government. 

Indeed, having seed capital is important, but it is not the only condition for implanting a startup destined to last. There are many other and different conditions that must occur and above all there are mistakes that the founders and management should avoid, because the money, and the public ones won't be many, run out quickly. 

Paul Graham has drawn up a reverse handbook (that is, the things not to do) of the minimum conditions to ensure a future a a startup. It is a list of 18 points that we offer below in the Italian translation. It is a brilliant reading, witty and full of food for thought, conditions that occur only if the writer really has "the ear in the bucket". 

Who is Paul Graham 

Paul Graham è developer, writer and investor. In 1995, he and Robert Morris have founded a software company, Viaweb in that 1998 was acquired by Yahoo, where it became Yahoo Store. In 2001 you started publishing essays on paulgraham.com, which in 2015 reached up 34 million page views. In 2005, together with Jessica Livingston, Robert Morris and Trevor Blackwell has fundato Y Combinator, the first incubator new type of startups. Since 2005 Y Combinator has funded over 1000 startups including Dropbox, Airbnb, Stripe e Reddit. Paul is the author seminal publications such as On Lisp (prentice Hall, 1993), ANSI Common Lisp (prentice Hall, 1995), and Hackers & Painters (O'Reilly. He graduated in jurisprudence  to the Cornell University e he got a doctorate in computer science ad Harvard. Has studied painting at Rhodes Island School of Design and at the Academy of Fine Arts in Florence. 

Premise 

Someone recently asked me and what makes it fail a startup. THEì for there I was speechless, because it's a big question. Equals ask what determines the success of a startup. One could answer that failure must be avoided. But it is not an answer and not even an answer can be given, like this, on the fly. 

Pensgoing there calmly ho concluded than having a to-do list it can be something very useful. You can transform this blacklist in a practical recipe for don't fail. It's easier to remember what not to do than what to do. 

There is only one capital mistake that can killre a startup. This: don't do something users wish. If you do what your users want, the matter will be fine no matter what is done o immense is done. If instead immense yes quello what users want, then has morti dead, regardless of what is done or not done. Here is a list of 18 things that cause that startups they don't quello that users want really. 

Un unique founder 

Have you ever noticed how few successful startups are founded by one person? Businesses too successfull who are believed to have a single founder, such as Oracle, usually disclose di have more than one. It seems unlikely that this is pure coincidence. 

What's wrong with having a single founder? For starters, it's a vote of no confidence. It probably means that the founder has not been able to agree with anyone, not even with a few friends, to start the company together with him. It's quite alarming, because friends are the ones who know him best. 

But even if friends of the founder is wrongNo. on his behalf and the company è a good bet, there is still a downside. To boot a startup it's too demanding for a'uniqueper person. Even if you could do it all yourself, there's a need of someone to deal with, talk about decisions and receive comfort when things go wrong. 

The last reason may be the most important. The drop points of a startup I'm this way abyss them that only a few could bear them in solitude. When there are multiple founders, the esprit de corps unites them in a way that seems to go beyond le themselves conservation laws of the species. Everyone thinks: "nI can't leave my friends at the bottom of the well". This is ao of the impulses more powerfull din human nature, and it fails when there is only one founder. 

Bad location

Startups they thrive in some places and not in others. There   Valley dominates, then Boston, then Seattle, Austin, Denver and New York. Toronto Canada. London, Berlin, Paris in Europe. There's not much out there. Even in one hub like New York the per capita number of startups is probably 1/20th of what it is in   valley. In cities like Houston, Chicago and Detroit it is still too small to be measured in meaningful terms. 

Why is the detachment so strong? Probably for the same reason as other industries. What is the sixth largest fashion center in the US? The sixth largest polo of the oil or of thefinance or of thepublishing? Nobody can tell. Whatever be that as it may, they are so far from the peak that it would even be misleading to call them centers say something. 

It is a question interesting including cities become hub of startsbusiness action, but the reason why startups thrive is probably the same as any other industry: they are born where They are the skills. Standards are higher; people are more in tune with the purpose of the enterprise; the type of people suitable for those tasks they want to live there; the supporting industries are there; people who si meetinga casually they are in the same business. Who knows exactly how these factors you are combine for concentration le startup of   Valley and forInstead, chase them away din Detroit, but it is clear that a determining factor is the number of startups per capita that already exist there. 

Marginal niches

Most ofaspiring entrepreneurs who ask me for advice suffer from a common problem: they prefer pick a small, obscure niche in hopes of avoiding competition. 

Observing young children practicing football, I observed that, below a certain age, they are afraid of the ball. When the ball goes to meet they retract. At eight years old Noi didnt many ball stopper as a midfielder, because every time a crossed ball came towards me, I closed my eyes and curled up more in search of protection than in the hope of stop it. 

Choosing a marginal project is the equivalent of my football strategy da child to deal with crosses. If you do something good, they will be there always of competitors, we might as well face them. Competition can only be avoided by discarding good ideas. 

I think that theimpulse to avoid big problems is mostly unaware. It's not that people, Rationally, I thought a great ideas, but then decide to pursue the smaller ones because they seem safer. The unconscious will not even leave the space of think of big ideas. So the solution could be the one of think about great ideas without be emotionally involved. We could work on thisa hypothesis: qwhich could be a as youthful  idea di startups da suggest to my neighbor? 

Ideacopyte 

Many projects of stArticleup are imitations of existing companies. The existing it's a good one source of ideas, but not the best. If you look at the origins of startups successful, there are few that have been started in imitation of some other. Where did successful entrepreneurs get their ideas? Usually from the identification of a specific problem e unsolved that have clearly identified because they have come across it personally. 

Our startup, for example, has created software to manage online stores. When we started, there weren't any; the few sites on which it could be ordered were programmed by hand with enormous development costs e la maintenance was entrusteda a external suppliers. We knew that if e-commerce was ever going to take off, these sites would have to be powered by software, so we wrote one. Quite directto as a solution. 

It seems that the best problems to solve are those that ci involving personally. Apple had happened because Steve Wozniak He wanted to build a computer. Google why Larry e Sergey they could not find la stuff online, Hotmail why Sabeer Bhatia and Jack Smith couldn't exchange emails at work. 

So instead of copying Facebook, with some variant and Facebook he rightly ignored, need searchre ideas in a'other direction. Instead of imitate the companies e linger su problems that have already been resolvedi, you have to find new ones problems and imagine ancompany that could solve them. What are people complaining about? What do you wantyear have? Here are the questions. 

Odestination

In some fields, the road to success is di have a vision and keep attachedi to this vision regardless of the difficulties encountered. The approach to stay glued is something like this with the goal of winning a gold medal at the Olympics, Where the purpose it is well defined. Lto management of a startup, on the contrary, hto something in common with the work of the scienceiato, that must follow a trail wherever it leads. 

Ebe too attachedi to the original plan, may be wrong. Startups most successful they have finished to do something abbastanza diverse from that that you were originally area codes   often so different not to seem more same thing. Need be readyi to grab the best idea when it comes. And the hardest part of all questo is often quella to set aside theidea original. 

But opening up to new ideas has to happen in the right way. Embracing a new idea every week is fatal. Is there some kind of test that can be used to verify the goodness of an idea? One is to ask whether the idea represents some sort of process improvement already started. If any new idea is able to reuse most of what has already been built, then it is probably part of a convergence process. While if it requires starting from scratch, it's a bad sign. 

Luckily, there's someone you can ask for advice: users. If think about changing direction e users post in response, then the idea is probably a good bet. 

Hiring bad programmers 

I forgot to mention: almost all the founders I know are developers. And this can immense be an advantage. Self, accidentally, some bad developer is hired, that's not it will sink the startup. In the blink of an eye the founders they can do whatever is necessary themselves. 

Eyet when I think of what killed the most of startups in the e-commerce industry in the 90s, I quickly got the answer: it was the bad developers. Many of these ventures were started by businessmen who thought the best way to make it work startups they were clever ideas and were rather overlooked in choosing i programmers to implement ite. It's not such an easy thing to do reality is a matter harder than it looks. Iincredibly difficult indeed, because businessmen don't know to understand what are the good programmers. They can't even find out the best, because no one really good wants a job that implements a businessman's vision. 

Choosing the wrong platform

A related problem (as it tends to be done by bad programmers) is choosing the wrong platform. For example, I think many startups during the bubble committed suicide by deciding to build server-based applications on top of Windows. Hotmail continued to turn on FreeBSD for years even after Microsoft bought it, presumably because Windows couldn't handle the load. If the founders of Hotmail had chosen to use Windows, they would be went under water. 

PayPal with difficulty ha dodged this boomerang. After the merger with X.com, the new CEO would wanted to switch to Windows, even after the co-founder of PayPal, Max Levchin, he had them proved that its software had difficulty climb on Windows. Fortunately for PayPal, il CEO was changed. 

Piplatform is a vague word. It could mean an operating system, a programming language, or a “framework” built on a programming language. It involves somethingto what at the same time supports and limits, such as le foundationsa of a house. 

The scary thing about platforms is that cand it are always some that appear to outsiders deresponsible choices, yet, like Windows in the 90s, they will destroy the project, se choices. The applets Java are probably the best example striking. Thatcry sembrava be the new way of develop applications. It is conceivable that Java has killed about 100% of startups that it they adopted. 

As you chooseFeatures the right platforms? The way top è enlist good programmers and leave them work. But there is a qualcosa and non-programmers couldthey would do: visitre an important department of computer science and lookking that platform, use in research projects. 

Extension of launching 

Le companies of all sizes have difficulty develop software. It is intrinsic to the medium same; the software is always 85% finished. It takes one great effort of will to pass to 100% and get something that can be given in meal to users. 

Startups they invent all kinds of excuses to delay the launch of a product. Most of these excuses are the same that people use to procrastinate daily decisions. There is always something that must still happen. Can be. But also if the software was 100% finished and ready to launch with un simple click, there would still be some reason to wait. 

A reasone to throw quickly un product is that this choice forces you to actually terminate a phase of Work. Nothing is really finished until it's released; you can see it from excitement around the release of a product, It does not matter at what level of completion it is. The other reason to launch the product quickly is that it is alone with the reaction of the users it is possible to test theidea in a signic wayficative. 

In the delay of a launching they manifest themselves different problems: it is working too slowly; Not you understand really the problem; there is the fear to face judgement deusers; there is the fear of being judged; yes it is working on too many different things; there is excessive perfectionism. However, can be fight them all with la simple decision di exit in rapid stages. 

Premature launch

There is however a problem. A launching too premature has killed more projects than it can did it un launching deferred. The danger here it lies in the fact that you can ruinare  la own reputation. If a launch too early negatively impacts the early adopters, the latter could never come back. 

So what is it the "package minimum” for one toss sure? We suggest to start-ups to think good to from do, for example identifiesndo a nucleus which result (a) useful by itself and (b) expansayable incrementally within the project. 

This is the same approach that I (and many other programmers) useYo soyor for to produce software. thoughto to the overall goal, then you starto developing un smallest subset that does something useful. This subset, It will be however developed, so in the worst case not he will throw himself dethe time. Si will find out also that implementing a subset is good for morale and helps you see more clearly functionality dethe rest. 

- early adopters to be impressed are quite tolerant. They don't expect a new product to do everything; it just has to do something right. 

Don't have anyone in mind specific user

No can be build a product for customers without understanding them. I already said that startups most successful started by trying to solve a problem identified by their founders. Maybe there really is a rule iIn this: Wealth is created in proportion to understanding the problem you are trying to solve, and the problems you understand best are your own. 

This is just a theory. But there's a reverse theory: if you try to solve problems you don't understand, It ends Fried. 

Yet a surprising number of founders seem willing to take that as a given in fact that someone, without knowing exactly who, is looking for that what they are working on. Do the founders want it? So? It is not enough? NotI'm not their the reference market. Who is allora? Are the Millennial o teenagers? Are people interested in local events? Or is it the “business” users? And which business users? THE gas stations? I cinematic? Defense contractors? 

Obviously you can build something for users who are different from us. We dido. But you have to understand that you are entering dangerous territory. It is flying above of thethe instrument e so one should (a) consciously shift the gears, instead of assuming they can be relied upon on your instincts, and (b) look inside the instrument. 

In this case, the tool is the users. When designing for other people you have to be empirical. There's no guessing what will work; they have to flush out users and measure their responses. So if there is an intention to create something for teenagers or for business users or some other group that is not the its, you must be able to talk about the project with these specs users. If not you do, we are on path wrongo. 

Collect some money

Startups most successful, at a certain point, they need money. Having more than one founder, it seems be statistically A good thing for this purpose. As you should to collect , However? 

Financing of startups it is measured on time. Every startup that isn't profitable (that is, almost all, to start with) deve have a certain amount of time before the money runs out e he introduces himselfi the need to go in search of new ones. This necessity is sometimes compared to an airstrip, like "qhow much track there is available?” It's a good metaphor because remember that when money they run out you are bound to run off the track without it  to take offare . 

Too little money may not be enough to get off the ground. What does it meanhi “flying up” depends on the situation. Usually one has to advance to an appreciably higher level: if all one has is an idea, a working prototype; you have to launch; if there is launch, there can be a significant growth. It depends on the investors, because until it is profitable, they are the ones we have to convince. 

And if you take money from investors, you have to takenand enough to get to the next step, whatever it is. Luckily you can control how much you spend as well to determine also what will be the next stepcessive. We suggest at the startup to keep costs down, At least at the start: spend practically nothing and set the initial goal of building a solid prototype. This gives maximum flexibility. 

Spendere too much

It is difficult to distinguish between spending too much and raising too little. If you run out of money, you can tell what caused it. The only way to decide how much to raise is to refer other startups. If you have raised five million and you are finishedi money, you've probably spent too much. 

Go belly up  with  too much money is not, However, more as common as it once was. The founders seem to have learned that lesson. Furthermore, today is more economic than once start a startup. Then, today, are poche le startup that they spend too. Noa of thate and we have founded lo ha done. (And not just because we make small investments, but also because we have continued to do more round.) 

The classic way to burn money is to hire too many people. This it hurts twice: in addition to increasing costs, it slows down Development   because the money that is consumed quickly they should have last longer. Most hackers understand this thing e Fred Brooks explained it in its The Mythical Man-Month (The book published in 1975 argues that adding people to a soft projecttlate ware ends up delaying it further). 

We have three general suggestions regarding recruitment: (a) don't do it if you can avoid it, (b) pay people with stock options and incentives rather than a fixed salary, not only to save valuable resources, but because you may want to the kind of people who feel committed to the project to such an extent to prefer this method of remuneration (c) to recruit only people who know how to write code ma they know also go out to look i clienti, because those are the only things that are needed at an initial moment.

Collecting too much money

It stands to reason that little money could kill the project, but it could happen also to have too much? 

Yes and no. The problem is not so much the money itself as what comes with it. As one Venture Capital said at an event di Y Combinator: "Once you take several million dollars of my money, time is running out." If VCs finance the project, they will not allow you to put your money in the bank and continue to operate as two living boys di Windows. They want and that money are needed to start working to do more money. They want that but also transfersca a office with adequate space e and they take on people. This will change the climate and not always for the best. La most people will be employees rather than founders. They won't be so busy; they will need to be told what to do; they will begin to engage in power games. 

When the company has a lot of money, he moves to the suburbs and people start to start a family and have children. 

Perhaps, more insidiously, one time that yes they are presi a lot of money becomes more difficult to change saytion. Suppose the initial plan be that to sell something to companies. After taking the money from VC, it is assumed a sales force to do it. What if now but also duringe I realize that we have to address aconsumers instead albusinesses? It's a completely different kind of selling. What happens, in practice, is that it doesn't ci se don't makee I count. More people there are, More si rimane trapped in the direction undertaken. 

Another disadvantage of large investments is the time they take to become operational. The time required to raise money grows with il cess the amount. When the amount is in the millions, investors become very cautious. VCs never say yes or no; they engage you in seemingly endless conversation. Scaling VC investments to scale is therefore a huge waste of time and requires more work, probably, than the boot itself of the startup. And no yes must wasting time talking to investors while competitors spend their funds building the products. 

We suggest to the founders who go looking of venture capital di prendere the first proposal reasonable that it is offered to him. If you get an offer from a reputable company at a reasonable valuation with no terms specifically onerous, need turn her on immediately and go ahead with the construction dethe company. Who cares about getting a 30% better deal with another VC? economically, startups are aroulette, take all or you lose everything. The search for bargains best among investors it is a waste of time. 

Bad investors

The founder must also manage the investors. Not need ignore them, why from them they might come of insights to evaluate. But no it should let them drive the company. This is the work dei founderi. If investors they think they have a vision to run the companies they finance, because they didn't start them directly? 

Pissing off investors, ignoring them, it is probably less dangerous than that give him the reins In our startup, we made a mistake similar. Much of our energy has been absorbed nedisputes with investors instead that in the improvemento dthe product. But this was less harmful of the yieldsrand their requests, which would probably destroy the company. If the founders know and what are they doing, it is better to have the attention focused on the product rather than the search for full attention of investors who do not I am directly involvedi in the product. 

The difficulty of working with investors is proportional to the amount of money they have put into it. 

The more you raise your VC capital, the more investors get un bigger check. If they have a majority on the board, they are the bosses. In the most common case, where founders and investors are equally represented and the casting vote is cast by neutral external directors, all investors have to do is bring to their side external directors and controlsre so the society. 

If things go well, this does not ha importance. Until the business marches, most investors will leave alone the founders. But things don't always work out so in startups. Investors have finished to create problems for even the most successful companies. One of the most famous examples is Apple, whose board made a near-fatal mistake firing Steve Jobs. Apparently Google has too had, in the beginning, a stormy relationship with the investors. 

Sacrificing users for (supposed) profit

In what I said at the beginning, that is se is done something users want No.n there will be problems, si could have noticed that I have not mentioned nothing about the right business moof the. It's not why i money not are importanti. I'm not suggesting ai founders to start thecompanya without the possibility of earning money in the perspective di do them all together. The reason we tell founders not to worry about the business model initially is that please people it is much more difficult than tearing down a business plan. 

I don't know why it's so hard to make something people want. It would seem like one thing simple. But you certainly can say that è very hard, looking at how few startups it they did or are doing it. 

Why to build something people want is much more difficult than obtainingand money: si dovrethey would leave the business models a after, just like you leave some element not important but difficult for version 2. In version 1, must be resolved the main problem. And the main problem of a startup it's like creating wealth and is in this equation: how many people want something x the number who want it. The concern now need not be to convert that wealth into cash. 

The companies that make it big are the ones that put their users first. Google, for example, before  did research, then builtì the product and finally he worriedò of how to make money from it. Yet some startup founders still think it's irresponsible not to focus from the beginning on the business model. They are often encouraged in this byThe investors whose experience comes from less malleable industries. 

It is certainly irresponsible not to think about business models, but it is ten times more irresponsible not to think about the product. 

Non wanting to get your hands dirty

Almost all programmers, and not just the lazy ones, prefer to spend their time writingendo code rather than manage the business convulsive to extract money from it. Early on, even Larry e Sergey (Google founders) were like that. After developing their new search algorithm, the first thing they have looked for was to sell it to someone. 

Start a company? Horrible! Most hackers would prefer work required on ideas. But as Larry and Sergey, there isn't much market for ideas. No one trusts an idea until it does it will take shape in a product and swill be used to grow la user base. At that time they'll get the money out, and several. 

Perhaps this state of affairs it will change, but I doubt it will change much. There is nothing come users to convince buyers. It's not alone a risk factor. Buyers are concrete people, and they have some difficulty to give A lot of money to of the  youngOtti just because they are Intelligent or have some good ideas. When un'idea takes shape in a startup with many users, they may tell themselves they are buying the users rather than the intelligence of the founders, and this is easier to digest. 

Se there is theintending to attract users, probably you will have to raisee from its computers and andarli a track down. It's an unpleasant job, but if it works, there are much more chance of success. In the first batch of startups we funded, in the summer of 2005, most of the founders spent all their time developing their applications. But there was also a startup in which someone rd away half the time talking to cell phone company executives, trying to bring home some chords. It succeeds to imagine something more painful for a hacker than go around making deals? But has worked, because that startup set up the business more successfulo of that group. 

Who wants to start a startup, he must convince yourself that it simply can't hack or spend all your time programming. At least a hacker will have to spend some time at cercare business. 

Dissidents among the founders

I I disagreed among the founders are on the agenda. about 20% of startups that we financed has lost On the road un founder. It often happens that we have had to transform our role in that of arbitration. We don't require it yet, but recommend founders to prepare for this eventuality so that there may be an orderly way for people to leave the business. 

However, leaving a founder doesn't necessarily kill a startup. È already successor many starsyou p successfull. Luckily, it is usually the less busy founder who walks away. If there are three founders and one it's a bit chilly, è a big problem. Self there are  two and one is chilly, or if a guy with critical technical skills leaves, il problem è even bigger. But this is too surmountable. For one who leaves a new one may come person same to you. 

Most of the disputes I've seen between the founders could have been avoided if they had been more careful about putting themselves together when starting the company. Most disputes are not due to the situation, but to people. Which means they are inevitable. And most of the founders who have been burned by such disputes probably already had doubts when the company started. Doubts should not be silenced. It's much easier to solve problems before the business is started than after. Therefore not it's wise enlist your roommate in the startup only because, otherwise, he would feel excluded. It's good not start a business with someone you don't know intends to just because he has some skills that are needed and there is a concern about not finding anyone else. People are the most important ingredient of a startupso you don't have to compromise. 

A half effort

Startups whose bankruptcy it feels more speak areGenerally do spectacular barrels. Those are the failure elite. But the most common type of failure it's the one we've never heard of, because it's a project di a couple of guys who started working without conviction and without leave the their main occupation. 

Statistically, if you want to avoid bankruptcy, the most important thing to do is give up any other day-to-day occupation. Most startup founders go belly up he hasn't given up on his day job, while most successful startup founders have ha Done. If the bankruptcy da startforgive was a disease, the Centro for disease prevention and control would issue bulletins that advise people ad avoid double work. 

This smeans that si must necessarily leave the its work? Not necessarily. I'm guessing, but I imagine many of aspiring founders of startups they may not have the type of determination what it takes to start a company, and that in the back of their mind, ne are also aware. The reason they don't invest anymore resources in their startup is that they are convinced that it isi di a bad investment of your time. 

I think also that there is a group of people who would could have happened if he had taken the leap and chosen full-time, but they didn't. I have no idea how broad this iso group, but if the group of skeptics has the kind of distribution that I think, the number of people who could have made it, had they quit their jobs, occupa probably an order of magnitude larger of that who did. 

The finding that most of startups that could succeed fail, why the founders not The dthey devote all their efforts, it matches what I see in the world. Most of startups they fail because they don't do something people want, and the reason is, in most cases, that they don't try hard enough. 

In other words, the boot of startups it's like everything else. The biggest mistake you can make is not trying hard enough. If there's a secret to success, it's yours to try. 

. . . 

Fanklin Delano Roosevelt: “Take a method and put it to the test. If it doesn't work, try another one. But above all, try!" 

Steve Jobs"The difference between a successful entrepreneur and an unsuccessful one is sheer perseverance that helps you stick it out until the end. 

Pia (Amanda Sandrelli) in the film Che non ci remained piangere while teaching to Mario (Max Troisi) to throw the ball: “Here… you have to practice. You have to try, try, try, try …are …are and then you succeed”