90% of start-ups go bust, often due to poor planning. Founders know they should write a business plan, yet many never do – spouting lame excuses based on misconceptions and cognitive biases (see footnote[1]).

Or else, they throw together a sketchy, useless plan. Or an over-optimistic one they don’t believe. Or produce a plan way too late; ignore expert feedback; or don’t actually use the plan.

Even if the start-up survives, poor planning probably wasted lots of time, money, and opportunities.

So here I explain why you should write a business plan – and why excuses for not doing so are flawed. (And my next post explains how and when to write a plan.)

A parable

A builder has an idea – no, a vision – for a skyscraper. It’s so vivid in his head – he talks about it so animatedly, so truly believes in it – that he just knows he doesn’t need an architectural plan or that kind of crap. Besides, it might take him weeks, he’s not sure he can draw, and it wouldn’t progress things anyhow. Don’t plan, act!

Admittedly, he’s more of a wannabe builder, as he’s never actually built anything before. But he’s googled an article called Build Your Very Own Skyscraper by a top property developer (D. Trump), and can figure out the rest as he goes along. Fired up with excitement and self-belief, he buys some tools and sets to work.

What are the chances his skyscraper won’t get built, or it collapses? Maybe 90%?


You, dear reader, are the builder; your start-up is the skyscraper; and here it is, built without a plan:

or, more likely, this:

What are plans for?

An architectural plan lets you examine a building design from different angles, check that all its elements make sense (e.g. elevators, lighting, plumbing, parking), and get expert opinions (from structural engineers, quantity surveyors, planning authorities) before construction begins. You can revise the design in response to feedback, or abandon it if it seems unsafe or uneconomic. This process takes time and effort – but saves far more during construction, and cuts the risk enormously.

A business plan is a simulation

Similarly, a business plan lets founders examine a start-up idea on paper, check it makes sense, run it past experts, revise and improve it, and decide whether to go ahead. It’s a simulation of a business, in words and numbers.

Writing a plan helps you flesh out your idea; clarify your assumptions; research missing information; consider alternatives and different scenarios; spot gaps and inconsistencies; pre-empt problems. Founders are forced to think through, discuss, perfect, and agree on all aspects of the business in advance. The process takes time and effort – but it’s far better than just starting the business and seeing whether it goes bust.

The financial projections in a plan are especially important, as start-ups ultimately stand or fall on whether they make money before they run out of cash. Once calculated, these projections produce your first year’s budget, ready to go.

They also show whether you’ll need to raise money via an investment, grant or loan. Most investors will want to see your business plan. Even so, the plan is mainly for you, not investors. Before convincing them your start-up has potential, you must first convince yourself.

It’s also a crash course

Businesses often fail if their founders haven’t run one before, and don’t know how to – that is, they are incompetent. Like the builder; or perhaps even you, dear reader. For everyone starts out incompetent at difficult tasks. You couldn’t walk, read, ride a bike, play the piano, or speak a foreign language on your first attempt – and running a business is just as hard.

Writing a proper plan is a safe way to learn and practise the basics, from budgeting to market analysis. It’s a crash course in business – Business 101. Experts mark your ‘homework’ (as you should get feedback on your draft plan), which you can keep amending till you get it right.

Minimising incompetence is also why you should start a business in a domain where you (or another founder) already have some expertise. If you know about cookery, you could consider starting a restaurant, but not an airline. Otherwise you’d have to learn everything from scratch – not just business management – and there would simply be too many ways to screw up.

Lame excuses

As a business plan is clearly crucial, you’d need a shedload of misconceptions, mental blocks and attitude problems to avoid writing one. Or to throw one together that’s sketchy and useless. Here are some excuses founders often give:

“Writing a plan is boring/difficult”

Did this ever save you from doing homework? If it seems like a reason not to write a plan, then you’re not cut out to start a business – which requires immense discipline, hard work, doing things you don’t want to do or know how to do, and generally acting like a grown-up. If there’s one thing worse than homework, it’s your start-up going kaput, because you didn’t feel like planning.

“Doing achieves more than planning”

As Nike says, just do it! But sprinting off in the wrong direction without a map is counterproductive. Planning is relatively fast and cheap – for instance, online research for a business plan can teach you things in minutes that might take months or years of doing to find out.

The daftest business idea I ever came across was for a tablet-like device – a digital video player – specifically to play classical music videos. (This was before the iPad.) The founder had spent loads of time and money doing. But a little bit of planning (assisted by Google) would have told him that the classical music market is small, and the classical music video market is minuscule – so his start-up was a non-starter.

“Business is easy”

…so you’ll just figure it out as you go along. But if it’s so easy, why do almost all start-ups fail?

Maybe you’re very intelligent or well-educated, and assume that gives you the edge. But book-smarts aren’t so useful: a zoology degree won’t help you build a skyscraper. Your competitors will be smart too, and with far more relevant experience. And you may lack other key abilities, from people skills to creativity. A co-founder can help cover your gaps – but it still won’t be easy.

“Only the product matters”

Founders often focus on creating a product, to the exclusion of all else. Maybe they think only products are important, because as consumers that’s what they see, buy and use. But the rest of a business is essential too: products don’t sell, market, distribute, support, strategize, invest in, or account for themselves. Even the best products can fail, if competitors are better at these aspects.

If product development is your expertise or sole interest, find a co-founder who’s keen to run the business side. Tech start-ups often have two founders: a product specialist (CTO) and a business specialist (CEO).

Product development can also be a displacement activity from other concerns:

“I don’t want to think about it”

…because of fear and denial. Deep down, you suspect your start-up has a fatal flaw, so you don’t want to examine it from all angles. Maybe you lack business experience, so you fear unknown unknowns; or you sense looming storm-clouds, such as superior competitors. But if you don’t deal with things like this the easy way (on paper), you’ll surely find out the hard way. And if you can’t contemplate failure, don’t start a business.

“Won’t a lean plan or pitch deck do?”

Some tech founders use the lean startup methodology. This involves rapidly developing and revising a minimum viable product (MVP, i.e. prototype) to get early feedback from customers. The start-up idea is summarized on a single sheet called a ‘lean plan’ or ‘business model canvas’, which is quick to revise as you try out different product iterations. This is fine, as there’s no point writing a full plan until you have settled on a product and market.

But once you’ve established those, you should still write a proper, detailed business plan (say, ten pages bare minimum). A lean plan is no substitute – for even if customers love your product, you won’t know if the market is big enough, or whether you could turn a profit, or need to raise money, or of other problems down the line. Some people might want a classical music video player, but not for what they’d cost if you could only sell 1,000 of them. Such things need careful research, modelling, and thought – i.e. proper planning.

If you’re considering a pitch deck (a pretty PowerPoint summary to interest investors), you probably think…

“Investors may not require a business plan”

While many professional investors will insist on a proper plan, some won’t. This is because they can spot a dud a mile off; but you can’t. They’ll only meet you if your idea has potential. And they will ask you difficult questions – the same points your business plan would cover, so you’ll need to have thought it all through. If they’re still interested, they will do their own research (‘due diligence’) before deciding to invest – again, the same kind of research your business plan should include.

So you may think you can get away with just a lean plan or pitch deck. But you’ll need a proper plan to deal with difficult questions, due diligence, and most importantly, to run your start-up.

Remember, a business plan is mainly for you, not investors. They don’t risk much money with a seed round, and they expect a high failure rate; but failure will hurt you, both psychologically, and in wasted time, money and opportunities. And if you raise money from family or friends, you won’t want to hurt them, either.

“My start-up isn’t like an old-fashioned business”

“…so we don’t need boring stuff like plans, budgets, board meetings, etc.”

Many late-1990s dotcom start-ups also thought this; they went bust.

In fact, all viable businesses are essentially the same: they provide products/services which they persuade people to buy for more than they cost to produce. So they use standard structures, departments and roles (owners, board, CEO, CFO, etc.) This is done because it works.

However disruptive you may think your start-up is, it’s not that special; you haven’t invented a whole new way of doing things, or new laws of economics. The same basic principles apply to you as they do to a florist, Netflix, a private school, and McDonald’s. All need a business plan.

“You can’t predict what will happen”

“…and plans often go badly wrong, so there’s no point making one. As Mike Tyson said, ‘Everyone has a plan until they get punched in the mouth’”.

But if you anticipate the punch, you may be able to dodge it. Unplanned start-ups often have a gloved fist heading their way, which they don’t even notice – because they stepped into the ring without knowing who their opponent was, or what a boxing match involves, or even what a gloved fist looks like.

A business plan helps you avoid disaster. Of course, it won’t guarantee that your start-up will succeed (due to the vagaries of talent and luck), but it may show that it can’t succeed – that a knockout is inevitable; worth knowing before you start the business.

The right attitude

Writing a business plan can almost be fun. Again, think of it as a simulation – an opportunity to try out different ideas on paper. You may enjoy the brainstorming, discussion, research, or financial modelling. You’ll discover a lot, perhaps with a “Eureka!” moment; e.g. Airbnb was initially for convention attendees to sleep on people’s floors, until the founders realised the potential.

A plan should make you feel better about the whole thing: more confident that your start-up will work; excited about a new, improved idea; or if you abandon it, relieved that you dodged a bullet.


A business plan is a simulation of a start-up, to try it out and improve it on paper before deciding whether to go ahead. Writing a plan is also a crash course in business. Both of these greatly reduce the chance of failure.

A plan is also useful for raising money, but you’ll need one before you reach that stage.

Avoid all excuses not to write a proper, detailed plan. You’ll be glad if you do – and probably regret it if you don’t.

My next post explains how and when to write a plan.

Thanks to Cat and Ari for helpful comments. Photos from shutterstock.com

[1] Such as these, plus others linked to in the text:

New to LessWrong?

New Comment
14 comments, sorted by Click to highlight new comments since: Today at 1:59 AM

It just occurred to me that it's often a good idea to write a plan more generally, not just for business plans. The benefits are similar to the ones you laid out for writing a business plan: it's a simulation that can prompt you to consider points you wouldn't otherwise consider, evaluate the strength of your current beliefs, get feedback from others, etc. If it is true that writing plans in general is a good idea, then I think the burden of proof for writing business plans specifically shifts towards proving that there's something specific about business plans that make them a bad idea.

Indeed, I've had many other thoughts about planning in general that I'll write up at some point. Like, planning your life :)

On occasion I've used a checklist of business plan headings & points when planning entirely different things, as many of them turn out to be relevant.

I think things like planning should be taught formally at school. They are key life skills (similarly, decision making). You're given vague advice sometimes to plan (e.g. plan your essay) but not really told how to, other than one or two very basic points (e.g. make a to-do list). Similarly if you're given something to plan, e.g. some college event. Just kind of left to figure it out for yourself.

Thank you for writing this. Overall, I updated my beliefs pretty significantly. I started two startups myself and plan to start many more, so I care a lot about getting the right answers here.

It's interesting, the vibe I get from "startup culture" is pretty anti-planning. That idea has always kinda made sense to me and I've sorta just rolled with it, but now I'm thinking that the decision to write a business plan deserves more careful thought.

I think the big critique people have for business plans is that when you're an early stage startup, there's bound to be a lot of iteration, so plans are bound to change significantly. Why spend a lot of time writing out a detailed plan for a podcasting platform when you are ultimately going to evolve into a microblogging social network? Another related critique is that numeric projections are extremely error prone at the early stages. You seem to mostly agree with these points in saying that it's fine to keep it light when you still don't know what direction you're pointing, but that once you do know where you're pointing it's important to write out a 10+ page plan.

After reading your post, I may be even more pro-planning than you! I think that even at the early stages when you don't know what direction you're pointing, a 10+ page plan probably makes sense. The big thing to me is that it doesn't actually take that much time to write out such a plan! Perhaps just a few days. So thinking about the cost vs the benefit, the cost side of the equation seems pretty low to me. However, if this low-cost assumption of mine is wrong and we're instead talking about taking eg. a month to plan when you're at an early stage, then that definitely seems silly. I think the big thing is remembering to ask yourself the question in the first place, rather than proceeding with some sort of cached thought/behavior. Once you ask the question, it seems like something that you can just be pragmatic about.

Still, even with all of that said, when I try to get concrete, things feel a little bit fuzzy to me. What are some good, real world examples of people who didn't write out business plans, and suffered the consequences? What does this look like? I'm not trying to imply that when you get concrete the answer ends up being that business plans aren't that useful. I'm just saying that I personally don't have a strong sense of what those examples look like. And that I think your post would benefit from such examples, with a pretty heavy amount of detail included. Another way I think your post would benefit would be to steelman the opposing arguments harder. It felt a little strawman-y to me.

Here's an example from my own life of not writing out a business plan harming me. I started a startup called Premium Poker Tools. Before diving into it, I did spend time thinking about what the market size is, how much of that pie I can plausibly capture, whether I could expand the pie, how likely all of this is, etc. Just getting a feel of whether the EV makes sense. I don't want to say I did a half-hearted job of it. I definitely spent some time googling around and trying to figure it out. I arrived at some answers by using things like poker forum participation, YouTube views, etc. as proxies, because I couldn't figure out any hard numbers for my competitors. I felt (over)confident in this and didn't want to "waste" too much time on it, so I moved on. It turns out I vastly overestimated the size of the market. One cool thing I learned is that if you don't know the revenue of your competitors, you could look at how many employees they have and multiply by something like $150k to get a ballpark idea. My competitors are all one/few person shops, so that should have been a sign. Also, more to the point I think, if I wrote out a business plan and shared it with people, someone who actually knows about this stuff could have pointed out that my market size estimates were dubious and that there are better ways of doing such estimates. It's sad to admit this, but it could have saved me a good two years of my life if I did this.

Anyway, looking at the benefit side of the equation now, I'm a big believer in writing to think. Writing is an enormously helpful way to develop and refine your thoughts. Especially when you share it with others and iterate on it. To the extent I see this as underrated, it makes sense to me that writing out business plans would be similarly underrated.

Btw, I don't have a good segue, but I really like the idea of business plans as simulations. I feel like that really clicked for me. I also really like and agree with the idea that it's hard to predict what success, but it's actually not that hard to figure out that a startup is bound to fail. Liron talks about this as well.

As a last point, I suspect that for startups/early stage businesses, the types of things you'd want to talk about in your business plan would be pretty different from what a more traditional business would want to talk about. It probably isn't worth going into more detail in this comment, I think it's an important thing to think about. What specifically are you planning in your business plan, and why?

PS re examples, I agree they are helpful in general (and I liked your post on the topic). Looking back at my notes I see I had a few more, but didn't use them because they didn't make crystal-clear points (and I was concerned the post was getting too long/complicated already). Actually in retrospect I realize I could have just made up examples if I couldn't produce actual ones. Anyways.

Looking back at my notes I see I had a few more, but didn't use them because they didn't make crystal-clear points (and I was concerned the post was getting too long/complicated already).

Personally, I'm a fan of kinda anti-conciseness. I used to think of conciseness as a big thing to aim for, but iirc, there was some Slate Star Codex post that convinced me otherwise (possibly Non-Expert Explanation, but I don't think so). A lot of things require you to look at it from various angles and see different examples of it before it really clicks.

There is a tradeoff of course though. Even if there's a benefit of being more clear, for more tangential points perhaps that benefit wouldn't be worth it. But for the more central points, I think the benefit usually is worth it, and that it would be worth it in the case of this article, IMHO.

The argument for conciseness these days would be that people are more likely to read it. Medium published stats a few years ago from which I inferred the optimum article length is v roughly 1200 words, if the aim is to maximise total words read (readers x mean words read).

Also analysing my own Medium articles suggests people typically stop reading after about 4 minutes regardless of the article’s length.

But I take your point re examples - they improve clarity disproportionately.

Thanks for this feedback. Re real-world examples of startups that went awry for lack of a plan, the classical music video one is of a standard kind - he thought 'this would be a cool product for people like me' but never thought how many there were like him, i.e. the market size. I assume similar examples exist for each of the headings in a business plan - people who didn't think about e.g. marketing, distribution, risks, and especially financials. But though I've advised various startups, I don't know many such examples, because I find if I insist on a business plan and they don't want to write one, I tend not to hear from them again and never find out what happened. Many founders seem to find writing one an insurmountable obstacle.

I agree maybe I should steelman the opposing arguments, though I don't know how strong they are. I thought a bit about the lean startup process, as it appears to contradict writing plans, but it's largely about the earlier stage of finding a plausible idea for a business. So it's not inconsistent with writing a plan once you've found one (even though people often don't). I sense the anti-planning mindset is more of an 'attitude' than a rationale, with the negatives that suggests. I don't think I'm too strawmanny though, as the lame excuses I give seem widespread (even if not always articulated).

As it happens I also had a poker software startup a few years ago (AI for playing/training with), which never succeeded - not for lack of a business plan, it was that making the software good enough took far longer than we expected (it being a particularly hard problem, and research as much as development), and we were chasing a moving target as the standard of online poker play was increasing fast. I spent a long time estimating companies' market sizes via various proxies, including indeed multiple of employees. I had a rare chance to test this, as I'd estimated PokerStars' revenue (from various things including traffic), which no-one knew since they were very secretive and published no figures - but when they were bought by Amaya they had to file accounts, and it turned out I had been only about 10% out. This was helped though by the fact various poker companies already published accounts, from which I had been able to calibrate the proxies quite well.

Re writing to think, indeed years ago a corporate finance company gave me this justification for writing a business plan. It forces you to think it all through.

I'm not sure a startup's business plan is that different from an established business's one - I've never noticed much difference when writing them myself, but maybe haven't thought about the differences enough. (Other than that a startup is naturally more speculative, and has to create departments & processes that would often already exist in an existing business.)

Many founders seem to find writing one an insurmountable obstacle.

Hm, people seem to feel that way about writing "long" things in general, and I've always found that weird. For example, I've received "ugh, you want me to read all of that?" responses from multi-paragraph text messages before. When you actually think about how much time it takes, it isn't that long, so maybe it's the shock value, but that doesn't feel like the right explanation either.

I don't think I'm too strawmanny though, as the lame excuses I give seem widespread (even if not always articulated).

Hm, I agree. At least that it's worth directly addressing the lame excuses due to them being widespread.

I agree maybe I should steelman the opposing arguments, though I don't know how strong they are.

Even if they aren't strong enough to win out, I think it's useful to go through a longer, more thorough steelman for the purpose of educating + convincing the reader. Eg. I see the lean startup stuff + wide error bars stuff as the big counterpoints, and you only spent a few paragraphs talking about them. Well, it was actually a decent length relative to the overall length of your post, so maybe this point is tied to my other point about anti-conciseness.

As it happens I also had a poker software startup a few years ago

Cool! That makes sense about the software's complexity. It's difficult to have good estimates there. When I worked on Premium Poker Tools I spent a few weeks trying to estimate how long it'd take to complete my tasks. For a while I was doing good, but then there were some tasks where the complexity just spiraled out of control, which caused me to lose motivation and stop trying to estimate. Right now I'm really liking Basecamp's idea of uphill vs downhill work as a substitute for estimating how much longer a task will take.

Thanks again for the detailed feedback. In practice I don’t think I’m going to improve this post further as I’ve spent far too much time on it already (I find assembling coherent thoughts painfully slow), but it’s useful for my future posts.

Sure thing. That makes sense.

It just hit me that this "plans as simulations" concept finally is allowing me to articulate my opinion on whether "ideas mean nothing and it's all about execution". Ideas can be thought of as simulations. Simulations that are imperfect and don't account for various variables that pop up in the real world. When these things pop up, you have to "execute". When your simulation turns out to be sufficiently inaccurate, you have to "pivot". But the "idea" is a simulation of a lot of crucial, big picture things. It is the premise that the company is built around. That's got to be important.

To try to steelman opposing arguments, I expect that they'd revolve around the idea that in practice, startups pivot and iterate so often that ideas are always being thrown out, and it is the ability to learn and adjust that matters, not the ideas themselves. My response would be that 1) Startups ideas are high-risk high-reward. Most fail, so most of the time you end up pivoting. That doesn't mean the idea was bad though. It could still be good if the upside is high enough to outweigh the low probability of success. 2) I suspect that if people planned properly like you describe in your post, odds of success would be much higher, and ideas wouldn't be these things that are just bound for failure.

My inconclusive thoughts around this:

In the phrase 'ideas mean nothing and it's all about execution', I think by 'ideas' people normally mean (good, novel) ideas for a product/service.

But indeed they also mean visualizing their product/service succeeding (without thinking about all the implementational details, especially ways it can go wrong).

For a startup to succeed, you need a good product/service idea AND good execution. (As a good product poorly executed will probably fail, as will a bad idea well executed.) So neither is optional.

But by saying 'ideas mean nothing and it's all about execution' people often imply that (good) product/service ideas are relatively easy, and execution is the hard bit. However the fact that only 5% of patents get commercialized, and only 10% of startups succeed, suggests that good ideas & good execution are both scarce.

So maybe 'ideas mean nothing and it's all about execution' is simply a counter to inventors who assume that execution is easy and ideas are the hard bit. As they often do, e.g. individuals who file patents and assume incorrectly that they're very brilliant/valuable and lots of companies will want to licence or steal their idea.

Or it may be that those who are good at execution can recognise good product/service ideas (fairly well), even if they can't come up with them themselves. Whereas those who can come up with good ideas usually can't do good execution. So good executors are more valuable for a startup.

Yeah, I think we agree that it depends on where you draw the line between "idea" and "execution". It just seems odd to me to draw the line in the place that most people draw it, and which you refer to in your reply, as the idea only being about the product. If you think about an idea as a simulation, it'd be silly to only simulate the product part and to not simulate anything about customer acquisition.

Yes I see what you mean re drawing the line in different places. Similarly an idea for a product can be distinguished from the execution, i.e. creation, of the product.

A few times I've been approached by people who claim to have 'invented' some software, perhaps with a patent for it, but on closer question they've merely had an idea for some software and want someone else to write it. (Again assuming that the idea is the hard part.)

This illustrates that a patent is a simulation (or indeed a plan) of a product. If it were to be created, it would do XYZ. (US patents are meant to include sufficient detail to enable people with relevant skills to implement it - i.e. a complete plan - though European patents needn't.)