We’ve spent 2 years brainstorming startup ideas and another 2 working on a bad one. All in all, It took us 4 years to find a successful idea – 4 years to arrive at the starting line.
What changed after 4 years? We didn’t get luckier or smarter, we just used the right methodology.
Looking for a successful startup idea and don’t have 4 years to spare? This is the recipe.
Note: This methodology applies to B2B startups. Use it for B2C at your own risk.
Stop looking for ideas
“Let’s build a pillow restuffing subscription service”
If you are looking for something to build then you are already screwed. Simply brainstorming things to build is a common mistake – one that I’ve painfully experienced. The problem with this approach is that it might take years or forever to converge, and when it does converge you are likely to end up with a solution in search of a problem – a problem that doesn’t exist.
Instead of ideas, what we are after is a startup scheme. A startup scheme has 3 elements:
A big market (or a quickly growing one)
An acute problem
A unique or improbable insight
This is all you need to start a startup.
Here’s the scheme for my startup:
Big market - Government procurement - this market is 12% of GDP
Acute problem - The procurement process is painfully ineffective and cumbersome, as any public servant will passionately vouch
Unique insight - We’re so good that we had two
Public data can be leveraged to make government procurement effective
By charging transaction fees instead of a subscription fees we can make more money and grow fast in a slow market
These 3 bullets are enough to start building a big company.
Scheme Generation
Now that you know what you are looking for (market, idea, insight), where might you find it?
There are 4 primary ways to generate your scheme. I’ve picked up this breakdown from Assaf Wand.
Domain expertise
If someone on the team has domain expertise in a sizable domain, this is a good place to start. They will be able to come up with problems, talk to people, and understand the trends in the domain.
A good example are 100% of the cyber security companies in Israel. The founders gain deep domain expertise in their army service, and go on to leverage their expertise to build massive companies.
Another example is my startup. My co-founder worked on both sides of government procurement and had enough insider knowledge to get us going.
Need-based
The goal here is to do research to uncover a need. You pick a big market and start broad. Because you’re starting from zero, this is a long process that takes months and hundreds of conversations. Don’t cut corners.
The idea is to narrow slowly and methodologically until your schema emerges by itself. Here’s an example:
You wake up one morning and decide to focus on business operations/finance
Through conversations, you discover that there are people whose full-time job is to run payroll
You interview many of them and learn about their process and the tools they use
You find out that in small-businesses, the owner runs payroll on their own. Wasting a lot of time and making painful mistakes in the process
You discover that all payroll systems are too janky and expensive for small businesses
You set out to solve this problem and create a massively successful company called Gusto as a result
What’s the emerging schema?
Market - payroll. undoubtedly huge with many expansion opportunities around it (benefits, HR)
Acute problem - small business owners spend hours running payroll and make costly mistakes along the way
Unique insight - Nobody solves this problem in the long tail of small businesses. It’s improbable, but if there’s a way to grow cheaply in this market then that’s an untapped opportunity
New tech
New technology can be the basis of your startup scheme.
There are ways to stumble upon a brand new technology. Perhaps you’re an avid reader, just finished your doctorate and discovered something cool, or you like to hang out with researchers. It is also possible to intentionally put yourself in a position to be aware of new tech trends.
Generating a schema from a new technology is tricky because there’s no guarantee that there’s a big market and acute problem that the technology can serve. However, because new tech enables something that was impossible before, this approach can sometimes work. And when it does it works massively.
Mammoth is a nice example in this category. They have taken the academic discovery of CRISPR and created an industry around it with Mammoth being the enabling platform.
Riding a Trend
Every moment in time there is a small wave starting that will be massive in 5 years. There are a few right now that few are seeing.
Catching such a trend is not easy because predicting the future is not easy. Yet some people know how to gamble well in this game.
Similar to identifying new-tech, you can put yourself in a position to feel the waves by “living in the future.” Then, If you are able to predict a trend semi-accurately then you can generate a good powerful scheme.
Riding a trend helps you in two ways:
When something gets 10x or 100x bigger quickly, acute problems are guaranteed to emerge quickly
Knowing of a big market before anyone else does is your unique insight
Coinbase is the perfect example. By seeing the future of Crypto they just built the obvious thing in the obvious timing, resulting in something extraordinary. It was only obvious to them because they predicted the future correctly.
Scheme Validation
A scheme can lead you to build the startup of a lifetime or waste a few years getting nowhere. Scheme validation is how you avoid the latter.
You are allowed to fall in love with your scheme, but only after objectively validating it. Validating it means ensuring that you’ve got founder-market fit and testing that the 3 ingredients are actually there.
Do you have a founder-market fit? Founder-market fit simply means that you can execute on this idea at the highest caliber, with enough passion, and not give up prematurely. The only way for you to know this is to be honest with yourself
Is the market measurably big? If it is, prove it with numbers. If you can’t, then it is not. A good benchmark is $1B+/year market
Is the problem deeply acute? Get 10 relevant people to rank the problem nine or ten out of ten. If you can’t then the only real problem is that you’ve got no problem
Is your insight real and meaningful? This is the hardest part because a great insight is elusive but not wrong – It must enable something new and differentiated. Force your insight to show it’s real face through experimentation
The goal of validation is to disqualify bad schemes, but we don’t want it to disqualify good ones, that would be a shame.
The following are bad reasons to invalidate and not pursue a scheme:
Too hard or too ambitious - Counterintuitive as it may seem, the harder an idea the more likely it is to succeed. An ambitious idea is more exciting for you, for your team, for your investors, and repels new entrants
Too boring - Boring ideas are an advantage because others don’t see them or don’t like them. If you can get yourself excited about a boring idea then the borness of the idea is a plus
Too much competition - This is a great sign of an opportunity. As long as you’ve got your one unique insight and exceptional execution then competition is not a reason to invalidate your scheme
You must be maniacally obsessed with speed and execution to build a successful startup. Put this obsession in a reinforced cage and go validate your scheme first. Once validated, and not a minute before, let your hungry obsession go wild.
Advice that did not age well
“Build something that you want or solve a problem for yourself”
This is terrible advice for 2 reasons:
It ensures that you work on a selfish idea
It prevents you from putting in the effort to find unique ideas in unique places. Good startup ideas have long moved from simple apps into complex solutions within hard markets. You have to do the legwork to uncover these ideas
If you can come up with a cure for cancer then better do it before you need it.
Other pieces of advice that are better ignored:
Timing is everything - timing is very important but it’s not everything. The important thing about timing is to not be way too early or way too late. Trying to get it perfectly is impossible and not that important
Look for an easy idea or a quick win - the fact is that the harder the idea the more chance you have in creating something meaningful
Look at industries that seem broken - this is a bad idea unless you understand why they are broken. Often times broken things will outlive you by many years
Ask people for ideas - good approach if you want to spend years not finding a good idea. You must methodically interview people about their problems within a well defined space
Conclusion
Looking for startup ideas the wrong way can cost you years of your life and prevent you from building something big and meaningful.
The wrong way is to randomly look for things to build. The right way is to look for a scheme that has 3 elements: a big market, an acute problem, and a unique or improbable insight.
Startup schemes come from 4 different sources:
Domain expertise - leveraging your expertise in a big area to uncover a problem and an insight
Need-based - methodically researching for needs in a big enough space
New tech - being aware of a new technology that may change something about the world
Riding a trend - predicting a very strong trend in its infancy
Any particular scheme can lead you to build the startup of a lifetime or waste years getting nowhere. Scheme validation is how you avoid the latter: Is the market measurably big? Is the problem deeply acute? Is your insight real and meaningful? Do you have founder-market fit? Your scheme is validated when all 4 questions are answered positively in an objective manner.
With a validated startup scheme at hand you are now at the starting line. Add your own blood, sweat, and tears, and go build something big.
Learning from what you're reading? Do me a favor and share this article with a friend who will find it valuable.