How we'd start another company
Spoiler: it's not by building a product
Getting the wrong things done
Many founders begin their startup like this: they think up a product idea, then decide to find some co-founders to help them build it. If they’re technical founders, they’ll often spend a lot of time building it. After all, this is what they’re good at.
In this post, we’ll argue that this is the complete opposite of what they should do. Borrowing from Maslow’s Hierarchy, where the foundation is the most important part, the founders’ true hierarchy of priorities should instead be:
Founding Team
Who do I want to spend the majority of my waking time with?
Firstly, solo founding sucks - just look at the numbers. But I’d also like to indulge myself in an abstract point. If you want to go straight to the practical bit, skip the next two paragraphs.
Companies do not exist in the material world. Try to look for Apple. Where is it? Their buildings aren’t “Apple”. If you gathered all its employees into one room, that’s not “Apple” either. Look hard enough and you’ll realise there’s nothing to find, because “Apple” only exists in people's minds: its employees, customers, shareholders (and the broader public of course). If everyone stopped believing in Apple as an entity, it would cease to exist.
At their birth, startups don’t (usually) have many customers, employees or shareholders. The only people who believe in the startup are the founding team, and this may be the case for quite some time. It’s important to acknowledge that an early-stage startup is no more than the founding team.
It’s also the most important factor in deciding whether you’ll find commercial success. On a day-to-day basis, how much you enjoy the work (and therefore how much you’ll do) mostly comes down to how much you like your colleagues who, in the early days, are just the founders. You spend a ton of time together, so it’s critical you enjoy that time. Otherwise, you’ll subconsciously find bullshit reasons to not spend time together and therefore not do the work.
Changing the founding team is not a pivot, it’s bigger than that. It means effectively forming an entirely new company. It’s possible the other layers, like mission and customer, remain the same. But, if you change the founding team you should consider these other parts again.
Mission
Why did we start this company?
A company’s mission is its reason for existence. While the mission is critical, it should be derived from the founding team i.e. you should agree on the mission together. Good missions will resonate strongly with some people, but not with others.
For example: the mission “Help all companies get value from AI” resonates with us. We like AI, and we think there’s huge potential for good but it’s untapped by most companies that can’t afford expertise. However, a different founding team may not care about AI, or even think it should be suppressed. This mission wouldn’t resonate with them and they wouldn’t be motivated to found a company based on it.
When thinking about mission, try to answer the following questions:
If we succeed in our mission, how will the world look different? This should be very tangible. For the AI mission above, we imagine all cafes and restaurants getting value from AI as easily as they can set up a Squarespace website today.
How would we feel if we spent 5 years working on this but didn’t find success? Hopefully, you’d still feel it was time well spent. This is very important, because startups are risky, missions are ambitious and the likelihood of failure is extremely high. Therefore, you have to feel like failure would still be a good use of your time.
A startup can survive if the mission changes, but the founding team all need to agree on that. Changing the mission is the biggest pivot of all.
Customer
Who are we selling to so we can reach our mission?
A business needs to find customers who want to pay you. Arguably, paying customers are the only thing a business needs - there are many successful businesses without a “product”. We spent too long ignoring this obvious point as first-time founders, instead focusing on vanity metrics like incorporating, renting office space and raising money.
Loving your customers is critical. It’s a major tailwind for reaching them, selling to them and solving problems. But the customers you pick should be derived from the mission: after all, something needs to be motivating you to reach them.
There are likely to be several customer segments you could choose from to achieve a certain mission. Taking the “help all companies get value from AI” mission above: the most obvious customer is companies themselves. However, this may not work for some reason. Instead, you may decide a better route is via education: if we can teach kids in schools how to use AI, then they’ll grow up, go to work in companies and it’ll be mission success! In this case, a legitimate customer could be schools or parents.
Changing your customer is a big pivot, but not as big as changing the mission.
Problem
What problem will we solve for our chosen customers?
People will only pay you if you solve problems for them. Crucially, the problem you pick should be derived from the customer you’ve chosen. While different customer segments may have the same problem, it’s often much more severe for some than for others.
This may seem obvious, but we made this mistake with our last startup. Weightlifting form in the gym was a huge problem for expert lifters like us. However, it was a much smaller problem for the vast majority of people. Instead, they lacked the motivation to work out consistently. While form was a problem for them, it didn’t matter unless they could motivate themselves to work out in the first place.
People won’t pay for problems they don’t know they have. Therefore, it’s important to frame whatever problem you’re solving in the customer’s language. It’s a huge red flag if you can’t do this, because it suggests the customer isn’t aware of the problem.
For example: you may know the problem most companies have with using AI is they’ve mislabelled their data. You pitch them something like “We’ll correctly label your data for you”. However, they don’t want to buy it: when they talk about AI, all they know is “our model isn’t working reliably”. Instead, you’ll need to pitch them on solving the “model isn’t working” problem. That pitch may look quite different.
You should change your thesis on what problem you’re solving more often than you change customer segment. If you change the customer segment, you’ll need to learn about a whole new stack of problems, and there’s no guarantee the problem you were working on before is top of their stack.
Solution
How will we solve this problem for them?
This part often gets the most attention from founders. It’s usually the technically interesting part, and emotionally easier to focus on. After all, your product doesn’t throw rejection at you in the way that customers (or potential cofounders) often do.
There are many possible solutions to a new problem. If your initial solution isn’t working, you can pivot to a new solution without throwing out all that you’ve learned about the customer and their problems. Your solution idea is initially just a vehicle for speaking to customers, so it needs to be plausible, but don’t view it as anything more than that. And you should be ready to throw it out if (more likely when) you learn customers don’t want it.
Spend as little time building the initial solution as possible. This is for two reasons: firstly speeding up time to market (and therefore speed of learning). But second: the sunk cost fallacy is extremely powerful, so the more time you spend building the less willing you’ll be to throw something away if the market doesn’t want it. And the odds are that the market doesn’t want your initial idea. Your job is then to learn why.
The right priorities
It’s not worth overthinking your solution at the start of the journey. That effort would be much better spent:
Searching for a strong cofounder
Agreeing on a mission together
Finding ways to speak to customers
Learning about their problems
While getting your priorities right is no guarantee of success, it should hugely improve your odds. In the uncertain world of startups, that’s the best you can do!