Lessons I learned at Prolific - part 1
After some time to reflect on my experiences at Prolific, I wanted to write about some of my reflections on building companies. Partly to help myself think this through, and to share some of the things I learned that might be useful for others.
These are some of the biggest lessons, that have stuck in my mind from Prolific. One is that only a few really important events and decisions really made the difference to success, surrounded by a lot of noise. This was hard to see at the time, but hopefully this might help you spot some of those inflection points ahead.
I think this topic is going to turn into 2 or 3 posts, and then I’d like to dive into some of these areas and others in more details
Let me know what’s interesting here that you’d like to hear more about
Why join a startup?
Why are you doing this in the first place? Founding or joining an early stage startup isn’t for everyone. If things go well it’s going to be HARD at times, and particularly if you are the Founder you are going to have to stick through those periods.
So you should know what you want to get out of it. I think the best reasons to do it are, to have a positive impact on the world by building something of value to others, and have a lot of learning and growth experiences doing it. These will come much faster than in a conventional role.
Even if you didn’t get into it for money and status, when things start going well it’s easy to get distracted by those things, and lose sight of why you were doing it. You need to consider when you’re no longer getting the right things out of the experience.
When I joined, I remember telling one of the founders Prolific, that I enjoyed extreme experiences. Little did I know what the years ahead would bring…
The risk/reward profile of joining as an early employee versus founding is different. Honestly, it’s probably not a good deal as an early employee, unless you are getting something other than money from the experience, or you’ve joined a guaranteed rocket ship.
So it’s best that you care about the mission of what you are building, or you are looking to learn and develop fast. Early stage startups need missionaries over mercenaries, later on as you scale and professionalise, you need mercenaries, but these require a different style of management and culture. If you are early in your career you are going to get a lot of chances to learn fast, and take on more responsibility quickly.
There'S no one way to succeed
Peter Thiel put a twist on Anna Karenina - "All happy companies are different: each one earns a monopoly by solving a unique problem. All failed companies are the same: they failed to escape competition.”
There are lots of ways to succeed, because successful startups by their nature have done something different from everyone else. There are common patterns in successful startups, but looking at these can be misleading, you don’t need to be in SF, have dropped out of Stanford, raise venture capital or appear in Tech Crunch to win.
These constraints can power your differentiation and success in important ways. Maybe you’ll learn to be capital efficient, get really good at hiring outside major centres, or solve hard regulated markets - “The obstacle is the way”.
Prolific was an unusual business, initially focused on academia as a market, bootstrapped for a long time (2014-2019 before going into YCombinator), based in Oxford, then 100% remote. Bootstrapping taught us capital efficiency that became desirable after the ZIRP era collapsed in 2022, but it took some time to adjust to being a venture backed business and learning to deploy capital well faster.
These more wandering paths to success seem common for successful non-US companies, where you can’t usually follow the VC funded hypergrowth path to winning.
The key is to stay alive long enough to get the opportunities (like staying in a poker tournament, until you eventually get dealt some good hands). As Paul Graham writes “Startups rarely die in mid keystroke. So keep typing!”
At Prolific we stuck around long enough, for the AI market opportunity to come into existence, fit the core product value we’d built for academia, and drive our next phase of growth.
On differentiation, it’s important to to decide what you believe different from other people, and when you should differentiate. Most of the time you should probably follow the market best practice, and not waste time re-inventing things, except for the few critical times you should do something unique.
The biggest mistake I saw from new hires coming into Prolific, was trying to apply the pattern of success from their previous startups, when it wasn’t appropriate to the context. And consistent hiring mistakes I made were overrating previous success in a different context. Now I think about Taleb’s surgeon paradox when hiring.
Network effects are powerful
What do I mean by network effects?
Network effects occur when a product or service becomes more valuable as more people use it. For startups, this creates a powerful growth engine where each new user increases the value for all existing users.
Network effects are so powerful that companies with strong network effects can do almost everything wrong and still win. Twitter is a great example, throughout most of its existence management has done almost everything wrong, and yet how many times have you seen users say they are leaving to Threads, Mastodon, Bluesky etc. none of these can begin to compete with Twitter’s network effects. As Mark Zuckerberg said "Twitter is such a mess – it's as if they drove a clown car into a gold mine and fell in”
2 types of network effects were critical at Prolific.
The network effects on the platform that drove growth
All kinds of positive network effects exist for Prolific as a marketplace platform; More participants, more data, more researchers, more studies, better data quality etc. all interact in positive feedback loops.
We spent a lot of time thinking about how to optimise these loops early on, creating very powerful self driving growth dynamics.
The network effects of ecosystems, we leveraged as a business
Joining YCombinator was the key acceleration moment for Prolific. YCombinator itself benefits from many network effects in its community of startups, leading to its position as dominant accelerator, in a power law effect.
It is much easier to build massive high growth startups as part of the biggest cluster, SF than it is elsewhere. I slightly regret not shifting more focus to the US in 2019, and building a US team then. I believe this would have allowed us to grow much faster in the US market, getting further ahead of competitors, being better positioned for the AI market opening up etc. I was very excited that we managed to open our NYC office before I left Prolific in 2024. There are many intangible benefits from being part of a bigger network that are hard to quantify, but compound towards success.
Some of the constraints we experienced building the company from the UK made us stronger, particularly working through challenges to succeed in European markets. While I’m an optimist about the UK and EU building more effective startup clusters, right now they are a long way behind the US, and for those that want to build huge, truly world changing companies, the US is still the place to be.
Teaching a horse to sing
Sometimes situations seem completely unfixable, but you’re still not dead. All startups experience these existential moments early on. During one such period at Prolific, I told the leadership team a story about teaching a horse to sign (which I’ve heard ascribed to Herodotus, although that’s probably not true).
”A condemned man who buys himself time through an impossible promise when sentenced to death by a king. To save himself, he promises to teach the king's favorite horse to sing within one year.
When his fellow prisoners mock him for making such an impossible promise, he responds - "I have a year now that I didn't have before. And a lot of things can happen in a year. The king might die. The horse might die. I might die."”
The point is that context changes. Your situation might truly be impossible now, but your market could change, your competitors could go bust, new technology could come along, just don’t die, and new solutions and opportunities may appear.
People really like stories, they are a very powerful communication and motivational tool. But we also broke the situation down with some first principles approaches.
We built a simple predictive model for the different possible outcomes for the company with a % assigned to each. I’m a believer in focusing on the optimal path first, even if least likely and trying to understand what needs to happen on the critical path to get there, and working to fix those blockers.
Situations like this can be allowed to drift on dangerously, until the situation worsens, so it’s also important to set checkpoints, to adjust your model and path analysis. Eventually things can change, and perhaps a horse can learn to sing.
