When do you stop your startup? Sebastien de Halleux on the brick wall

Monday 21 September at the Wintercircus in Ghent, I had the opportunity to ask Sebastien de Halleux one question. He had just spent fifty minutes on 25 years of building: Glu Mobile to a Nasdaq listing, Playfish to an exit at Electronic Arts, Saildrone to 400 ocean robots, Graph Ventures to 300 investments. And one company that did not make it, Matchday. So I asked the question I face from time to time with the startups I coach: “When do you decide to stop?”

His answer had two walls in it.

The brick wall

Matchday had a partnership with Leo Messi, a team he calls the best in the world for that job, and metrics good enough to raise a Series A. Every investor said the same thing. You clearly know what you are doing. But my alternative is to put ten million into OpenAI and my LPs prefer that. When people you trust all give you the same answer, that is the wall. They decided not to take an extra loan to stretch it, because the same thing would happen three months later. Read the message, embody the message, take action, he says. Then the meeting every founder dreads: telling people you personally convinced to leave their jobs that it is over. He let the whole team go.

The second wall has no investors in it. His example was a mathematician in Ghent who thinks he can build something groundbreaking in AI and has to decide whether to move to where the AI companies are. My wife has a good job, my kids are happy at school, is a legitimate answer. So is: we are moving. The test is whether you are ready to change your life for it. Nobody else can answer that one for you, and he was clear it does not always pay off. Entrepreneurship is a shitty job, he said, and he does not recommend it.

Global from day one

His playbook is the global startup. At Glu they went country by country from London and when they reached the US, the carriers already had local providers who had copied their games. So at Playfish they opened San Francisco, London and Beijing on day one, with a general manager and local investors in each. Everyone saw them as the local success story, the talent pool became the whole world, and they hired 300 people in two years. His point for Ghent: Netlog had better technology than Facebook and could not grow fast enough by hiring only here.

Fundraising upstream

The best hack of the evening came from a founder preparing a trip to San Francisco. Do not send a hundred cold emails. List the funds, go through their portfolio companies, pick the founders you admire, not necessarily in your sector, and write to them: I admire what you built, we are raising, you are backed by X, would you share your experience or make an intro. That founder has the investor on WhatsApp and can make you skip the 10.000 emails a day that investor gets. Like Sebastien said: Swim upriver through the portfolio.

And when the no comes, do not stop but ask two questions. Please share the investment memo that explains the no; if they cannot, they do not have one. And recommend a fund you admire that we should talk to instead. How they react to being pushed is how they will react after they invest. Some VCs sat with him for two hours at a whiteboard explaining why they passed. Those are the ones he respects.

Two more from the same block:

• Raise from people you identify with, because if you feel out of place at the golf club you will feel out of place with them on your board.
• And when you raise, do not raise your lifestyle. Put all of it on red, whether red is marketing or AI credits. No company car, no nicer office, no more expensive beer.

Take the wrong step on purpose

A founder asked how you know you are going in the right direction. His answer is the one thing I will steal for my own. Pick a direction and go, knowing in advance it is the wrong one. Take a small step. Ask what mistake you made. Then take the next step trying to make half as many mistakes. Then halve again. Most of the time you do not hit a wall, it kind of works, and the halving compounds. He gave the CTO version: monthly deploys, then weekly, then daily, then continuous. The point is not to make a better first decision, it is to make the second one twice as good based on experience and knowledge.

The envelope

Someone asked whether you need a co-founder and how you split the equity. That got the story I liked most, and the one he never finished. Four Playfish founders around a table. Each writes on a piece of paper the amount of money that would transform their life, seals it in an envelope. Then they split the company 25 percent each. People asked whether he was mad to give a quarter to his CFO. He calls him the best CFO he ever had. His philosophy today: he is the guy who assembles the IKEA furniture, makes the coffee and brings the bags of money, and hires the rock stars for the stage.

What he did not say is what happened to the envelopes. Were they opened? Did the numbers match, and did that decide the equal split, or was the split already decided and the envelope a ritual? Sebastien, if you read this, can you please clarify ;).

Sebastien de Halleux is founding General Partner at Graph Ventures, co-founder of Playfish and Saildrone, and co-founder and shareholder of Wintercircus. This session was part of Winter Circus Collective in Ghent. I’m just making these as my personal notes for future reference. Maarten Laruelle, Tsjirpa.