Watch Them Grow

I spent 48 hours in a room with about 200 people, all trying to build a company from scratch. Pitch on Friday night. Form teams. Hand something back to the judges on Sunday afternoon.

We didn’t win. We did build something — watchthemgrow.net, a private social network for family photographs of children. Four people who hadn’t met each other 48 hours earlier, and by Sunday afternoon we had a working product.

The itch behind it was personal. My sister had started posting baby pictures to Facebook at volume, and it bothered me — not the pictures, the platform. Every one of those photographs was a child who had not consented to anything, uploaded to a company whose business model was working out what could be inferred from them. I wanted somewhere the grandparents could see the children grow up that wasn’t also a data collection exercise.

Looking back at this from years later: Cambridge Analytica was still a year from being founded when we built this, and it took another six before everyone else worked out what the pictures were worth. The site itself is long gone — the domain stopped resolving a long time ago. The problem it was pointed at did not.

Here’s what I took away.

Rogue teams are a feature, not a bug

At some point on Saturday, our team quietly ignored the assigned brief and built what we actually thought was right. We weren’t the only ones.

Startup weekends have a structure: pitch, vote, team up, build. What nobody says out loud is that the best teams often drift from the original pitch within hours. The person who pitched the idea has a vision. The people who joined have a different interpretation of that vision. The tension between those two things either destroys the weekend or produces something sharper than either version alone.

Forming and storming happens on fast-forward

Team dynamics that normally take months to surface will surface in about six hours when the deadline is real. By Saturday afternoon we’d had an argument, resolved it, and were moving again. That’s the good version. Some teams never resolve it. You can tell by Sunday morning which ones those are.

Vision and delivery are different skills

The person who can articulate why something matters is rarely the same person who can scope a two-day build. Both are necessary. The ideas that land with the judges are usually the ones where someone held the vision clearly while someone else kept the vision from being ridiculous.

Mentor feedback has a shelf life

We got feedback from a mentor on Saturday afternoon — smart, well-intentioned, and completely wrong for where we were in the process. We were six hours from presenting. He was asking about the five-year business model.

Take mentor feedback, adjust for where you are, and then do whatever you were going to do anyway.

London needs an all-night co-working space

By 11pm on Saturday we were out of places to go. The venue had closed. Coffee shops were closing. We finished in someone’s flat.

For a city that runs financial markets around the clock and describes itself as a startup hub, the after-hours infrastructure is surprisingly bad. If someone opens a properly equipped, 24-hour co-working space in Zone 1 or 2, startup weekends will thank you.