In 2004 I already had a CDP. It was called “the pupils' database”

There's one line, in the section headed "Target", that stopped me: Total pupils enrolled: 465. And underneath, the breakdown, the way you'd break down a market: Little Friends, ages 6–8, 145. Under-10s, 102. Under-12s, 102. Under-14s, 62. Under-16s, 54.

The document is called Business Plan, Website, Season 04/05, version 1.0. I wrote it, early in my career, for a youth football school's website. Not one line of code in it: all plan, all trade, all assumptions.

Calling it a CDP is generous. It was a list. But a list built a certain way.

What was actually in those pages

Four sections. Photo gallery, news and standings, kit and apparel, and a pupils' database — one card per child, searchable by age group, by team and by name, refreshed monthly.

A price list, already decided: team album €20–25, single photo from €3, the "sticker photo" 10x15 at €5, the large format of the team photo at €12. Photos published in low resolution only, and orderable exclusively through the site, collected at the school office. Shirts personalized with your own name: same, site only.

An editorial cadence written down in black and white: news twice a week, results and standings once every seven days.

And then the part that, rereading it today, actually made me grin. The sponsor slots — top banner, side logos, banners inside the news pages, logos on the t-shirts — came with a two-part pricing model: a fixed up-front cost and a cost triggered on reaching a certain number of visits. And right after it, a sentence that today is worth an entire industry: the compliance and validity of the traffic data is guaranteed by an external service. That data will be made available to the sponsors.

The two things that don't hold up

First. That pupils' database held, for every child, a photo and a date of birth, searchable by name, on a public website. In 2004 that was a feature. Today it's the first thing I delete — and not to comply with a regulation, but because I learned, building the app for a school pizza night, that children's data gets handled with tongs, collected only if it's needed, and thrown away the moment it's done its job. Version 1.0 of that plan didn't even have the vocabulary to ask the question. That isn't a detail: it's the part of the trade that has changed the most, and for the better.

The second is more embarrassing. Whole sections on who the customers are, what they buy, at what price, who pays for the advertising and how we measure it. Zero lines on who builds the thing, with what, in how long, and for how much money. The plan was all demand and no supply. And those two news posts a week, every week, for a whole season — the part that later kills projects — were a line in a calendar, not a person with a name.

What collapsed, and what didn't

The plan wasn't wrong. It was unbuildable by one person, in 2004. It needed a photographer, a webmaster, a back-office system, an office clerk, and someone to sell ad space.

Today that same document is a weekend. The cost of execution has collapsed, and that's the real news of these twenty years: not that ideas got better, but that the wall between the idea and the finished thing has thinned out into tissue paper.

What hasn't collapsed by a millimetre is the line that was missing. Who updates the standings on the seventh day, all season long, including the days they don't feel like it. Machines brought down the price of starting. The price of keeping going hasn't moved, and it's still paid in people.

In 2004 I already had the plan. What I lacked were the tools, the vocabulary to protect 465 kids, and somebody to keep the database up to date.

Two of those three, I have now.