Notes

How we build, and why we build it that way.

Written by the people doing the work. Method rather than commentary.

The full stack world

Point of view · 12 October 2026

AI is compressing the layers between strategy, design and execution. The work concentrates on the people who decide what is worth making and the people who make it.

Marketing needs a spine

Point of view · 12 October 2026

Measuring marketing means connecting a dollar of spend to a dollar of revenue through every step in between. As tracking is restricted, the platforms are left holding the only complete chain, and they report on their own results.

Measure the incumbent before you train anything

Method · 7 August 2026

A holdout score tells you a model is consistent with its own test data. It tells you nothing about whether the model beats the spreadsheet it was built to replace.

The approval gate is the product

AI systems · 7 August 2026

Getting a model to write a decent client update is now the easy part of the job. The hard part is the two minutes a human spends approving it, and the thing that governs those two minutes is the clock.

When you cannot call the real data source

Architecture · 7 August 2026

The numbers everyone trusts often sit behind a desktop add-in, a portal with an export button, or a licence the client has not bought. That should change the architecture. It should not stop the build.