The Ghostart SME Visibility Panel — Edition 1

What we're testing

Preregistered descriptive cuts of AI answer-engine visibility across 240 UK SME businesses in 9 sectors, per the cuts bank signed off 9 Aug 2026 (#2090). No directional hypothesis: Edition 1 reports rates with N and confidence intervals, nulls included.

We're testing it because we don't yet know the answer.

What publishes, and when

The question above was recorded before any data was collected, and it won't change to fit the data. The results publish here whatever they say, with the full method and, where we can share it responsibly, the data behind them. Until then, nothing early: no “early data suggests” figures while the study is running.

How the panel was built, where its bias is, and which cuts publish are on How Labs works.

The rules every Labs study follows are on How Labs works.