Did the ads cause the sales, or would they have happened anyway? Now you know.

The test

A lift test on Meta, set up and launched from your account. A random group of people does not see the ads. The difference between their purchases and everyone else's is what the ads caused. The test runs on Meta's own tools. Adella designs it, watches it, and reads the result.

The result

The sales and revenue the ads caused, from your store's real orders, by channel and audience. Shown as the share of claimed revenue that was real, and as the real return on ad spend.

What it changes

Your targets are reset to the real return. Budget is ranked by real return, so an audience that only catches people who were going to buy anyway stops outranking one that brings in new buyers. The forecast curve is scaled to the real share, so “peak profit” means real profit.

Freshness

A lift result goes stale. After a set time it stops driving the targets and the page asks for a new test. A six-month-old result is never shown as today's truth.

When there is no study

The page says so. Returns are labeled “claimed, not proven”, and every budget decision carries that label until a test has been run.