Bring your clients a lift you can prove
Measured against a held-out control.Or you don't pay.A held-out control on every client test, and a number you can report.
The first test we'd run on the client's highest-traffic page
The same page twice. One block different. Your traffic decides, and the control never changes.
Clicks are not a result.
CTR and ROAS never say whether the page converted better because of your work. A measured lift against a control, on their own orders, does.
One store, one test, one number.
Start with the client's highest-traffic page. You approve the variant, traffic splits, the dashboard fills in live.
The number in your report is the number in their dashboard.
Control rate, variant rate, sessions per arm and P(beat control). Nothing blended, nothing modelled.
Is there a partner rate or a multi-client view?
Not yet. Today each client store is its own account and login, and agency pricing is a conversation: book it from the onboarding flow. The part that matters for your reports, a held-out control on every test, is already here.
Your conversion rate grows against a held-out control. Or you don't pay.
One install per client store. First five attributed conversions free on each.