How long does a holdout take?

Two to four weeks for most Shopify merchants. The duration depends on order volume per audience; Ripplux computes the minimum sample size per experiment.

Does my campaign stop running during the holdout?

No. The holdout splits the audience: the test arm continues to receive ads, the control arm does not. You see the spend pattern in the dashboard while the experiment runs.

Prove incrementality

Move from estimated to tested. Run platform-side holdout experiments and get an answer that survives privacy change.

Problem: Pixel-based attribution rewards the campaigns closest to the conversion. Estimated lift is not measured lift.
Outcome: Ripplux runs holdout experiments on Meta and Google, ships the verdict, and updates the ROAS engine with the measured lift coefficient.

How experiments run

Ripplux orchestrates a Conversion Lift–style holdout on Meta or a Geo / Audience holdout on Google. The platform splits the audience; we read the order data from Shopify and the spend data from the platform; we compute incremental conversion rate and incremental ROAS.

What changes after

The ROAS engine recalibrates. Decisions move from estimated to tested. Future audits cite the holdout result as authoritative ground truth.

Frequently asked

How long does a holdout take?: Two to four weeks for most Shopify merchants. The duration depends on order volume per audience; Ripplux computes the minimum sample size per experiment.
Does my campaign stop running during the holdout?: No. The holdout splits the audience: the test arm continues to receive ads, the control arm does not. You see the spend pattern in the dashboard while the experiment runs.

How experiments run

What changes after

Frequently asked

See how much of your ad spend is wasted