INDEPENDENT MINDS. SHARED AMBITION.
Back to the newsfeed
Paid search2 min read

Google Ads · Status at source date: Update

App campaign experiments gain a "Video testing" subtype that compares up to 20 groups of videos

Written by Paid Media Collective
IN BRIEF

Thomas Eccel found a new experiment subtype under Experiments > Assets provided by you for App campaigns: "Video testing (Beta)" tests up to 20 groups of videos against each other to find which perform best, next to the existing Uplift experiment, which only measures the effect of adding videos and still excludes iOS apps.

App campaign experiments gain a "Video testing" subtype that compares up to 20 groups of videos ; source illustration
LinkedInView full-size image

What changed

No help page describes it yet. For you, this means app teams can compare creative inside the campaign rather than inferring it from asset reports, so shortlist video groups before the holiday install push.

A video-testing subtype can organise a larger experiment, but the number of groups is not a reason to use the maximum. The learning question should determine how many meaningful variants are needed. Twenty weakly differentiated groups can be harder to interpret than a smaller set built around a clear creative hypothesis.

Before setup, define what changes between groups and which outcome will guide the next decision. Keep the app, audience and campaign objective in view when deciding whether a variation tests a message, a visual treatment or a different offer. Avoid changing several unrelated elements and then attributing the result to only one of them.

Review each video for the placements and device experience in which it will appear. Confirm that the experiment configuration matches the intended group structure and that the reporting can distinguish those groups. A successfully created experiment does not by itself establish that the test has enough evidence to support a conclusion.

Check the documented eligibility and account access before allocating production resources. The new subtype expands the testing workflow; it does not promise a winner, a fixed learning period or a performance gain from increasing the number of variants. Plan the interpretation before launch so that the experiment produces a decision rather than only a table of results.

Sources & contributor credit

  1. Original sourceLinkedIn

Original creator unverified

Credits reflect the editor's latest changes. Original authorship has not been independently verified for this version.

Attribution evidence and limitations

Full attribution review pending.

A source link alone does not establish original authorship.

Original image and video creators have not yet been verified.

FOLLOW THE PEOPLE BEHIND THIS UPDATE
Published on this site
Article updated

This update reflects the dated source reporting. Availability may have changed. Further coverage of this same development will be added to this page.

Explore the source reporting