Screenshots

• The phone screenshots on the store page.
• The caption text in each image, read by the vision model and tagged as marketing, interface or branding text.
• The listing text, the storefront language, and the app's rating and review count.
• How often the competitors' screenshots carry particular elements, such as a rating or an award.

A 0 to 100 score over these measures. Friction reducers and variety count on their own and again inside coverage. The weakest measure other than friction reducers caps the total, and the cap is lower when no caption is in the storefront language.

• Clarity: whether each frame's message reads at a glance.
• Friction reducers: whether the set shows an outcome, the solution and proof.
• Variety: how many different messages the frames carry, rather than repeating one.
• Coverage: how many use cases, scenarios and audiences the set shows, combined with friction reducers and variety.
• Visual cohesion across the set.
• Fit between the captions and the listing text.

• Changes to a named frame, with its score on the measure that is weak.
• Captions in the storefront language for the opening frames, when they are written in another language.
• A preview video, whenever the store page has none. The engine does not check whether the leading listings in the category use one, so treat it as something to test, with Product Page Optimization on the App Store or a store listing experiment on Google Play, before keeping it.
• Elements from a fixed checklist, such as a rating or an award, that competitors' screenshots carry and these do not.

• Tablet screenshots.
• The Google Play feature graphic.
• The preview video's content.
• Whether a video lifts conversion in the app's category. The video row is not weighed against competitors' use of video.
• Installs. The score rates the images, not what visitors did.

Included on every plan: Free, Basic, Pro, Elite and Enterprise. Compare plans