Should you test a new ad design when the phone preview is broken?

Inspect the actual placement

For a fictional product poster comparison, preview both candidates in the intended phone placement. Include captions, buttons and surrounding controls. A source canvas with generous margins may still lose its headline when the publisher crops it differently. A screenshot from an editing tool is not the final placement.

Record the file and layout actually shown. If one candidate is sharper because a larger rendition was requested, that difference belongs in delivery review before it is credited to a better composition.

Separate defects from hypotheses

The sheet prevents a failed implementation from being called an audience preference. It also preserves real creative differences for a later interpretable comparison.

Retest the repaired delivery

Check the smallest supported layout, the default option and any campaign-selected variant. Use the same practical quality standard for all candidates. If a destination cannot show the design reliably, adapt the composition or choose another delivery method.

Record the comparison boundary

State which defects were removed and which creative variable remains. Do not reuse performance numbers from the broken version as if they came from the repaired asset. Use cross-channel preflight and ratio and pixels for delivery planning.

A validity gate example

Two fictional ad candidates use the same offer, but one loses its headline under the destination’s controls. Mark that candidate as failing delivery review. Repair the text position and inspect the exact uploaded file before comparing it with the other candidate. Keep the broken and repaired versions separately identified in reports. Otherwise a weaker observed result could be attributed to the concept when people never saw its intended message. The gate is practical: both assets should communicate as designed before a design conclusion is drawn.