Case Study: Why 2 out of our 4 Videos Failed?
Keep reading to understand not only where we succeeded, but also what the failed creatives taught us.
A Google YouTube Demand Gen campaign executed in a much tougher market, with four video creatives tested within the same campaign environment.
The four creatives received materially different levels of distribution and audience response, providing evidence that scripting, opening structure and production quality influenced campaign performance.
A historically strong industry faced much tougher market conditions in 2026.
Global tensions, geopolitical uncertainty, trade pressures and broader economic volatility weakened confidence and made demand more difficult to generate.
The change increased the difficulty of generating qualified demand and placed greater importance on how effectively the campaign could attract and retain attention.
The campaign was therefore executed in a market where established approaches could no longer be assumed to produce the same response.
Four video creatives were deployed within a single Google Demand Gen campaign.
All four videos formed part of the same campaign and the same objective. The purpose was to test different scripting and presentation approaches within the same media environment.
One campaign. Four video creatives.
Keeping the four executions within the same campaign environment created a direct basis for comparing how individual creative assets performed under the same campaign structure.
Distribution was not evenly shared across the four videos.
Google delivery concentrated disproportionately on specific creatives, creating a useful comparison without introducing differences in campaign structure or objective.
Campaign delivery reached approximately 449K impressions and 41K interactions.
Confessing Where We Failed
Creatives 02 and 03 failed to generate comparable audience response. The explanations for what was different about those executions are outlined below.
The presenter was already familiar to the audience through the brand’s organic content. The video deliberately opened in a way that suggested she had previously provided incorrect information. This created a specific information gap: the viewer needed to continue watching to understand what had been wrong. The video then transitioned into the importance of listening to expert guidance.
This execution emerged from a combination of client-side and agency-side opinions. The resulting approach differed from the executions that received stronger distribution within the campaign.
This execution involved input and collaboration from another party rather than being developed entirely from the agency’s strategy. Its delivery and audience response were materially weaker than the two videos that received the strongest distribution.
The campaign launched in July, immediately after the first half of 2026 had ended. The opening used that specific point in the calendar as the psychological context: half of the year had already passed. Rather than introducing an artificial urgency message, the script used a situation that the audience could immediately recognise from their own circumstances.
The strongest distribution was concentrated around two videos.
Two creatives accounted for the overwhelming majority of campaign delivery.
The same video production was subsequently used for Meta advertising.
Same creative production
The same four video creatives were also used for Meta advertising by the client’s in-house team.
Performance attribution remained separate
Meta performance was excluded from all campaign performance, attribution and commercial-value calculations presented in this case study.
Platform attribution recorded approximately 42 conversions.
Google campaign reporting recorded approximately 42 platform-attributed conversions. Separately, additional landing-page enquiry activity was observed during the campaign period.
Approximately 10 campaign-attributed sales were reported.
Creative selection became a measurable performance variable within the campaign.
The comparison indicates that creative development was not simply an asset-production exercise. Within the same campaign environment, differences in scripting, opening structure and production quality materially affected which videos earned distribution and engagement.
Future creative development should be based on the observed performance differences.
Develop further variations of successful opening structures.
Continue testing time-specific contextual hooks.
Use the strongest creative territories as starting points.
Retarget meaningful video engagement.
Connect landing-page enquiries with qualification data.
Continue separating platform attribution from broader enquiry activity.