If creative is now a form of targeting, as we argued in part two of this series, then the volume and variety of what you put into the system matters as much as the quality of any single execution.
Performance marketing has long fixated on metrics such as cost per acquisition and return on ad spend. But as automation deepens, performance is increasingly shaped before a campaign even launches at the creative stage.
Algorithms optimise among the assets they receive. If two ads are nearly identical, the system is choosing between similar inputs. Learning remainsnarrow and performance plateaus quickly. To unlock scale, algorithms need contrast: multiple hooks, different emotional framings, varied value propositions across formats: short-form video, statics, user-generated content, founder-led storytelling, product demonstrations.
The algorithm doesn’t invent variation. It selects from what it’s given.
This is why structured creative testing before large media investment has become so important. Entering platforms with assets already validated for attention and resonance means faster optimisation, reduced wasted spend, and more efficient scaling.
Media efficiency is no longer primarily about bidding strategy. It is about creative experimentation. Brands that treat creative as a systematic input, not a one-off output, will outperform those relying on incremental tweaks.
For brand and marketing teams, the practical shift is this: creative development needs a testing phase before media spend begins, not after. Validating which concepts have the attention and resonance to perform means the algorithm starts learning from stronger inputs. It is front-loaded investment that pays back in efficiency at scale.
Testing creative before media investment is one of the highest value changes a brand can make to its performance approach. If you’d like to know how dentsu X supports that process, get in touch.
.jpeg)