Winston arrived asking about advertising performance. That is almost always the first question, because the ad account is where the spend is visible and therefore where the problem appears to be.
In week one we told them the ads were not the problem. The checkout was.
Why that was an unpopular answer
It is an uncomfortable thing to hear when you have come in with a media brief, and the team pushed back on it — reasonably. Advertising is measurable, immediate and easy to change. Checkout is further from the budget and closer to the engineering roadmap.
But a conversion problem at the end of the journey makes every pound of acquisition spend worse, and no amount of creative testing upstream compensates for it. Spending more to send more people into the same friction is the most expensive way to grow.
What changed
We addressed the checkout first and left the advertising broadly alone, so the effect of the change could be attributed cleanly rather than lost in a simultaneous media overhaul. Once that was working, the rest of the journey began performing the way the team had expected it to all along.
Where it landed
Revenue grew 41%, with a 17% increase in ROI and 23% better ROAS efficiency. The advertising that was fine in week one became genuinely productive once what sat behind it stopped losing people.