Specialist online retailer.

Inherited a paid account optimising to the wrong conversion event with no offline feedback loop. Measurement was rebuilt before a dollar of spend was touched — then paid restructured and an SEO program layered in, compounding to eight-fold organic growth inside a year.

The situation

On paper, the account was simply expensive — cost per acquisition high and climbing, and the standard responses of new campaigns and fresh creative changing little. Underneath was a measurement problem. The account was optimising to a conversion event that did not represent a customer, so the algorithm was diligently buying more of the wrong thing. With no feedback loop returning real outcomes to the platform, every optimisation compounded the error.

This is a common shape, and it is usually misdiagnosed. When the numbers look bad, the instinct is to change the media — the targeting, the creative, the agency. But no restructure can work while the account is learning from the wrong signal. The media was not the problem. The definition of success was.

What was done

Measurement was rebuilt before a dollar of spend was touched. Conversion events were redefined around genuine commercial outcomes, tracking was reconstructed so those outcomes flowed back into the platform, and reporting moved onto the client’s own data rather than the ad account’s version of events.

Only then was paid restructured — consolidated so that signal concentrated instead of fragmenting across campaigns, with budget following tracked cost per acquisition rather than platform-reported metrics. The same budget, learning from the right signal, began behaving like a different account.

With paid stabilised, an SEO program was layered in: technical foundations, content architecture matched to how the category is actually searched, and authority built steadily rather than bought. Paid delivered while organic compounded — and inside a year, organic growth had reached eight times its starting point.

Why the number is real

Both figures are measured in the client’s own analytics, against consistent definitions. The 90% CPA reduction compares the cost of a real acquisition — the same commercial outcome, measured the same way — before and after the rebuild. The eight-fold organic figure is year-on-year growth in tracked outcomes from organic search, not rankings and not estimated traffic.

That consistency is the point. A number that moves because the definition moved proves nothing; these moved because the inputs did. The account was not given a bigger budget or an easier market — it was given a truthful signal to learn from.

The pattern generalises further than this account. Most “underperforming” media budgets are learning from the wrong signal — which is why every engagement, whatever its brief, starts with measurement rather than media. An independent marketing audit exists to find exactly this kind of problem before more budget is spent into it.