When Retargeting Stops Working

Retargeting always looks like the best performing part of an ad account, and that appearance is largely an artefact of how attribution works rather than a measure of what the ads did.

Understanding why changes how the budget gets split.

Why the numbers flatter it

Retargeting reaches people who already visited your site. Many of them were going to come back anyway.

When they do, the retargeting ad is usually the last thing they touched, so it collects the conversion. The ad did not create the demand, it stood next to it at the moment of purchase.

That is why retargeting return on ad spend is high in the dashboard and why increasing the retargeting budget so rarely increases total sales.

The test that settles it

Hold out a portion of your retargeting audience. Show them nothing.

Then compare conversion rates between the group that saw the ads and the group that did not. The difference is the actual incremental effect, and for most accounts it is dramatically smaller than the reported return.

This takes a couple of weeks and it is the only way to know. Every other method is inference.

What retargeting is genuinely good at

Abandoned carts and abandoned enquiry forms, where there is a specific unfinished action and a clear reason to return.

Long consideration cycles, where staying visible for months has real value.

Specific objection handling, where the ad addresses the thing that stopped them rather than repeating the offer.

Post-purchase, for the next product or the renewal.

Where it is wasted

Broad site visitor audiences with a generic ad. Anyone who loaded the homepage once, shown the same message for ninety days.

Windows that are too long. Someone who visited seventy days ago has a different intent to someone who visited yesterday, and treating them identically wastes the budget on the first group.

Frequency without a cap. There is a point where repeated exposure stops persuading and starts irritating, and it arrives faster than most accounts assume.

Audiences too small to matter. Retargeting a few hundred people consumes management attention out of all proportion to what it can return.

Prospecting is the constraint

The size of your retargeting pool is set entirely by prospecting. Cut prospecting and the pool shrinks with a delay, so retargeting keeps performing for a few weeks and then quietly stops.

Accounts that shift budget toward retargeting because it looks efficient are consuming an asset they have stopped replenishing. The decline arrives a month later and gets blamed on the platform.

A sane structure

Most of the budget on prospecting, because that is what creates future customers and future retargeting audiences.

A smaller portion on retargeting, segmented by recency and by action, with different messages for someone who abandoned a form and someone who read one article.

Then measure the whole account against total revenue rather than against platform-reported return, because the platform figure counts the same conversion in more than one place.

The single number to watch

Total new customers against total spend, across everything.

If retargeting return is climbing and that number is flat, the budget is being moved toward the part of the account that reports credit rather than the part that creates demand.