Retargeting is almost always the best-looking line in a last-click report. That is not because it is the best channel. It is because of where it sits. Retargeting shows ads to people who already visited your site, already know your brand, and in many cases were already planning to buy. When they click one more ad on the way to checkout, last click gives that ad all the credit. The prospecting ad that brought them in the first time gets none.
Over a few months, that pattern quietly reshapes a budget. Prospecting looks expensive, retargeting looks efficient, money moves toward retargeting, and the pool of new visitors to retarget starts shrinking. Revenue flattens, and the report still says retargeting is your best channel. I first wrote about this in last-click attribution is starving your top-of-funnel, and it is still the most common measurement mistake I see.
Why last click flatters retargeting so much
Three things stack up. Selection, because retargeting audiences are people with high purchase intent to begin with, so their conversion rate would be high with or without the ad. Position, because a retargeting ad is often the last paid touch before a purchase that was going to happen anyway. And view-through credit inside the ad platforms, where a retargeting impression someone scrolled past can still claim the sale. Each one inflates the number. Together they can make a channel with little incremental effect look like your most profitable spend.
The signs it is happening to you
Look at new-customer share. If most of the orders credited to retargeting come from returning customers, the campaign is largely reaching people who already buy from you. Compare first click and last click for the same period. A channel that gets far more credit under last click than first click is closing, not opening. And watch what happens to total revenue, not channel revenue, when you shift budget. If you moved money into retargeting and total sales did not grow, it was collecting credit rather than creating sales.
How to measure retargeting properly
The direct answer is a holdout. Exclude a random share of your retargeting audience from seeing ads for a few weeks and compare their purchase rate with the group that did see ads. The difference is what retargeting actually adds. Meta’s lift testing can run this for eligible advertisers, and a simple version is possible on most platforms with audience splits. Most brands that run it find retargeting adds something, just much less than last click claimed. I walked through what to expect in is retargeting ROAS real?.
Day to day, switch your primary view away from last click. A multi-touch model shares credit with the touches that started the journey, which brings prospecting and retargeting closer to their real contribution. It is not perfect either. No click-based model can tell you what would have happened without the ad. But it stops rewarding position alone. The trade-offs between models are in what are the attribution models, and which one should you use?.
Then judge retargeting on new-customer metrics, not blended ROAS. A retargeting campaign that converts first-time visitors into first-time buyers is doing real work. One that mostly reaches past customers belongs in your retention budget and should be judged against email and SMS, which do the same job far more cheaply.
What a sensible retargeting budget looks like
There is no universal number, but for most brands I have worked with, retargeting ends up as a modest share of paid spend once it is measured honestly, with frequency capped and recent purchasers excluded. If yours is a large and growing share and your new-customer count is flat, that is the pattern to break. Move budget back toward prospecting gradually, watch total revenue and new-customer counts rather than channel ROAS, and give it a full purchase cycle before judging.
Any attribution tool that offers multiple models and splits new from returning customers can show you this. ThoughtMetric, which sponsors this blog, reports first touch, last touch, linear paid, position based, and its own multi-touch model, with new-customer ROAS by campaign, which makes the retargeting gap easy to see.
The short version
- Last click gives retargeting credit for intent it did not create.
- Check new-customer share and the gap between first and last click for retargeting campaigns.
- Run a holdout on your retargeting audience to measure what it really adds.
- Use a multi-touch model as your daily view and judge retargeting on new customers.
- Shift budget back to prospecting gradually and judge it on total revenue.
Leave a Reply