I know. It feels productive. The ROAS looks incredible. The CPAs are low. But here is what is actually happening: you are spending money to re-show ads to people who were already going to buy.
That "8x ROAS" remarketing campaign? Strip out the people who had items in cart, who were already on your site, who searched your brand name — and your incremental return is close to zero.
What I See in Almost Every Audit
The pattern is consistent across clients and industries:
- 40 to 60% of "remarketing conversions" would have happened without the ad
- The CPA looks cheap because you are claiming credit for organic behaviour
- Meanwhile, the top-of-funnel campaigns that actually drive growth get their budget cut because the numbers "don't look as good"
This is how remarketing budgets quietly grow while new customer acquisition quietly stalls. The numbers in the dashboard look great. The business is not growing.
Why the Numbers Lie
Remarketing audiences are, by definition, people who already know you. They visited your site. They searched your brand. They were already in consideration.
When they convert, your remarketing campaign takes credit. But would they have converted anyway through a direct visit, an organic search, a sales email? In many cases, yes. The ad did not cause the conversion. It just happened to be visible when the conversion occurred.
This is the difference between correlation and causation, and most marketing dashboards are built to show correlation.
The Research Behind This
Byron Sharp's research across hundreds of brands is clear: growth comes from penetration, not loyalty. Brands grow by reaching new people — people who have never bought from you, who do not yet know you exist.
Your remarketing campaign reaches the same 3% of people who already know you. The other 97% — the ones who have never heard of your brand, who are actively searching for what you sell — are getting none of your budget because your remarkable ROAS campaign is hoovering it all up.
Your remarketing is not a growth engine. It is a rearview mirror with a nice ROAS sticker on it.
The Question to Ask
Next time you are reviewing campaign performance, ask one question: "What would have happened if we did not run this?"
If you cannot answer that honestly — if you cannot separate the conversions that the ad caused from the conversions that would have happened anyway — then you are not measuring. You are decorating.
What to Do Instead
This does not mean remarketing is useless. Targeted retargeting of specific high-intent audiences — people who visited your pricing page, who started a trial, who downloaded a specific resource — can drive real incremental value. The key word is incremental.
What does not work is blanket retargeting of everyone who ever visited your homepage, running indefinitely, claiming credit for every conversion that happens to occur within the attribution window.
Tighten your remarketing audiences. Shorten the attribution windows. Set frequency caps. And use the budget you save to reach people who have never heard of you. That is where growth lives.