Retargeting campaigns have the highest ROAS in almost every account. That is also why they are the most misunderstood.
A high ROAS doesn’t mean the campaign is doing a lot of work. It usually means you showed one more ad to someone who was going to buy anyway.
The question you should ask
Not “what is my retargeting ROAS?”.
“Would these sales have happened without retargeting?”
A customer who adds to cart and comes back two hours later to buy would most likely have come back even if you had shown them no ads at all. But that sale is credited to the retargeting campaign.
That is why retargeting ROAS is inflated by nature.
The wrong structure everyone builds
The typical setup we see:
- One campaign
- One audience: “site visitors in the last 30 days”
- One ad: the product catalogue
- 35-50% of the budget
There are four problems at once.
Problem 1: The audience is too broad and mixed
“Site visitors” includes people who bounced off the homepage, people who read a blog post, existing customers checking a delivery, and people looking at the returns page.
Showing all of them a product ad burns budget.
The fix: split audiences by intent.
| Audience | Window | Message |
|---|---|---|
| Started checkout, didn’t finish | 3 days | Urgency / reminder |
| Added to cart, didn’t buy | 7 days | Objection handling (shipping, returns, warranty) |
| Viewed a product | 14 days | Product + social proof |
| Watched a video / engaged | 30 days | Introduction, brand |
| Existing customer | 180 days | Cross-sell, new products |
Problem 2: The window is too long
A single 30-day window means advertising to people whose intent to buy passed long ago.
The faster the purchase decision, the narrower the window should be. 30 days is long for a 400₺ product; it’s short for a 40,000₺ piece of furniture.
Problem 3: The same ad gets shown
The person has already seen the product and didn’t buy. Showing them the same product image again gives them no new information.
On the second contact you need to answer the objection:
- If they think it’s expensive: instalments, comparison, lifespan
- If they don’t trust you: reviews, returns guarantee, a user video
- If they’re undecided: size guide, FAQs
- If they simply forgot: a plain reminder
Problem 4: The budget share is too high
The retargeting audience is made of traffic you brought in earlier. If no new traffic arrives, that audience shrinks and runs out.
Giving half the budget to retargeting produces lovely numbers in the short term and stops growth in the medium term.
A healthy split: new customers 70-85%, retargeting 15-30%.
Set a frequency cap
The most common complaint about retargeting is that it becomes irritating. Someone who sees the same ad five times a day cools on the brand.
Set a frequency cap at ad set level: at most 3 impressions in 7 days is a good starting point.
How to measure it
To see retargeting’s real contribution, don’t look at total ROAS. Look at this:
Does account-wide contribution rise when you increase the retargeting budget?
If you raised the retargeting budget 50% and account-wide revenue didn’t change, that money simply reassigned existing sales to its own column.
The cleanest test: switch the retargeting campaigns off for a week. If total revenue doesn’t fall noticeably, those campaigns aren’t doing as much as you think.
Few brands dare run that test. Those who do usually change their budget split permanently.
Summary
- Split the audience by intent; don’t build one pool
- Narrow the window to the product’s decision time
- Answer the objection on the second contact; don’t re-show the product
- Give it at most a third of the budget
- Set a frequency cap
- Measure its contribution by switching it off