Media & Performance
Remarketing
We bring back people who showed real interest, with messages suited to where they stopped, and we stop paying for sales that were coming anyway.
The problem
Remarketing looks like your most profitable campaign, and you can't tell how many of those sales it actually caused.
Remarketing reaches people already close to buying, so it nearly always looks brilliant in reports. Someone adds to cart, sees a banner, returns the next day, and the ad takes the credit. A high ROAS then pulls more budget toward people who might have converted regardless.
Customers feel the other side. A catch-all audience shows the same ad for weeks, even to people who already purchased.
Consent narrows things further. Visitors who decline cookies never enter your lists, so planning has to start from permissioned audience size.
You may recognize this
- Remarketing ROAS sits far above prospecting, and budget keeps drifting toward it.
- Customers who just bought still see ads for the same product.
- You run one audience, one message and a long membership duration.
- No frequency cap exists, and people say the ads follow them around.
How we work
Segment by intent and recency
Audiences reflect page depth, cart activity and days since the last visit, with membership durations set to your sales cycle.
Exclude with discipline
Recent buyers, open leads and current customers come out of the pool. GA4 audiences, Customer Match and Meta custom audiences stay synced with the CRM.
Sequence the message
Dynamic remarketing handles product recall through the feed. Separate creatives address hesitation at each stage, whether price, proof or fit.
Prove the lift
A holdout group or platform conversion lift study shows what the campaign adds. Budget and frequency caps follow from that.
What we deliver
- Audience architecture and membership duration plan
- Exclusion rules and CRM sync flow
- Message sequence with creative briefs
- Frequency cap rules by audience
- Holdout test design and readout
How we measure
- Incremental conversions: the gap between exposed and holdout groups.
- Frequency per person: reveals when repetition has become annoying.
- Click-through versus view-through share: a check on how much credit is earned.
- Reachable audience size: sets a realistic ceiling on spend.
An adult education provider
Illustrative scenario: A provider of adult certificate programs shows one discount ad to all visitors for months, including enrolled students. The team separates audiences by program page, application and enrollment status, removes students, and leaves part of the audience unserved as a holdout.
Common questions
What happens to remarketing when people reject cookies?
It shrinks. Those visitors never join your audiences, so Customer Match from permissioned lists and on-platform engagement audiences matter more. Consent mode modeling helps measurement but won't rebuild the audience.
Can we upload our customer list to ad platforms?
Technically yes, with hashed data. Legally, confirm with counsel that consent and cross-border transfer requirements under KVKK, Türkiye's data protection law, are met first.
What is the right frequency cap?
No universal number exists. We set a conservative starting cap, then watch how incremental conversions respond as frequency changes.
Before we work together, let's think together.
Requests are answered by the founder.