The Segmentation Audit: Finding the Leaks in a Retargeting Account Export
Objective: Given a real retargeting ad-set export, apply the lesson's segmentation, exclusion, and frequency rules to find which mistakes are quietly burning budget.
You're auditing the retargeting account for Chewy's cart-abandonment campaign after cost-per-acquisition crept up 40% over the last quarter with no obvious cause.
Review the ad-set export against the lesson's Common Mistakes checklist: exclusions, segmentation depth, and frequency caps.
Before you start
What you'll need
Free path (everything below is enough to finish)
Free, no account access needed, works from a CSV export alone
The process
3 steps
Step 01 of 03
The lesson's Common Mistakes section flags showing a 'buy now' ad to someone who bought yesterday as the single biggest waste in retargeting, and requires a recent-purchasers exclusion on every campaign.
Four ad sets are running. Which ones are missing a purchaser exclusion, and what is that costing?
Procedure
- Import ad-set-export.csv and freeze row 1
- Filter the 'Exclusion audiences' column for blank cells
- Cross-reference blank rows against the 'Spend (7d)' and 'Purchases from ad (7d)' columns to estimate wasted spend
Ad Set Audience Exclusion Audience Spend (7d) Purchases from ad (7d) Cart Abandoners Cart 7-day Purchasers 30d $412 3 All Visitors Site 30-day (blank) $598 11 Product Viewers Product 14-day (blank) $301 6
Healthy
Every ad set targeting past visitors carries a 'Purchasers 30-90d' exclusion; purchases attributed to the ad after a same-user purchase are near zero.
Unhealthy
Two of three ad sets have a blank exclusion column, and 17 of the week's attributed 'purchases' are actually repeat buyers who already converted before the ad ran.
What this means
A missing exclusion doesn't just waste impressions, it inflates the campaign's reported conversions with people who were going to buy anyway, making the whole campaign look more efficient than it is.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Ad set has no value in the Exclusion Audience column | Build a 'Recent Purchasers (30-90 days)' audience and add it as an exclusion to every retargeting ad set | 30 min |
Step 02 of 03
The lesson's Three Core Audience Segments splits visitors into homepage/blog visitors (low intent), product viewers (medium intent), and cart abandoners (high intent) — each needs its own message.
The export shows one ad set called 'All Visitors' running a single generic ad to everyone. What is that costing relative to segmented ad sets?
Procedure
- Sort the export by CTR and CPA
- Compare the single 'All Visitors' ad set's CTR/CPA against 'Cart Abandoners' and 'Product Viewers'
- Note the audience size and creative angle used for the merged ad set
Ad Set CTR CPA Creative Angle All Visitors 0.6% $58 Generic 20%-off banner Cart Abandoners 2.1% $19 Urgency: 'Still in your cart' Product Viewers 1.3% $31 Social proof: star ratings
Healthy
Each segment's CTR and CPA reflect its intent level — cart abandoners convert cheapest, homepage visitors are the most expensive segment to convert and use a trust-building message, not a discount.
Unhealthy
One merged 'All Visitors' ad set shows a CPA 3x higher than the cart-abandoner segment because a low-intent browser and a near-buyer are seeing the identical discount ad.
What this means
Blending intent levels into one audience forces one message to serve people in completely different mental states, which drags average performance toward the worst-performing segment.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| One ad set covers 'All Visitors' with a single generic creative | Split into at minimum 3 ad sets (browsers, product viewers, cart abandoners) with distinct creative angles | half day |
Step 03 of 03
The lesson caps frequency at 5-10 impressions per person per week; beyond that, users stop noticing the ad and start resenting the brand ('ad stalking').
The export includes a 'Frequency (7d)' column. Which ad set is at risk of ad fatigue, and what should change?
Procedure
- Sort by the Frequency (7d) column descending
- Flag any ad set above 10 impressions per person for the week
- Check whether that ad set has a frequency cap set in its settings
Ad Set Frequency (7d) Frequency Cap Set? All Visitors 18.4 No Cart Abandoners 6.2 Yes (8/week) Product Viewers 7.9 Yes (10/week)
Healthy
Every ad set has a frequency cap between 5-10 impressions/week, and actual delivered frequency stays inside that cap.
Unhealthy
The 'All Visitors' ad set shows 18.4 impressions per person in one week with no cap set, well past the point where the same message stops converting and starts annoying people.
What this means
An uncapped ad set spends the most impressions on the people least likely to respond to another repeat, which is the fastest way to burn budget without moving CPA.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Frequency (7d) exceeds 10 and no cap is set | Set a frequency cap of 5-10 impressions per week in the ad set's delivery settings | 5 min |
Final deliverable
A one-page account audit flagging every mistake found (missing exclusions, single generic audience, uncapped frequency) with a prioritized fix list ordered by estimated impact.
See a reference example
Rent the Runway, Retargeting Account Audit (excerpt) FIX NOW 1. 'All Visitors' ad set has no purchaser exclusion — est. 22 wasted impressions/week on repeat buyers 2. 'All Visitors' frequency at 21.0/week, no cap set — add 8/week cap FIX THIS SPRINT 3. Split 'All Visitors' into browsers / product viewers / cart abandoners, each with its own creative angle
Success criteria
You're done when you can:
- Correctly identifies every ad set missing a purchaser exclusion
- Correctly flags the merged 'All Visitors' audience as under-segmented
- Correctly flags any ad set exceeding a 10/week frequency with no cap
- Prioritizes fixes by estimated spend impact, not just by list order