Dynamic ads run themselves. The feed, the audience, and the exclusions behind them still need someone watching.

Catalog ads remove much of the manual work of building a thousand ads. The trouble starts when a stale feed shows sold-out products, a missing exclusion retargets a recent buyer, or a pixel mismatch breaks the personalization. We verify the feed, tracking, and exclusions before relying on the automation. You end up with a catalog advertising setup with scheduled checks for feed health, audience exclusions, and product-set logic before automated delivery is trusted. For ecommerce or marketplace businesses running (or about to run) catalog and dynamic product ads on paid social.

Some of the 500+ brands we've worked with

See all references

AI runs continuous feed and match-rate checks. The catalog owner and strategist decide catalog scope, exclusion policy, and when an automated bid strategy can be trusted. AI does not change feed or exclusion rules.

How we hold ourselves to it

A catalog ad is only as good as the feed and pixel behind it
Recent buyers stay out of retargeting
Each product set carries a targeting or merchandising decision
AI flags feed and match-rate problems. A human approves catalog scope

Audit the feed and the tracking behind it
Check the product feed for accuracy, freshness, and required fields, and confirm pixel or Conversions API events are firing and matching catalog item IDs. The catalog owner confirms which flagged issues get fixed before launch.
Feed and tracking health record covering accuracy, freshness, required fields and pixel or Conversions API match against catalog item IDs.
Build product sets on purpose
Group products into sets (best sellers, category, sale items) that carry a real targeting or merchandising decision, and document the rationale behind each one. Ecommerce or catalog owner approves the product-set logic.
Product sets with the targeting or merchandising decision behind each one documented.
Set exclusion rules for recent buyers and edge cases
Decide who gets excluded from seeing ads for a product they just bought, and how out-of-stock or discontinued items get pulled from rotation automatically. The strategist approves every exclusion rule before it's built.
Exclusion rules covering recent buyers and the automatic removal of out-of-stock or discontinued items.
Build and validate the campaign, paused
Configure the catalog campaign (audience, product sets, exclusions, creative template), while paused, and verify it against the feed and rule set before it can spend. A second reviewer signs off on the paused build before launch.
Paused-build QA record verifying audience, product sets, exclusions and creative template against the rule set.
Let it run, and keep watching the feed
Monitor feed health and match rate continuously alongside performance, since a personalization engine can look fine on the surface while quietly serving stale data. An authorized media owner approves any change to product sets, exclusions, or budget.
Continuous feed-health and match-rate monitoring recorded alongside delivery.
Review catalog health and campaign performance together
Check whether feed quality, exclusions, and product-set logic still hold, and whether the campaign's results reflect real catalog performance or a data problem. Catalog owner and strategist approve any change to scope or rules.
Joint review separating a real catalog result from a data problem, with the decision recorded.

Before anyone trusts a catalog campaign's numbers, there's a record showing the feed and tracking behind it were actually checked.

Feed and tracking health record
Feed accuracy, freshness, and pixel or Conversions API match rate, checked against catalog item IDs.
Product-set and exclusion register
Every product set's logic and the exclusion rules attached to it, including recent-buyer and out-of-stock handling.
Paused-build QA record
Evidence the live campaign matches the approved feed, sets, and exclusions before spend started.
Catalog and campaign decision log
What changed in the feed, sets, or exclusions, why, and how performance responded.

The automation is only as honest as the feed and the exclusions behind it. Catalog ads can look effortless from the outside. Upload a feed, connect a pixel, and let the system personalize. Yet a stale price, a missing exclusion, or a broken event match rarely causes an obvious failure. The campaign keeps showing the wrong product or retargets someone who already checked out.

A good fit when

Your catalog has too many products for manual ad builds, but the feed still needs a repeatable way to keep prices and stock current.
Pixel or Conversions API events are firing, yet nobody can tell whether their item IDs still match the products in the catalog.
Feed errors keep reaching campaigns without a clear response, so sold-out items, old prices, or broken links can continue serving after the source data changes.
Stale prices, sold-out items, or broken links remain unchecked in the feed.
Recent buyers keep seeing the exact product they purchased.
Catalog item IDs do not reliably match pixel or Conversions API events.
Product sets still reflect an older catalog.

A product feed with stale items and a mismatched pixel event feeding a personalization engine that can't see the gap

Better handled as other work when

A few manual ads already cover the catalog, so product-set automation would add upkeep without solving a real scale problem.
No working pixel or event tracking connects product views and purchases to catalog item IDs, so personalization would run on missing signals.
Product data changes faster than the feed can be trusted, and stale prices or discontinued items would keep reaching live ads.

Paid Search, Paid Social, CRO, and Programmatic each run under a named owner at Zeo. The consultants below are matched to the channel this page is about, so you can see who you'd actually work with.

DataFeedWatch
keeps the feed itself accurate in real time, which is what the automation is actually trusting

Start with the feed, match rate, recent-buyer exclusions, and product sets. We'll identify where weak data may be shaping delivery and performance.

Check your catalog setup

Do you build the product feed for us?

We audit the feed against the platform's needs, including accuracy, freshness, and required fields. The underlying feed infrastructure usually sits with an ecommerce platform or developer. If building or maintaining it falls outside scope, we'll name that explicitly. You still get a record of what needs fixing before catalog ads rely on it.

How is this different from just turning on catalog sales in Ads Manager?

This method adds a feed audit, exclusion rules, product-set structure, and ongoing health checks to the one-click Ads Manager setup.

Can AI manage the catalog campaign on its own?

Product-set logic, exclusion rules, and every change to scope require approval from a catalog owner and strategist. AI may run feed and match-rate checks and flag problems. It does not manage the campaign on its own.

What happens when the feed breaks mid-campaign?

A feed error or falling match rate triggers a flag. An authorized owner then decides whether the affected product set needs to pause while the feed is fixed.

By team

By industry

By tool

Catalog & Dynamic Product Ads

Automation starts with feed health

Audit the feed and the tracking behind it

Build product sets on purpose

Set exclusion rules for recent buyers and edge cases

Build and validate the campaign, paused

Let it run, and keep watching the feed

Review catalog health and campaign performance together

A record behind every automated campaign

Feed and tracking health record

Product-set and exclusion register

Paused-build QA record

Catalog and campaign decision log