Audience Targeting

Optimized Targeting vs Topics Targeting: Model or Category

In short: Both decide where a Display or video ad group spends when you do not have a precise audience list. Topics targeting draws the line around content - any page or channel Google files under your chosen subject is eligible. Optimized targeting draws no line at all - it predicts which people are likely to convert and serves to them wherever they browse. One is static and auditable, the other adaptive and opaque. If you need to know and control where ads run, use topics; if you have clean conversion data and want volume, optimized targeting will find it.

By the AdFlint research team · Fact-checked against current Google and Meta platform behavior · Last reviewed July 2026

Optimized Targeting

Lets Google find additional converting users beyond the audiences you selected in a Display, Demand Gen, or Video campaign, treating your segments as a starting signal.

Enabled at the ad group level, it uses your chosen audiences and live conversion data as seeds, then bids on people outside those lists who look likely to convert. It is the intended successor to similar segments. It only behaves well when conversion tracking is reliable, and advertisers routinely confuse it with audience expansion, which widens reach without optimizing toward a conversion goal.

Full definition

Topics Targeting

Places Display or video ads on pages and channels about a chosen subject, targeting the content being consumed rather than the person consuming it.

You pick from Google's topic tree and your ads become eligible on any inventory Google classifies under that topic, at whatever depth you select. It is a fast way to get broad contextual coverage across the Display Network or YouTube. Because topics are wide by design, they pull in far more inventory than expected, so placement exclusions and a capped budget keep the waste visible.

Full definition

Side by side.

The differences that actually change what happens in your account.

	Optimized Targeting	Topics Targeting
What constrains delivery	The model's conversion prediction - nothing else except your exclusions.	Google's topic classification of the page or channel.
Follows the person or the page	The person - the same user is eligible on any site they visit.	The page - the same user is in scope on one article and out on the next.
Data required	A live, trustworthy conversion feed to learn from.	None - it works from day one with no tracking at all.
How it changes over time	Re-targets continuously as new conversions come in.	Static until you edit it - only the pages under the topic change.
Auditability	You cannot see who it chose or why - only aggregate results.	The placement report names every site and channel it bought.
How you rein it in	Turn it off, fix the conversion action, or add audience exclusions.	Go deeper in the topic tree and exclude offending placements.
What it costs you when wrong	Confident spend against a polluted conversion signal.	Broad junk inventory you must weed out placement by placement.

What actually separates them.

Topics decides eligibility per impression from the page, while optimized targeting decides it from the predicted person, so the two can disagree about the same auction.

Enabling optimized targeting in the same ad group loosens your topic from a hard boundary into a signal, so delivery is no longer guaranteed to stay under the topic.

Optimized targeting is unusable without conversion tracking, while topics is the tool you reach for precisely when there is no data yet.

A topics mistake is recoverable through placement exclusions the same day, while an optimized targeting mistake requires fixing the conversion action and waiting for the model to relearn.

Topics caps how wrong things can get at the edge of the category, while optimized targeting has no category edge - budget and exclusions are the only fences.

Which one should you use?

Use Optimized Targeting when

You have steady, deduplicated conversions and want more of them without listing audiences.
Your manually picked segments have saturated and volume has flattened.
You do not need to defend a placement list to a client or brand team.
You are consolidating after the similar segments sunset and need the modern equivalent.

Use Topics Targeting when

You are launching with zero conversion history and need a sane starting boundary.
Brand adjacency matters - ads must sit on content about the subject, full stop.
You need to explain and audit exactly what inventory the budget bought.
Conversion tracking is unreliable right now and you refuse to let a model learn from it.

Common questions.

Can I run topics and optimized targeting together?

You can, but understand what happens: the topic becomes a seed signal rather than a fence, and ads can serve on content outside it. If strict adjacency to the subject is the point of the campaign, keep a topics-only ad group with optimized targeting switched off. Use a separate ad group if you want to test the model's expansion.

Why is my topics campaign serving on pages that have nothing to do with my topic?

Three usual causes: the topic node is broader than its label suggests, Google's classifier filed a page under it that you would not have, or optimized targeting is enabled and expanding beyond the topic. Check the ad group's targeting setting first, then read the placement report and exclude the offenders. Moving to a deeper subtopic cuts the breadth problem at the source.

My optimized targeting ad group shows great conversions but sales are flat. What happened?

The model optimizes toward whatever the conversion action counts, not toward revenue. If the action includes soft events, duplicates, or a value that never changes, it will happily scale an audience that produces those numbers and nothing else. Audit what the conversion action fires on before blaming the targeting.

Does optimized targeting replace topics targeting?

No - they answer different questions. Topics controls what content the ad sits beside; optimized targeting chases who is likely to convert. Accounts that need contextual control keep topics with the toggle off, and accounts chasing conversion volume lean on optimized targeting, often both within the same campaign in separate ad groups.

Or stop choosing between them.

AdFlint picks the setting, writes the ads, and keeps optimizing inside the Google and Meta accounts you already own.

Try AdFlint free See how it works

Related comparisons

All comparisons