What do the Targeting and Observation settings change for a placement plan?

What is the difference between targeting and observation in Google Ads?

Google's help page About Targeting and Observation settings (opens in a new tab) gives the short version: Targeting restricts the reach of your ad group, and the Observation setting does not restrict the reach of your campaign or ad group. With Targeting, the advertiser is telling Google Ads who to reach or where ads should show. With Observation, the advertiser monitors how ads perform for selected placements, topics or audiences while reach stays as it was.

The Google Ads API names the same switch more bluntly. Its TargetRestriction reference (opens in a new tab) has a field called bid_only. A value of true means the criteria will only apply bid modifiers, and not affect targeting. A value of false means they restrict targeting as well as applying bid modifiers. Observation is bid only; Targeting is bid and restrict.

Which targeting methods can use each setting?

The choice is not available everywhere. Google's settings page lists, for Display Network campaigns, display or video keywords, placements and topics as methods that can target, and placements and topics as methods that can be observed. It states plainly that display or video keywords cannot be used in Observation mode. For Search and Shopping campaigns, the page lists Observation only, for audiences.

The API's targeting settings guide (opens in a new tab) lists the dimensions that accept a restriction: audience, topic, gender, age range, placement, parental status and income range. It also notes that the Google Ads interface and the API can apply different defaults for some dimensions, so a setting created by a script or a bulk tool should be checked, not assumed.

How does the setting affect a placement list?

For a client who approved a list of sites, the placement dimension has to be set to Targeting. That is what turns the list into a condition: ads in the ad group may serve on the listed sites and nowhere else, subject to bids, eligibility and exclusions. Set to Observation, the same list fences nothing in. It becomes a set of rows that collect statistics and can carry bid adjustments, while the ad group's other targeting decides where ads run.

If nothing else in the ad group restricts reach, the result can be wide. Google's settings page says that for Display campaigns with no targeting method on an ad group, ads can run anywhere on the web within the campaign and account settings. The API describes bid only criteria as targeting all values for that dimension and serving on the targeting in other dimensions. Either way, an observed placement list is not a plan.

What happens when audiences and placements are layered?

Google's Display targeting page (opens in a new tab) states the general rule: each time you add a new layer of targeting, you narrow the potential reach of your ads, but adding items within a given targeting type broadens it. Placements, topics and keywords are one family; Google's page on combining content targeting (opens in a new tab) says ads show on content matching any of them. Audiences are a separate layer, and the setting on each layer decides whether it narrows.

Google adds one more detail on observed audiences: their data is reported only where it overlaps with the existing targeting. An observed audience on a strict site list therefore tells the planner who among the list's visitors fell into that segment, not how the segment behaved elsewhere.

When should observation be used?

Google recommends Targeting for all advertisers in their Display campaigns, and Observation for Search campaigns and for the Display campaigns of more advanced advertisers. For a planner running an approved site list, Observation has three honest uses, all on dimensions other than the list itself.

Observation also feeds automated bidding. Google's settings page says that for Smart Bidding campaigns, first party audience segments added under Observation are used as signals. That is a reason to add them deliberately, and to record them in the plan, rather than a reason to leave them off.

How do you check which setting an ad group uses?

Google's settings page says the Observation setting is applied from the Edit all targeting page, and that an observation method can also be added from a method's reporting page if it is not already used for targeting in that ad group. It also says an observation method cannot be added when an ad group is first created, which is one reason settings drift after launch: they are changed later, by someone else, on another screen.

Teams that manage accounts through the API can read the setting directly, since each ad group's targeting setting carries its target restrictions with a bid only value per dimension. The API guide warns that updating the restrictions as a full list removes any restriction left out of the payload, so a bulk script can switch a dimension back to its default without anyone touching the interface. Display campaigns are also moving to Demand Gen: in June 2026 Google began a phased rollout of a migration tool, so confirm the settings again in any migrated campaign. The Demand Gen migration guide covers that review.

What does a settings review look like on a real brief?

How does PlacementGraph support a placement plan that must stay on its list?