Now Reading: Google Ads Audience Signals Explained: Signals vs Targeting
-
01
Google Ads Audience Signals Explained: Signals vs Targeting
Audience signals are hints, not commands. Here's the simple version of how they work, and why PMax and Demand Gen treat them so differently.
Audience signals are hints, not orders. You tell Google's AI who your best customers tend to look like. The algorithm takes that hint and goes looking for more people like them, even ones outside your original list.
Think of it like giving someone directions to your house. You mention the neighborhood and a landmark. They still find their own way there.
That's the shift worth understanding in 2026. Keywords used to be the main lever. Now, audience signals are just as important, sometimes more.
01Audience Signals vs Audience Targeting
These two terms get mixed up constantly. They mean very different things.
Targeting is a wall. "Only show my ads to these people." Nobody outside that group sees your ad.
Signals are a nudge. "Here's who I think converts best." Google can still show ads to people outside that group if it finds a good reason to.
Mind Over Maze's guide sums it up in one line worth remembering: targeting locks you in, signals give the AI room to explore.
02The Simple Version
In a Demand Gen campaign with "optimized targeting" turned off, audience signals behave more like real targeting. Search Engine Land's reporting confirms this: turn that setting off, and your ads only show to the audience you picked, even if it underperforms.
That's the opposite of how Performance Max works. In PMax, there's no way to turn signals into a hard wall. The AI always keeps some freedom to look elsewhere.
03The Signal Types You Can Actually Use
ClickedOn's 2026 guide breaks the main options down clearly. Each one fits a different stage of the funnel.
- Affinity segments — long-term interests, good for top-of-funnel awareness.
- In-Market segments — people actively researching a purchase right now.
- Custom segments — built from keywords, URLs, or apps you specify.
- Customer Match — your own first-party customer data, uploaded directly.
- Detailed demographics — age, income, parental status, and similar filters.
- Life events — recent moves, graduations, marriages, and other big life changes.
First Pier's data found In-Market segments deliver a 10% CTR lift over Affinity segments in direct comparisons. That's a useful default for mid-funnel campaigns.
04How Signals Work Differently in PMax vs Demand Gen
This is the part that trips up a lot of advertisers. Same word, different behavior.
| Behavior | Performance Max | Demand Gen |
|---|---|---|
| Can signals become a hard limit? | No, never | Yes, with optimized targeting off |
| Applied at what level? | Asset group | Ad group / campaign |
| Lookalike-style expansion? | Built in automatically | Lookalike Segments, opt-in |
| Reporting detail | Limited channel visibility | Placement and segment-level detail |
Groas' comparison guide puts it simply: PMax groups assets and lets Google assemble them. Demand Gen lets you build distinct creative for distinct audiences within the same campaign.
05What Changed in March 2026
Demand Gen's Lookalike Segments got a quiet but important update. The "reach" slider used to be a hard constraint.
Now it's just a signal strength control. ALM Corp's coverage confirms Google added a "signal" tag next to Lookalike segments in the Audience Reporting table, marking exactly which ones now work this new way.
ALM Corp's report is specific: this change applies only to Lookalike Segments inside Demand Gen campaigns. Video campaigns, Display & Video 360, Search, Shopping, and Performance Max are unaffected.
06Getting Signals Right the First Time
Sarah Stemen's 2026 guide reframes the whole job simply: you're not targeting users anymore, you're training the algorithm.
Start with first-party data. Customer lists and site visitors give Google the cleanest starting signal.
Don't overload one asset group. Multiple signals confuse the AI rather than sharpening it.
Give it two full weeks. Google's own documentation says signals take up to 14 days to fully integrate.
Check new lists after 24-72 hours. That's the window before audience size populates accurately.
Require click attribution for conversions where possible, this removes noisy view-through data.
07Are Your Audience Signals Set Up Well?
Answer a few quick questions to check your current setup.
Are Your Audience Signals Set Up Well?
Select the option that matches your account
08Common Questions
No. Targeting restricts who sees your ad. Signals suggest who to prioritize, but Google can still show ads to people outside that group.
No. PMax always keeps some freedom to expand beyond your signal if it finds conversions elsewhere. Only Demand Gen lets you lock this down.
The reach slider stopped being a hard limit and became a signal strength control instead. Google now marks affected segments with a "signal" tag.
Up to two weeks for Google's models to fully integrate it. New audience lists take 24 to 72 hours just to populate accurately.
Keep it focused. Too many overlapping signals in one group makes it harder for the algorithm to identify a clear pattern.
What We Learn Today
Audience signals are suggestions, not hard rules
PMax can never be locked to a signal, only Demand Gen can
In-Market segments beat Affinity by 10% CTR in tests
March 2026 turned Demand Gen's reach slider into a soft signal
New signals take up to 2 weeks to fully train
First-party data remains the cleanest starting signal
Build a Complete Google Ads Targeting Strategy
Audience signals work best alongside Demand Gen and value-based bidding. Explore both guides next.
