Programmatic Advertising: Is Something Wrong with Google's Audience Segments? 🕵️

Why are Google audience segments so large, yet often so effective? Let's take a closer look at a curious pattern in Display & Video 360 (DV360).

In the French programmatic advertising market, DV360 is a major DSP. Two of its most compelling advantages are:

• Access to YouTube inventory through Google's buying platforms, including DV360 and Google Ads.
• The ability to activate Google's own audience and contextual segments without a separate third-party data fee.

Google's audience offering is particularly attractive to marketers because of the breadth of its ecosystem:

🔎 Google Search
📊 Google Analytics
🌐 Chrome
📱 Android
📧 Gmail
🎬 YouTube
🗄️ Google Ad Manager (GAM)
👤 Google Accounts

These products represent a vast ecosystem of potential signals, although that does not mean every product's data is used directly to build every advertising segment.

And then there's the price: Google's own audience segments can be activated without the extra data charges typically associated with third-party audience marketplaces. Depending on the campaign's media and data costs, that can make a significant difference to the total cost of targeting. 💲💲

But behind this attractive proposition, some questions remain unanswered.

How Does Google Build Its DV360 Audience Segments?

From an advertiser's perspective, the exact composition of many Google audience segments is not fully transparent. What signals are used? How are users classified? How broad are the underlying models?

Like the Coca-Cola recipe, some of the formula remains secret. 🤫

And there's something particularly intriguing about their scale.

They're big.
Very big.
Really, REALLY BIG. 👀

How Large Are Google's In-Market Audiences in France?

Here's an example from our own analysis of the DV360 planner.

With the filters set to France, Google Ad Manager (GAM) supply, and web/mobile web inventory, the planner showed an estimated universe of 40.8 million cookies.

Across 37 top-level Google audience categories examined in the planner — including categories such as Consumer Electronics — the average estimated segment size was approximately 35 million cookies.

That's roughly 86% of the 40.8 million-cookie reference universe for an average category. 😲

Some examples from the same analysis:

• Home & Garden: 94%
• Financial Services: 94%
• Business & Industrial Products: 83%

These are strikingly broad audience estimates.

One important caveat: these figures are planner estimates, not a verified count of unique French people. Cookies can overlap across segments, and estimated reach does not necessarily equal the number of individually identifiable users.

Still, the question is worth asking: how can such broad segments remain useful for targeting?

Why Can Broad Google Audiences Still Perform Well?

Here's the surprising part.

In our experience, when a travel advertiser activates Google's Travel-related interest or intent segment, that segment often performs better than alternative audience selections — despite its enormous estimated size. 🤯

This is an observation from campaign practice, not proof that the same outcome occurs across all advertisers, objectives or markets.

But it raises a fascinating question: has Google found a way to combine massive reach with meaningful targeting efficiency? 🧙

Could Google Audience Segments Behave Like Giant Lookalike Models?

We have a theory. 👬

What if the audience segments visible in DV360 acted less like fixed lists of users and more like very broad modeled populations?

And what if Google's delivery systems could prioritize more or less relevant subsets within those populations, depending on campaign objectives, available signals and bidding conditions?

In that scenario, a segment displayed as extremely large in the planner would not necessarily mean that every eligible user receives the same priority in the auction.

Instead, automated bidding and delivery could potentially balance audience relevance, available reach and expected performance. ⚖️

To be clear: this is our hypothesis, not a confirmed description of Google's proprietary audience-building systems. Audience modeling, segment eligibility and bidding optimization are distinct mechanisms, and the planner figures alone cannot establish how they interact.

But the underlying strategic idea is compelling: perhaps the real competitive advantage is not simply having the most precise audience definition, but knowing how to balance scale and relevance dynamically.

That could give other AdTech players a few ideas. 😉

What Does This Mean for Programmatic Media Buyers?

For programmatic traders, the lesson is to question the relationship between audience size, targeting precision and actual campaign outcomes.

A huge estimated audience is not necessarily a bad audience. And a narrowly defined audience is not automatically a high-performing one.

Rather than judging targeting solely by the size of a segment, buyers should evaluate incremental reach, campaign performance, supply quality, media costs and the transparency of the underlying methodology.

How GreyFox Ads Approaches Programmatic Efficiency

At GreyFox Ads, we're a French programmatic advertising company specializing in Curated Master Deals designed around real media efficiency.

Our approach combines:

📡 Supply Path Optimization (SPO): prioritizing direct publisher supply paths.
🚫 MFA prevention: excluding made-for-advertising websites and subdomains.
📐 Screen-share optimization: favoring display formats and video players that occupy a meaningful portion of the screen.
🎯 Addressability: combining cookie-based, cookieless and privacy-compliant approaches for environments where consent or identifiers are unavailable.
🛡️ Brand Safety: applying safeguards against unsuitable environments.
🪟 Fee Transparency: making the programmatic cost chain easier to understand.

Because in programmatic advertising, the goal isn't simply to reach the largest possible audience. It's to deliver effective advertising through high-quality, transparent supply paths.

Looking for a Programmatic Advertising Partner?

Whether you're an agency, trading desk or advertiser working on Display and Video campaigns, we'd be happy to discuss your programmatic challenges.

📩 Get in touch with GreyFox Ads.

Original French LinkedIn post: https://lnkd.in/eWgzkZV4