Most publisher inventory behind privacy restrictions just doesn’t monetize. Buyers can’t see who they’re bidding on, so they don’t bid. We wanted to know if Secure Signals actually fix that, or if it’s another config checkbox that doesn’t move the needle.
Context
The situation most publishers know too well: a user opts out of data collection or hits a restricted ad-serving environment, and the AdX auction goes quiet. Buyers lose the identity and interest signals they rely on. CPMs drop, fill rates crater, and a big chunk of impressions go unmonetized.
Google’s Secure Signals framework is supposed to address this. Publishers can pass encrypted, privacy-safe user signals into the AdX auction, giving buyers enough to bid on without exposing raw user data.
| We ran this study to answer a direct question: does enabling Secure Signals improve AdX bid CPMs, fill rates, and yield on privacy-restricted traffic? Or is the lift too small to bother with? |
What are Secure Signals?
Encrypted identifiers that publishers send to authorized AdX bidders during the auction. The buyer gets a privacy-safe signal that helps them recognize and value the user, but never sees the underlying data. Only the winning bidder can decrypt it.
In practice, these can be first-party identifiers from the publisher, hashed emails or authenticated tokens, or signals passed through Prebid, Google Tag, or a direct publisher integration.
The challenge
This isn’t a one-click setup. Several moving parts need to line up:
IMPLEMENTATION HURDLES
• Picking the right signal type for your audience and tech stack
• Deploying it correctly (publisher-side, Google-deployed, or via Prebid)
• Configuring GAM so the right bidders receive the signals
• Measuring impact with only GAM reporting, since there’s no built-in A/B framework for this
• Separating the signal effect from floor changes, demand mix shifts, or geo/device fluctuations
Any one of these can quietly undermine the exercise if it’s off.
Strategy
Benchmarking first
Before enabling anything, we locked down a baseline: AdX CPM, RPM, match rate, fill rate, and non-monetized rate. Each metric was cut by device, geo, and serving restriction status to control for traffic quality differences.
Enabling Secure Signals in GAM
Three things had to be true before we flipped the switch. Secure Signals needed to be enabled on non-personalized ad traffic, which is where the biggest impact sits. The right GAM integration had to be activated (web vs. app, depending on inventory type). And the deployment method had to be verified, whether publisher-deployed, Google-deployed, or Prebid-deployed.
Measuring the impact
We compared signal-present traffic against signal-absent traffic on the same KPIs, then sliced the analysis by signal delivery status, signal presence, and individual signal name to isolate what was actually moving the numbers.
Results
Traffic with Secure Signals: $0.34 RPM. Without: $0.03. Roughly an 11x gap.
CPMs went from $0.25 to $0.38 on signal-enabled impressions, a 52% increase. Fill rates moved from 13.62% to 88.38%, more than 6x higher. Non-monetized inventory dropped across the board on enriched traffic.
| ~11xRPM gap$0.34 vs. $0.03 | +52%CPM increase$0.38 vs. $0.25 | 6xFill rate jump 88.38% vs. 13.62% | ¯Non-monetized inventory down |
| Metric | With Signals | Without Signals | Delta |
| RPM | $0.34 | $0.03 | ~11x |
| CPM | $0.38 | $0.25 | +52% |
| Fill Rate | 88.38% | 13.62% | +6x |
| Non-Monetized Rate | — | — | Significant drop |
Revenue Opportunity :
A publisher with 100,000 daily impressions on privacy-restricted inventory faces a significant revenue gap without signals. At a 13% fill rate and $0.25 CPM, only 13,000 impressions clear, generating approximately $3,250 in daily revenue. The remaining 87,000 impressions go unsold.
With Secure Signals, the same inventory achieves an 88% fill rate at $0.38 CPM. This results in 88,000 impressions clearing and about $33,440 in daily revenue. In effect, 87% of previously unused inventory is converted into revenue-generating auctions.
The impact is clear: Secure Signals not only increases fill rate but also lifts CPM, driving an 11x improvement in RPM and closing the gap between unused and active inventory.
That’s $30,000 more per day on this single publisher’s restricted-traffic segment alone. Scale that across a month, 900,000 dollars. The setup work isn’t trivial, but neither is leaving $900,000 on the table monthly while competitors who’ve already enabled signals capture that revenue.
The real kicker is what happens to your inventory quality perception in the market. When buyers consistently win auctions on your restricted inventory, they get smarter about bidding on it. Your signal becomes a known quantity. Your fill rates stabilize. The CPM premiums stick around because buyers know what they’re bidding on. That’s compounding value.
Why Signal Strategy Is Now a Revenue Strategy
Most publishers treat privacy-restricted traffic as a write-off. Users usually opt out, environments restrict data collection, and the assumption is that those impressions just don’t monetize. The numbers in this study show that assumption is costing real money.
The reason signal-absent traffic goes dark isn’t mysterious. Buyers in a programmatic auction are making split-second decisions on whether a user is worth bidding on. Without any identity context, they can’t answer that question. So they don’t bid, or they bid the floor. That’s not a buyer problem, that’s an information problem. And information problems can be solved.
What Secure Signals does is give buyers enough to work with without exposing the fundamental data. This way the user stays protected. The buyer receives a signal they can decrypt and act on. The auction becomes competitive again. That’s the entire logic behind the 52% CPM lift and the jump from 13% to 88% fill rate. It isn’t demand that wasn’t there before, but demand that was always there but had nothing to bid on.
That distinction matters for how publishers think about their signal strategy going forward. The question isn’t whether to enable Secure Signals, the question is how much restricted traffic you’re currently sitting on, and how much of it is generating zero revenue because buyers are walking into a blind auction. Every impression in that bucket is recoverable. The infrastructure exists. The demand exists. The gap is configuration.
What this actually means
For publishers:
The fill rate move is the real story. Going from 13% to 88% on previously dark traffic means impressions that were generating zero revenue are now clearing in competitive auctions. That’s not a marginal optimization; it’s dead inventory coming back to life.
The 52% CPM lift confirms what you’d expect: when buyers can identify a user through an encrypted signal, they pay more. That increase isn’t from inflated floors or bid manipulation. It’s demand showing up because buyers finally have something to work with.
And the drop in non-monetized rate is just the flip side. Fewer wasted page views, tighter monetization.
For SSPs and DSPs:
SSPs that support signal passthrough and identity enrichment will be more competitive in AdX auctions going forward. Those that don’t are going to fall further behind, particularly on restricted inventory where signals matter most.
On the buy side, better signal quality gives DSPs more targeting confidence, which shows up as higher bids and better campaign performance. That loops back to publishers: more demand, better CPMs, more incentive to keep investing in signal infrastructure.
Bottom line
Signal-enabled traffic generated 11x more revenue per thousand page views, cleared at 6x the fill rate, and bid 52% higher on CPM. On privacy-restricted inventory, Secure Signals turned dead auctions into competitive ones. The setup takes work, but the payoff is obvious. If you haven’t enabled this, you’re losing revenue on every restricted impression you serve.
How DataBeat fits in
We help publishers figure out which signal configurations drive incremental revenue, not just which ones look good in a dashboard. That means benchmarking across signal types and traffic segments, isolating real impact from noise, and handling the full rollout from signal selection through GAM config to ongoing optimization.
If you’re trying to quantify how much you’re losing on restricted traffic, or you want to test whether a Prebid-deployed signal outperforms a publisher-deployed one, that’s the kind of thing we do. The goal: fill more impressions at better CPMs on the inventory that’s hardest to monetize.
