Data published by Piano in partnership with Digiday earlier this year found that publishers who lost search traffic did not lose revenue at the same rate. Search traffic fell 36%. Revenue fell 16%. Overall audience was down 9%.

That gap is not luck. It is the result of session-level monetization doing more work per visitor. And for gaming publishers, it is about to matter a great deal more.

Google expanded AI Mode to all US searchers in May 2025, following its I/O announcement, and made it the global default at I/O 2026. The content formats that AI Overviews handle best are exactly what gaming publishers produce: factual, informational, single-answer queries. Character guides, patch notes, tier lists, wiki-style content that ranks well in traditional search and answers in one snippet. The Piano data already reflects the early damage. Their projection for the next three years is that media leaders expect 43% more search traffic loss, with one-fifth expecting losses above 75%.

The traffic contraction is already here. The question is whether your ad setup is configured to maximize revenue from the sessions you still have.

The Shift From Pageview RPM to Session RPM

Publishers have historically optimized for pageviews, sessions, and RPM at the page level. Those metrics make sense when traffic growth is the primary lever, but when traffic is declining, the unit of optimization often comes down to the session.

Samuel Youn, VP of Programmatic at Chegg, argued in AdMonsters that publishers should optimize for "user value, not just pageview yield," and that a lighter ad experience reducing short-term revenue per session can increase total revenue by driving longer sessions and higher engagement over time. The logic applies directly here. If a visitor arrives from a search result that now competes with an AI answer, they still land on your site, but the path that used to produce 3 or 4 pageviews now produces 1 or 2. If each of those pages earns more, total session revenue can hold.

The practical change is in what you optimize against. If you are A/B testing ad placements, sticky vs. static, lazy load delays, or viewability thresholds, session RPM is the outcome metric. A placement change that lifts RPM on an individual page but reduces scroll depth and kills the second pageview is a net loss. Session RPM shows you that. Pageview RPM does not.

How Static Floor Prices Are Leaving Money on the Table

Most small and indie publishers set a floor price once during setup and do not revisit it. A single floor applied uniformly to all inventory ignores the fact that a 300x250 on your homepage on a Saturday afternoon in the US commands a meaningfully different market price than the same unit served at 2am to a mobile visitor in Southeast Asia. A fixed floor calibrated to the average leaves money on the table in high-demand situations and blocks fills in low-demand ones.

In two documented cases, AdExchanger found the following results from adaptive floor implementations: a sports media website saw a 76% RPM lift and a 36% revenue increase during a period when traffic was down 22% year-over-year, and a casual gaming site saw a 40% RPM lift after enabling adaptive flooring. The mechanism is straightforward: a bid below your floor is rejected, and the impression either goes unfilled or falls to a lower-priority demand source. Dynamic floors reduce both outcomes by calibrating the threshold to the point where fill rate and CPM together maximize session revenue, rather than holding a fixed floor that either blocks too much or accepts too little. Results vary by publisher, inventory mix, and baseline floor configuration.

Dynamic floor pricing adjusts floors in real time based on historical bid data, time of day, geography, device, and ad unit. You do not need to implement this manually. Most header bidding solutions and SSPs offer some form of floor optimization, either built into their wrapper or as a configurable feature. The implementation for a solo publisher is: turn on dynamic floors if your current setup supports it, set a minimum CPM floor per unit based on your 30-day average CPM for that unit, and audit the results monthly.

One thing to watch: floors set too aggressively reduce fill rates, which can suppress session RPM even if per-impression CPM rises. The target is not the highest possible floor. It is the floor that maximizes revenue per session across the full inventory. If fill rate drops more than 8-10% after a floor adjustment, check whether the CPM lift on filled impressions is offsetting the lost volume — if not, the floor is likely set too high for that unit and context.

Why More Competing Bidders Per Impression Drives More Revenue Than Traffic Volume

According to Digital Applied's 2026 programmatic advertising data, 87% of programmatic publishers now run header bidding, and among publishers using header bidding the average wrapper includes 14 SSP integrations. That average hides a wide spread. Publishers with fewer integrations are running auctions with less competition per impression, which suppresses CPMs regardless of how much traffic they have.

The same data puts blended open-exchange CPMs in the range of $5–6 across verticals, with private marketplace inventory averaging roughly 2x that, though gaming-specific CPMs vary significantly by ad unit, geo, and demand source. PMP deals typically require direct relationships with buyers, a minimum traffic threshold, and ongoing management. For publishers who cannot or do not want to manage PMP relationships directly, increasing SSP count in open bidding is the more accessible lever.

Running this audit is straightforward but requires per-impression data your reporting dashboard does not always surface by default. Look at your impression count by SSP, your average CPM by SSP, and your win rate by SSP over a 30-day period. Any SSP that wins fewer than 3-5% of auctions and consistently bids below your floor is not adding competitive pressure. It is adding latency. Removing it and replacing it with an SSP that has stronger demand for your specific vertical (gaming, entertainment) typically produces a net CPM lift without changing anything else about your setup.

The practical test is to run a 30-day period with your current SSP mix, then swap one underperformer for a new integration and run another 30 days. Session RPM is the comparison metric, not individual CPM, because SSP latency changes can affect ad load times and therefore total impressions served per session.

How Nitro Maps to AI Traffic Loss

Publishers managing this shift need two things from their ad stack: more competing demand without more configuration burden, and reporting granular enough to run the bid density audit described above. Nitro's header bidding setup addresses the first problem. Nitro handles header bidding setup and demand management on the publisher's behalf, running simultaneous auctions across multiple demand sources. For a solo or two-person team, this removes the primary operational barrier to competitive demand breadth.

The reporting side is where the revenue protection work actually gets done. Transparent reporting broken down by bidder, ad unit, and geography is the data layer that makes the bid density audit possible. Without it, publishers are optimizing blind: adjusting floors or SSP configurations without knowing which demand source is bidding on which inventory at what price. Nitro's reporting gives publishers the signal they need to make floor and demand decisions that are specific to their inventory rather than based on averages.

Liquipedia ran Nitro's header bidding setup and saw a 170% CPM increase, a 96% increase in programmatic revenue, and an 18% RPM increase. These results reflect what happens when you increase bid competition on inventory that was previously undermonetized. The RPM increase, specifically, is the metric that matters for publishers facing AI-driven traffic reduction: more revenue from the same sessions.

Nitro is dedicated to reinventing website monetization for the gaming industry. Our ad tech platform delivers uncompromised user experience alongside high performance revenue, with Net 7 payouts, same day support, and fully transparent real time reporting.