Yield Audit and Revenue Bridge for Ad-Funded Publishers

The answer first: In the audits I have run, most of the gap between what inventory earns and what it should earn sits in the demand path and in latency, not in floor prices. A Prebid timeout set at 2,000 ms is not buying more bids. Past roughly 1,000 ms on Mexican mobile connections the incremental bid rate flattens while first ad render, INP and viewability all degrade, and the net effect on session RPM is negative. Raising floors is the intervention most shops sell, because it is the one that shows a CPM increase in the interface while fill and revenue drop behind it. What I look at: Ad Manager Data Transfer files loaded into BigQuery rather than the GAM interface, because the interface cannot tell you which bidder wins only when another bidder times out. Prebid analytics at the auction level. Session RPM as the governing metric and page RPM as a diagnostic. Duplicate and abandoned ad requests. Unified pricing rules against the floors your SSPs actually receive. Open Bidding and client-side Prebid compared at the same floor for the same demand partner, which regularly shows one path eating the other. Refresh policy against measured Active View rather than against the interval you configured. Lazy-load thresholds against real scroll depth. ads.txt and supply chain gaps that quietly disqualify you from parts of DV360 and Trade Desk demand. The deliverable: A revenue bridge. One table decomposing the gap between current and achievable into named causes, each with a dollar figure per month, the evidence it rests on, and a stated confidence where the evidence is thin. Under it, a remediation list ranked by recoverable dollars per engineering day rather than by number of findings. Most audits I deliver contain three items worth doing and eleven that are true and not worth your sprint. I mark which is which. What I will not recommend: More ad slots. A sticky footer on every template. Outstream video dropped into the article body. Infinite scroll that manufactures pageviews. A 900-word article paginated across four screens. If ad density were the answer you would not have needed an audit, and every one of those moves borrows next year's revenue in a market where your direct advertisers read your site. Who should not buy this: If programmatic revenue is under roughly 20,000 USD a month, the recoverable amount will not repay this work, and I will say so on the first call rather than after the invoice. Sites whose traffic is majority paid acquisition have a different problem and belong in a policy and MFA review. And if you want a second opinion as ammunition in an internal argument that is already settled, I am the wrong person. I have handed clients a finding that said their previous vendor was right. Access I need: GAM at reporting level, Data Transfer enabled or the ability to enable it, your Prebid configuration as deployed rather than as documented, SSP dashboards, and one engineer for two conversations. Without Data Transfer this becomes a shallower piece of work, and I price it as one. After delivery: I walk your team through every finding, and if you want, your monetization vendor as well. I will defend the numbers in front of the vendor whose configuration I am criticising, and that conversation is included. Implementation is quoted separately, and you should ask your own team first, because several of these fixes are an afternoon of work for somebody who already has commit access.