Begin-to-Render Is Coming to AdSense: What Publishers Lose

Key Points

• Google moves AdSense and Ad Manager display impressions from count-on-download to begin-to-render on February 17, 2027, according to reporting from strategicrevenue.com.
• Impressions that never render will stop counting, and Google itself says publishers "may notice a decrease in total display impressions."
• Reported impressions dropping does not mean revenue drops by the same amount, because the ads being removed were largely unmonetizable anyway.
• Publishers should start benchmarking begin-to-render against current impression counts now, well before the transition date.
• RPS and session-level yield become the metrics that matter once impression volume stops being a reliable denominator.

What Google Is Changing

Google is narrowing the definition of a countable display impression in AdSense and Google Ad Manager. Strategic Revenue reports that the switch from count-on-download to begin-to-render takes effect February 17, 2027.

The mechanics are simple. Today, an impression can register the moment the creative starts downloading to the device. Under begin-to-render, the ad has to successfully load and start painting before it counts.

Google says the change aligns AdSense with measurement standards set by the IAB and the Media Rating Council. That is accurate. Begin-to-render has been the accepted desktop display standard for years, and most of the buy side has been transacting against it through third-party measurement for a long time.

Here is the practical difference:

Why Your Impression Count Is About to Drop

The impressions disappearing from your reports are the ones that were never worth much. A user who opens a page and leaves before the creative renders generated a download event and nothing else. No viewability, and nothing an advertiser would pay for.

Strategic Revenue makes the point directly: Google has confirmed publishers "may notice a decrease in total display impressions." The size of that decrease depends entirely on your traffic profile.

Sites with high bounce rates, slow page loads, or heavy mobile traffic on poor connections will see the biggest gaps. Fast pages with engaged sessions will barely notice.

Publishers running aggressive lazy-load configurations come out ahead here. Lazy loading already delays the ad request until the slot approaches the viewport, which means a larger share of your requests were already rendering. The gap between your old count and your new count shrinks accordingly.

Essential Background Reading:

• Ad Revenue: The baseline framework for how publisher ad revenue is generated and measured.
• Ad Monetization: Core concepts behind turning inventory and impressions into revenue.
• Ad Yield: How yield is calculated and why it matters more than raw impression volume.
• Target CPM Price Floor Strategy: Background on how CPM and price floors interact, relevant to the CPM shifts this change causes.

What Happens to Your Revenue

Revenue and reported impressions are two different numbers, and the source article is careful about the distinction. Strategic Revenue notes that fewer, higher-quality counted impressions could affect CPMs and advertiser economics differently than raw volume suggests.

Work through the math. If an impression never rendered, it was not viewable, which means it was not billable under most viewability-based deals and it dragged down your measured viewability rate. Removing it from the denominator raises your CPM and improves your viewability metrics at the same time.

Your revenue line should stay roughly flat while your impression count falls and your CPM rises. The actual result depends on which demand partners settle on Google's count versus their own measurement.

The risk here is operational. Publishers who benchmark against impression volume, compensate teams on impression targets, or forecast direct campaigns off historical impression counts will have a bad quarter explaining what happened.

Related Content:

• Ad Monetization Platform Comparison: See how different platforms handle measurement and reporting methodology differences like this one.
• Ad Monetization Platform Scorecard: A framework for evaluating whether your current platform reports revenue the way it should.
• Self-Service vs. Managed Service Ad Monetization Quiz: Determine whether you have the internal resources to manage a measurement transition like this one alone.
• Ad Revenue Resource Center: A broader library of resources on ad revenue strategy and monetization fundamentals.

What Publishers Should Do Before February 2027

Google does not require any changes ahead of the transition, but running a comparison now costs you nothing and prevents surprises later. Google recommends that Ad Manager publishers compare existing impression counts against begin-to-render metrics to estimate the decline.

Start with these steps:

• Pull the comparison report in Ad Manager: measure your current impression total against the begin-to-render metric across at least 30 days, segmented by device and by ad unit.
• Segment by traffic source: social and search referral traffic tends to bounce faster than direct or newsletter traffic, and the delta will show it.
• Audit page load performance: every second shaved off time-to-render converts a would-be lost impression into a counted one, so Core Web Vitals work now pays twice.
• Review lazy-load thresholds: loading too early wastes requests that never render, and loading too late costs you fill. Test your viewport offset rather than guessing.
• Reset internal forecasts and comp plans: if anyone on your team is measured on impression volume, move them to revenue or RPS before the switch.
• Brief your direct sales team: guaranteed campaigns sold on impression counts need their delivery assumptions updated ahead of 2027 IOs.

Next Steps:

• Revenue Per Session: A deeper look at why RPS is becoming the standard metric for publisher monetization.
• Revenue Per Session Resource Center: Additional guides on building an RPS-first reporting structure.
• Google Search Console for Publishers: How to use GSC alongside Google's ad products to protect revenue through platform changes.
• Ad Monetization Platform ROI Calculator: Model how a CPM increase and impression decline actually net out for your site.

Move Your Reporting to Session-Level Metrics

Impression counts have been a shaky business metric for a while, and this change makes that harder to ignore. Different measurement methodologies across SSPs, viewability vendors, and your ad server already produce discrepancies that no reconciliation process fully resolves.

RPS, revenue per session, sidesteps the problem. A session is a session regardless of how any vendor chooses to count a render event, which makes it the cleanest way to measure whether your monetization is improving.

Publishers who report on RPS will see the February 2027 transition as a footnote. Publishers who report on impression volume will spend a week explaining a chart to executives who assume the money went with it.

Render performance also becomes a revenue variable under this methodology. Ad slots that render fast count. Ad slots that hang behind a bloated tag chain do not.

See It In Action:

• Unbeatable Ad Revenue Case Study: A real publisher outcome from optimizing yield rather than chasing raw impression volume.
• Year-Over-Year Ad Revenue Increase Case Study: How sustained render and yield optimization compounds into long-term revenue growth.
• Letterboxd: Increasing App Revenue With Playwire: A session-level yield approach applied to a high-engagement publisher property.

How We Handle This

We built our stack around session-level yield rather than impression counts, so the begin-to-render transition touches our reporting inputs and leaves our optimization logic alone. Our yield teams already optimize against rendered, viewable inventory because that is what advertisers pay for.

Render speed is where the work sits. RAMP manages tag load order, timeout tuning, and lazy-load configuration across the full demand stack, so fewer requests die before they paint. Publishers running with us target viewability in the 70 to 90 percent range, and inventory that hits that range clears begin-to-render measurement without issue.

Our reporting gives you the log-level detail to see exactly which placements lose impressions to slow rendering and what that costs. No black box, no reconciliation guesswork.

The publishers who spend 2026 fixing render performance will walk into February 2027 with a smaller impression drop and a higher CPM. Talk to our team if you want to see where your inventory stands before the deadline arrives.