Ad count looks fine; what about the reader experience?

For years, publishers have been asked to make a difficult trade-off: earn more from each visit while still giving readers a reason to return. In this calculation, ad load is often reduced to a count of placements – but two pages with the same number of ads can impose very different costs on the person reading them. One may be unobtrusive and quick; the other may fill the screen and drain a device’s resources.

Chrome’s new ad metrics make that distinction harder to ignore. The Chrome User Experience Report (CrUX) now publishes four experimental measures drawn from eligible real-user visits: the average number of ads visible at once, the share of the viewport they occupy, and the CPU time and network data consumed by ads. Publishers can inspect eligible sites through public tools, and so can advertisers and agencies. Ad experience is becoming a shared data point rather than a private judgement.

These measures give us more information to work with, but they still need context. Chrome has set no recommended thresholds or penalties, and the measures are not part of Core Web Vitals. The data covers qualifying Chrome visits, not every browser or app. Origin-level results can conceal large differences between an article page, a homepage and a live blog. The metrics are experimental and may evolve. Anyone turning one number into a universal ‘quality score’ today would be getting ahead of the evidence.

In a comparison in Assertive Yield’s publisher playbook, 3,743 publisher domains carried the same measured ad count as a cohort of the top 1,000 domains by programmatic supply share. Yet the publisher group showed two percentage points less ad density, 25% less ad-related network weight and 21% less ad-related CPU time. Those are cohort comparisons, not proof that one setup caused better commercial outcomes. But they illustrate the gap between how many ads a reader sees and what those ads ask of their device and attention.

An ad’s economic contribution has to be weighed against what it asks of the page, the device and the person reading it. A large sticky placement may attract attention but obscure content. A lightweight unit might add value without noticeably interrupting a session. A third-party creative can consume resources even when a publisher’s own page code is lean. A publisher may well be able to sell any one of these, but what they need to ask themselves is whether the revenue it generates outweighs the cost it imposes on the experience – and, ultimately, on the lifetime value of the user.

For publishers, that starts with establishing a baseline. Look at the four measures across eligible domains and pages, then put them alongside page type, device, scroll depth, time on site and advertising setup. Bring in revenue per visit, viewability, engagement and page performance, alongside the measures that contribute to user lifetime value. A high ad count might reflect a long, engaged session rather than a cluttered screen; a low count can still conceal heavy code.

The next step is to test the decisions behind them. Does a different refresh rule create meaningful incremental revenue? Is a video placement worth its network and CPU cost? Could lazy loading stop ads fetching before a reader gets near them? Do some demand partners add more resource weight than the value they return? We shouldn’t seek to remove ads indiscriminately, but we do want to find out which ones are earning their place.

Advertisers have a role in this conversation too. Public ad metrics may offer another way to investigate the environment in which a campaign appears. But they do not tell a buyer whether an audience is relevant, a placement is viewable or a campaign is effective. Nor should a single origin-wide figure determine the value of every impression on a site. Used carefully, the data can sharpen questions for publishers and supply partners; used carelessly, it could reward superficial compliance over genuine quality.

The wider opportunity is to change what the market rewards. For too long, volume has been easier to count than the cost of delivering it. Shared measures make it possible to discuss that cost more openly, while publishers’ own revenue and audience data provide the missing context.

An industry-wide race to run fewer ads is not the solution. The goal is to understand when each ad contributes enough to justify what it asks of the reader – today and over the lifetime of that relationship.

Leela Fallah is VP of customer success at Assertive Yield