When ads cover the content, the sixteen ad experiences the Better Ads Standards name and the density lines behind them

If your ads are covering your content, two measurements decide whether it costs you anything beyond annoyance: how much of the screen the advertising occupies, and whether the reader can get rid of it. The Coalition for Better Ads publishes the list of experiences that fail on those grounds, and on 31 October 2026 the standards page at betterads.org/standards named six on desktop web and ten on mobile web. Three of the sixteen are density thresholds rather than formats, which is the part most layout arguments miss.

The named experiences, in the Coalition's own words

The page heads the list "Least preferred ad experiences". Desktop web, as read on 31 October 2026:

• "Pop-up Ads"
• "Auto-playing Video Ads with Sound"
• "Prestitial Ads with Countdown"
• "Large Sticky Ads"
• "Ad Density Higher than 50%"
• "Ad Density Higher Than 30% when Combined With a Sticky Video Ad"

Mobile web, same page, same date:

• "Pop-up Ads"
• "Prestitial Ads"
• "Ad Density Higher Than 30%"
• "Flashing Animated Ads"
• "Auto-playing Video Ads with Sound"
• "Postitial Ads with Countdown"
• "Full-screen Scrollover Ads"
• "Large Sticky Ads"
• "Sticky, Pop-Out Video Ads"
• "Sticky Video Ad with Large Inline Ad"

The same page states how the list was built. In its words, "Extensive consumer input and empirical data shaped the Better Ads Standards", and "More than 150,000 consumers have participated to date in the Coalition's research to develop its set of Better Ads Standards." That matters when somebody on your team argues the list is arbitrary. It is a research output with a published sample description, which is more than most layout opinions carry.

The density numbers are the ones that catch honest publishers

Nobody sets out to run a pop-up over their own article. Density is different. It accumulates. A header unit, a sticky footer, two in-article units and a right rail are each defensible on their own, and together they can push past the mobile 30 per cent line on a phone screen without anybody adding a single new placement.

Density is measured as the proportion of the vertical space taken by advertising, so the number moves with viewport height and with how much body text sits between units. Two pages carrying identical ad code can land on opposite sides of the line because one has long paragraphs and the other has a listicle of two-line items. Measure the pages you actually publish most of, not the longest article on the site.

Three practical consequences follow. A short post is the riskiest template you own, because the same ad load covers a bigger share of it. A sticky video unit changes the arithmetic twice over, since it appears in the desktop list at the lower 30 per cent threshold and again on mobile paired with a large inline ad. And infinite scroll feeds that inject a unit every few cards need a counted ratio rather than a cadence that felt right in a design review.

The three layouts that bury an article without naming a banned format

First, the collapsing hero. An ad slot above the headline that loads late pushes the article down after the reader has started, so the first thing they do is scroll back up. Reserve the space in CSS so the slot occupies its final height before the creative arrives.

Second, the double sticky. A sticky header and a sticky footer together leave a reading window that can be shorter than one paragraph on a small phone. Pick one, and let the other scroll.

Third, the interstitial on internal navigation. A full-screen unit between two articles on your own site is the postitial pattern the mobile list names, and it is usually added by a tag rather than by the design team, which is why it survives reviews. Audit what your ad stack injects, not only what your templates declare.

The order to fix a layout that is covering its own content

Work in the order that changes the reader's experience fastest for the least revenue risk. Remove anything undismissable first, because a unit a reader cannot close is the one that ends the session. Then reserve space for every slot, which costs no inventory at all and removes the shift that makes a page feel broken. Then count density on your three most published templates at a phone viewport and cut the lowest earning unit until you are inside the thresholds. Only then revisit formats, because format changes need creative and seller conversations and the first three steps do not.

Keep the measurement honest by taking it on a real device and a throttled connection. A layout that behaves on a desktop at a wide viewport is not evidence about the phone, which is where the lower 30 per cent threshold applies. Our note on why five speed tools give five answers for one page is worth reading alongside this, because ad units are usually the single biggest contributor to the numbers those tools disagree about.

What a machine reads when the advertising dominates the markup

There is a second reader that nobody checks. Crawlers and the fetchers behind assistant answers take your HTML and keep the text. When half the document is ad container markup, script payload and placeholder divs, the article still gets extracted, but the parts of it that rank and get quoted are decided by what sits near the top of the body. An ad stack that injects three units before the first paragraph pushes your actual answer down the document, and the text a machine quotes first is not the text a reader sees first.

The fix is the same one that helps the reader. Put the direct answer in the first paragraph of the body, above any injected slot, and keep the opening section free of units that load asynchronously. That is a layout decision with an editorial payoff, and it costs one placement.

Questions about ads that cover the content

Does Google penalise a page for ad density on its own

Ad density is not a documented ranking factor with a threshold attached. What it affects is measurable behaviour: layout shift, interaction delay and readers leaving. Those are the things to fix, and the Better Ads list is a useful shortlist of what readers reported as worst rather than a penalty schedule.

How do I measure ad density without building a tool

Take a full-page screenshot at a phone viewport, measure the vertical pixels occupied by advertising, and divide by the full page height. Do it on three templates. It is a ten-minute job per template and it gives you a number you can defend in a revenue meeting.

Is a sticky footer ad always a problem

No. A small sticky footer that can be dismissed is not the "Large Sticky Ads" entry on either list. Size and dismissability are what move it across the line, so specify both in the slot definition rather than leaving them to the creative.

What about advertising on an app rather than the web

The standards read on 31 October 2026 are written for desktop web and mobile web experiences. An in-app placement is governed by the store policy and the mediation partner's own rules, so take the density reasoning across and read the app rules separately.

Our ads pay for the newsroom. How much revenue does this cost

Run it as a test rather than a decision. Remove the lowest earning unit on one template for two weeks and read revenue per session rather than revenue per page view, because the sessions themselves get longer when the layout stops fighting the reader. Judge the trade on your own numbers, not on a benchmark from somebody else's site.

The reader you did not put a slot in front of

There is a version of your article nobody has designed for: the one an assistant reads before it answers somebody's question, with no advertising in it at all. The sentence it repeats is whichever part of your text it found most answer-shaped, and the layout above it decided where that text sits. AI Knows Us exists to show you which of your pages get used that way and what gets quoted out of them, which is a different report from the one your ad server produces. Keyword Planner reports 210 searches a month in the United States for this group of phrasings, read on 31 October 2026 against the United States geo target and taken as the group maximum rather than the sum of the phrasings, so the question is small and the layout decision behind it is not.