Ad viewability by format: which units actually get seen

A display impression is viewable when at least 50% of the ad's pixels are in the viewport for at least one continuous second. For video it is 50% of pixels for two continuous seconds. That is the IAB and MRC standard, and it is the bar your buyers measure you against.

Two things follow from that definition and both cost publishers money. An ad that loads below the fold and is never scrolled to is served but not viewable, so you gave away inventory for nothing. And an ad that renders after the reader has already scrolled past it is also not viewable, which makes page speed a revenue question rather than a technical one.

Ranked by how reliably they meet the threshold, and by what each one costs you elsewhere:

The pattern is that viewability tracks attention, not position. A unit placed where the reader is going anyway beats a unit placed where you wish they would look.

Mobile viewports are small, so fewer ads are on screen at once and each one depends entirely on the reader scrolling to it. A sidebar unit that performs on desktop often collapses or stacks at the bottom on mobile, where it is effectively invisible.

Mobile is also where render speed bites hardest. Slower connections mean the ad arrives later relative to the scroll, so the same slot in the same template measures worse. If your mobile viewability is far below desktop for identical placements, look at rendering before you look at layout.

The levers, in the order most publishers should try them:

• Lazy load below-the-fold units, with a trigger distance far enough ahead that the ad is ready when the slot arrives. Too close and you lose the impression entirely.
• Move units into the content flow rather than around it. Mid-article beats sidebar on almost every template.
• Cut density before adding units. Fewer, better-placed slots usually produce more viewable impressions than more slots.
• Fix render speed. Every hundred milliseconds of delay costs impressions on fast scrollers.
• Test sticky units against engagement, not just against revenue.

Notice that four of the five reduce your raw impression count. That is the trade being made on purpose: buyers pay against viewable inventory, so fewer impressions at a higher viewable rate is usually the better business.

This is the part that sends publishers redesigning the wrong thing. When a blocker uses cosmetic filtering, the ad request goes out, an ad comes back, and then a filter rule hides the container. Your ad server counts a served impression. Your verification vendor counts zero pixels on screen. The impression lands in your report as unviewable.

So a growing share of blocked traffic drags your viewability down while your layout has not changed at all. The instinct is to move units, change templates, add lazy loading, and none of it moves the number, because the cause is not where you are looking.

The tell is in the shape of the data. Layout problems vary by placement and template. Blocking varies by audience: by browser, by device, by traffic source. If the same unit in the same template performs very differently on Chrome and Firefox, or on desktop versus mobile beyond what render speed explains, you are looking at blocking.

The deeper cost is the one nobody sees. Where a blocker cancels the request instead of hiding the element, there is no impression at all, so it never shows up as bad viewability, it simply vanishes from your reporting. Fill rate and viewability explains why that loss is invisible in both metrics at once.

Work it in this order, because each step needs the previous one to mean anything.

1. Measure how many of your page views arrive with a blocker. Until you know that, every viewability figure describes only the traffic that got through.
2. Separate the two blocking modes: requests cancelled outright, and containers hidden after the ad loaded. The first is missing inventory, the second is depressed viewability.
3. Apply your own CPM to the recoverable share.
4. Then, and only then, judge whether your viewability problem is a layout problem.

Most publishers discover the ordering was backwards: they had been optimizing placements for months against a number that a fifth of their audience was silently distorting. What adblock costs publishers walks through putting a figure on it, and if you want to measure your own rate rather than estimate it, our script reports detections, recoveries and recovered revenue per site with a one-line install.