Banner blindness

Also called ad blindness. Also in Psychology

Quick answer

Banner blindness is the tendency of web users to overlook page elements that look like advertisements or sit where ads usually appear, even when those elements hold exactly what they are looking for. It was named in 1998, after test participants missed large, bright links. It shows up mostly when people are searching a page for one specific thing.

~20 sec read

011 min

The training-page test behind banner blindness

Jan Panero Benway and David Lane of Rice University were running a usability test of a corporate intranet. A usability test is a session where people try real tasks on a site while researchers watch. The participants were new to the web, and each had one job: find information about Internet courses.

The training page had a link made for exactly that job. It was large, it was marked as new, and it stood apart from everything around it. All participants found their way to that page without trouble. When participants reached the training page, almost every one of them scrolled past the large link and picked a small link from a list near the bottom.

The small link led to a list with nothing about Internet courses, so the participants gave up. The researchers then took them back to the page and showed them the large link. Most were very surprised that they had missed it.

A 1997 study by Spool and colleagues had described a similar moment: a participant needed information that was in an animated banner and did not seem to see it. Benway and Lane named the pattern banner blindness. The term dates from 1998.

training pageInternet Coursescourses
Look at where the participant goes: past the large link, to the small one. The large link was made to be seen first.

023 min

How banner blindness was first measured

A story from one usability test is not proof, so Benway and Lane built a small website to test it. The site had four levels, from a home page down to content pages. Each page above the content had a title, one short paragraph that did not matter for the task, and a menu of three to six text links.

Six working professionals with web experience did 24 short searches, such as finding a hotel's email address. For most searches, the text menus were enough. These were the control tasks. For four searches, the only way to finish was to click a red banner that offered a shortcut to the goal. The banners were large and bright, and they used the same words as the task.

In this pilot, participants found the banners about 58 percent of the time, while they solved 94 percent of the control tasks. Six people is a small group, so the result shows that the effect exists, not how large it is for everyone.

How the test tells a missed banner from an absent one

A person who misses a banner cannot report it, because they do not know it was there. So the researchers used success as the signal. Because only the banner tasks needed the banner, comparing success on banner tasks with success on control tasks showed whether participants had noticed the banners. Participants who missed a banner often concluded that the site did not hold the answer at all.

What finding the banner still cost

Some participants did find the banners in the end, especially when no other route existed. Even then, they rated those tasks as harder than the control tasks, and the tasks took longer.

The original studies used no eye tracker

Many guides describe Benway's work as an eye-tracking experiment. The paper reports no eye tracker. Its evidence is task success, the time people took to pick a category, and what they said afterwards they had seen. Eye tracking, which records where a person's eyes rest on a screen, entered the story later. Nielsen Norman Group, a usability research firm, repeated the finding in a large eye-tracking study in 2007 and found it again in a new study in 2018.

A second experiment with plain text items

In a second experiment, 72 undergraduate students searched a larger site with an 18-item menu. On some pages a large line of text, a small line of blue text or a graphic told them which menu cell held the answer. Only 17 of the 71 participants who answered said they had seen any of these items. A separate set of tasks showed ordinary advertisements, and only 20 percent of participants said they remembered seeing any ad at all.

Putting the banner on the same gray background as the menu did not help. A small amount of animation did not help either. The plain text items looked nothing like ads, and they were still missed. This is why the researchers used "banner" for anything meant to stand out and attract attention, not only for advertising.

The design guideline the results questioned

The guidance of the time said the opposite. Ameritech's web design guidelines, quoted in the Rice paper, held that the larger an item is, the more likely it is to attract attention. The Rice results suggested the advice could fail for people who are searching, because they skipped the large, bright items.

033 min

Why banner blindness happens

If the participants skipped a link made to be seen, what were they doing instead? They were searching. A person looking for one answer does not scan a page evenly. They look in the places where earlier pages taught them answers appear, such as menus, small text links and headlines.

Nielsen Norman Group describes the lesson people have learned: pay attention to navigation bars, search boxes and headlines, and skip areas that usually carry nothing useful. That is selective attention shaped by experience. It works as top-down processing: what a person expects decides where they look, before the page has shown them anything.

Users have learned to skip content that looks like an ad, sits close to ads, or appears where ads have traditionally been. That gives three signals that an area is probably an ad:

  • Where it is: the top of the page, or the column at the right edge, called the right rail.
  • How it looks: bright color, animation, or a box that differs from the content around it.
  • What is next to it: content placed close to ads is often treated as an ad.

Scanning habits add to this. On text-heavy pages, eye-tracking maps record fixations, the short moments when the eyes rest on one spot. They often show many fixations along the top and down the left side, a path called the F-pattern. Ad slots at the top and right sit away from that path. Guides also describe a Z-pattern and the Gutenberg diagram, which make a similar point, and say that content below the fold, the edge of the first screen, gets less attention.

Marketing guides name two more causes: crowded pages, and seeing the same ad formats again and again, which dulls the reaction to them. Both are plausible. Neither was tested in the 1998 experiments, so this article treats them as likely rather than shown.

How the eyes avoid a banner they have not looked at

Judging a banner as an ad before looking at it sounds impossible. Eye-tracking work suggests it is not. In an early eye-tracking study of web search, Drèze and Hussherr (2003) found that people looked directly at just under half of the banner ads shown. Ads were looked at less often than their size and position predicted. The authors concluded that people spot banners at the edge of their vision, where detail is poor, and choose not to look at them.

Why one ad lowers attention to the area around it

Once people decide an area is for ads, they stop going back to it. Nielsen Norman Group calls this the hot-potato effect: an ad can make people turn their eyes away from a region of the page and never come back to it. It has a second cost. Content placed in the same part of the screen as an ad is often taken for an ad too, and is also ignored.

041 min

Banner blindness on a travel website

That second cost is easy to miss, so a second case shows it. Nielsen Norman Group watched a participant use a travel website while looking for hikes in the Andes. The right rail held promotions, and also content about the very trip the participant was researching. They looked at the rail once, took its items for ads, and never looked there again.

Of the 148 fixations on the page, 132 fell in the content area, and only one of those landed in the right rail. The rail covered about a quarter of the content area, so it received about 33 times fewer fixations than its size would predict.

Compare this with the intranet test. There, the ignored item was a real link styled like a banner. Here, the ignored items were useful content that sat where ads usually go. The one difference is what was skipped, and it teaches something neither case shows alone: the cost does not stop at the ad. Whatever sits beside it can be skipped with it.

travel pagecontentrailfixationscontent: 132rail: 1
Compare the two bars: they share one scale. Almost every fixation fell in the content area.

052 min

Banner blindness in interfaces today

The same skipping appears on ordinary pages today. Nielsen Norman Group's 2018 eye-tracking sessions showed several forms of it.

  • On the Apartment Therapy website, the combined eye-tracking map of 26 participants showed the article text read heavily, while the ads in the top banner and the right rail drew little attention.
  • On the MedicineNet website, a participant researching cold symptoms swiped quickly past two ads. Their eyes passed over them without taking them in.

A promotion ignored in the middle of the text

Ads do not need to sit at the top or the side to be skipped. On the bacnyc.org website, most participants ignored a blue rectangular promotion placed after the first block of text, because it looked different from the text and photos around it. On a desktop screen, such a box is easy to ignore for that reason.

The first banner ad and what came after

The first banner ad on the web ran on HotWired in October 1994 and is reported to have started with a click-through rate of 44 percent. Click-through rate is the share of people who saw an ad and clicked it. Display ads today get far less. Marketing guides quote figures well under one percent but disagree on the exact value, so this article gives none.

Real content mistaken for an ad on mobile

Mobile screens make the problem harder in two ways. Less of the page is visible at once, so it is harder to see which elements differ from the main content. And a large inline ad can fill so much of the screen that people cannot look away from it as they scroll, even when they know it is an ad.

The reverse also happens. On the Generac website, a participant looking for a portable generator scrolled past a section that looked like an ad, because it differed from the white page and the images. The section held installation information they would likely have wanted. Ad-like styling alone can hide real content.

A 2023 study on matching ads to their page

A 2023 eye-movement study published in Cognitive Processing, led by Bo Ning, found that text ads placed inside pages dominated by pictures, with a theme matching the page, were the more effective choice against banner blindness.

062 min

Designing around banner blindness

If people skip what looks like an ad, designers can act on how a page looks, where its parts sit and how they are mixed. Say a product manager puts a new-feature announcement in the right column of a settings page, as a bright card with an illustration. People who came to change a password pass by it. The designer moves the announcement next to the setting it concerns, in the page's own type and spacing. Now it looks like part of the task.

Moves that follow from the research:

  • Do not give important content the look of an ad. Bright boxes, animation and unusual backgrounds teach readers to skip it. Keep to the page's own visual hierarchy.
  • Test whether people see what you put in the top banner or the right rail. Nielsen Norman Group recommends a usability test for this.
  • Keep ads and content in separate sections, since content next to an ad is often ignored with it.
  • Measure more than clicks. Click rate shows what people did. Heatmaps and recordings show where they looked, and viewability shows whether an ad was on screen at all.
  • Change one thing at a time in an A/B test, such as placement or format, so the result has one cause.

For people who buy or sell ad space, industry guides add a list of tactics: fewer ads, ads relevant to the page or the reader, native ads that match the style of the page, white space and a clear call to action, sticky or welcome-screen ads, and formats that are not rectangular boxes, such as video with a transparent background. They also suggest interactive and rich media formats, rewarded ads that trade attention for a benefit, and personalised or dynamically varied creative. The evidence for these is mixed. The Rice result that animation did not help is a warning that motion alone is not a fix.

These moves can backfire. Guides on native advertising warn that readers who feel tricked trust the page less, so ads that pass for content need a clear label, and deliberately exploiting the habit is an ethical choice, not only a design one. No format has ended the effect for good: the 2018 Nielsen Norman Group study found it still common on mobile and desktop.

071 min

How to spot banner blindness on your own pages

Banner blindness shows up in data and in test sessions:

  • A promotion with a low click rate on a page that has high traffic and strong engagement elsewhere.
  • Heatmaps or session recordings where visitors scroll straight past one region, often the top banner or the right column.
  • Usability sessions where people choose a small text link although a larger, brighter link for the same goal is on screen, as in the 1998 test.
  • Support questions asking where to find something that a banner already announces.
  • Ad reports with many counted impressions and low viewability. That points to ads that were never on screen, which is a different problem from blindness.

081 min

Banner blindness vs. nearby concepts

Several ideas sit close to banner blindness. The question to ask is what is being skipped, and why.

ConceptWhat separates it from banner blindness
PsychologyInattentional blindnessA person misses an unexpected object because attention is on another task. It is a one-time lapse, not a learned habit of skipping a kind of page area.
PsychologySelective attentionThe general ability to focus on some input and filter out the rest. Banner blindness is one result, with learned targets.
Not in the library yetAd fatigueResponse to one ad drops after many views. Banner blindness can skip an ad on its first view because of where and how it looks.
Not in the library yetAd blockingSoftware removes the ad. With banner blindness the ad is still there and the eyes skip it.
DesignChange blindnessA person misses a difference between two views. Banner blindness concerns an element that does not change.
Not in the library yetPop-upAn interruption that asks for a response before the task goes on. A banner sits in the page and can be skipped.

Guides also use the term content blindness for skipping any content, not only ads. It is the same habit, applied to content that looks like an ad.

091 min

Where the evidence on banner blindness is contested

The strongest objection says the name overstates the effect. In a 2011 eye-tracking study by Hervet and colleagues, 32 people read eight pages about digital cameras, and 37 percent of the 128 ad exposures were looked at at least once. Their point is that earlier evidence rested largely on whether people remembered ads after browsing, which cannot show whether they looked.

The Drèze and Hussherr result points the same way: people looked at just under half of the banners, not none. So the blindness is partial. Neither study asked people to do a task as urgent as the Rice searches, and the Rice pilot had six participants.

While the picture is unsettled, treat any single percentage as a local result. Marketing guides quote precise figures for how many users are affected, often from one survey by an advertising company, and this article gives none. Test on your own page, with your own task, whether people look at the part that matters.

?3 questions

Questions people ask

Does banner blindness mean banner ads do not work?

No. Looking, remembering and clicking are separate results, and a low click rate alone cannot show whether an ad was seen. Measure where people look before concluding that the ad failed.

Does banner blindness affect search results pages?

Yes, in some cases. Nielsen Norman Group observed participants who had learned to skip the ad at the top of Google search results, although its design looks little like a traditional banner.

How can I test for banner blindness before a launch?

Give testers a task the banner would help with, then compare their success with a control task that does not need it. Eye tracking or heatmaps can confirm where they looked.

§10 sources

Sources

Show all 10 sources
  1. Jan Panero Benway and David M. Lane, "Banner Blindness: Web Searchers Often Miss 'Obvious' Links", Rice University.

  2. Kara Pernice, "Banner Blindness Revisited: Users Dodge Ads on Mobile and Desktop", Nielsen Norman Group, 22 April 2018.

  3. Xavier Drèze and François-Xavier Hussherr, "Internet advertising: Is anybody watching?", Journal of Interactive Marketing 17(4), 2003.

  4. Emily Higgins, Mallorie Leinenger and Keith Rayner, "Eye movements when viewing advertisements", Frontiers in Psychology, 2014 (a review that summarises Drèze and Hussherr).

  5. Guillaume Hervet, Katherine Guérard, Sébastien Tremblay and Mohamed Saber Chtourou, "Is banner blindness genuine? Eye tracking internet text advertising", Applied Cognitive Psychology 25(5), 2011.

  6. Christian Jarrett, "Are we really blind to Internet banners?", British Psychological Society Research Digest, 17 October 2011 (a summary of Hervet et al.).

  7. Bo Ning et al., "Effects of banner ad type, web content type and theme consistency on banner blindness: an eye movement study", Cognitive Processing, 2023.

  8. "How the Banner Ad Was Born", Digiday, 12 April 2013.

  9. "Banner blindness", Wikipedia.

  10. "F-Shaped Pattern of Reading: Misunderstood, But Still Relevant", Nielsen Norman Group.

Keep reading

More from Design

All of Design
All of Design

Missing a concept?

Tell us what to write next. We’ll email you when it’s live.

Several? Separate them with commas.

Only to tell you when it’s live.

How we use your email: privacy.