Banner blindness is the habit web users have of ignoring page elements they take to be ads, whether or not those elements really are ads. Anything that sits where ads usually sit, moves like an ad, or is styled like one tends to get skipped, even when it holds the information the visitor came for.
Where the term comes from
The name comes from Rice University researchers Jan Panero Benway and David M. Lane, who wrote it up in 1998. In their tests, people looking for specific information missed it when it sat inside a big, colorful banner.
As Don Norman summarized their work, "it was the rare individual who noticed that this bold and salient banner contained the information being sought."
Nielsen Norman Group has kept testing the effect with eye tracking.
In its 2018 update, built on a study with 26 participants, one page's right rail made up 25% of the content area but got only 0.8% of fixations.
NN/g lists three triggers: ad-like placement (the top of the page or the right rail), ad-like styling (animation, colored boxes, fancy type), and sitting next to real ads.
How it shows up in CRO and A/B testing
Most stores don't run third-party ads, but they build things that look like ads: rotating hero carousels, scrolling marquee strips, promo tiles and press-logo bars.
Visitors who have learned to skip ads skip these as well.
The result is that the element you spent the most design time on can be the one nobody reads.
In testing, banner blindness usually shows up in one of three ways:
- Removing a banner doesn't hurt, or helps. If visitors were ignoring it, taking it out just frees up space for content they do read.
- Moving the same message into the content works better. An offer placed next to the button people are already looking at gets seen. The same offer in a strip at the top of the page gets skipped.
- Click maps show a dead zone. Heatmaps or scroll maps show almost no clicks or attention on a large, prominent area. That makes it a good place to test something new.
Banner blindness is a tendency, not a rule.
Some banners get plenty of attention, so check click maps and run tests on your own pages rather than assuming any top-of-page element is invisible.
Examples from the database
Dave Diederen hid a scrolling marquee bar and saw +2.31% revenue per visitor.
A moving strip at the top of the page fits two of NN/g's triggers at once: ad-like placement and ad-like motion.
Clean Commit swapped a trust-focused homepage hero for a sales carousel, and conversions fell 46.2%.
The result is consistent with a promotional, rotating format being read as advertising, while the plain trust message it replaced had been doing real work.
The opposite also happens. Dave Diederen saw on heatmaps that a collection banner was drawing attention, so they turned it into a clickable countdown timer, which brought in +$161K revenue.
The lesson is to measure first: a banner that people actually look at is worth putting to work.
A few more tests from our full test database on banners that were removed, shortened or moved closer to the task:
@DaveDiederen
Blend Commerce
@RubendgrfNone of these tests measured attention directly, so they don't prove banner blindness was the cause.
They do show that how a banner is placed and styled matters a lot, and that the outcome can go either way.
For more examples, see the tests on announcement bars and collection banners.
How to work around it
- Put key messages (offers, guarantees, shipping terms) in the content near the main action, not in a separate strip.
- Don't style important content like an ad. Avoid animation, rotating slides and boxed, colored panels for things people need to read.
- Keep real ads or cross-promotions visually separate from core content, so the content isn't skipped along with them.
- Test removing a banner before you redesign it. If removing it makes no difference, the space can go to something else.
Related terms
Banner blindness ties closely to the F-pattern, the way people scan text-heavy pages, which explains why some areas get far less attention than others.
Browse the rest of the CRO glossary for related terms.