The Stats Explain Why We Don't Read Banner Ads (And What We Can Do)

Jakob Nielsen recently published two interesting research statistics that help explain why display ads don't work.

First, Jakob says we read a very small amount of the text on most pages. Apparently we only, on average, read at best 28% of the words on the page and most likely around 20% of the words.

Second, in an older article from late last year, Jakob also says users avoid anything that looks like a banner. This may seem obvious, but the interesting thing to consider is that all web pages on a given site (and even across the web on different sites) are likely to have the same pattern for what looks like an ad vs what looks like content. Your eyes quickly learn that anything that is on the sides, or anything with a border around it, or anything that looks photographic is an ad.

Newspapers and magazines generally do not have any such easily to discern pattern. Pages are generally laid out manually by a designer on an issue by issue basis. Smaller ads are blended in with the content, breaking columns and such so that advertising will have unavoidable and unpredictable proximity to content. Moreover, many ads take up the full page which requires at least some minimal scanning before turning the page. And because the ads are often so big, you almost can't avoid at least seeing a bit of it before turning the page.

The bottom line is that users see too much stuff on the web and so they focus on very little. On top of that they have learned effective patterns for avoiding ads. And while we try to avoid ads in print, the physical size of the advertising, the unpredictable patterns, and the high degree of mixing advertising and content make pattern based ad avoidance much harder in print.

In my view, the model for how we organize information on the web is fundamentally broken as it relates to siphoning off attention for advertisers. We need to inject more randomness into our designs. Indeed, the kinds of thing that Jakob Nielsen promotes as improving usability do in fact promote usability while *reducing* monetizability.

Magazines and newspapers are *not at all useable*. I am often frustrated that it is very difficult to find a table of contents in a magazine, or that some pages are missing page numbers that would help me find an article. This is all on purpose!

Print publications are very effective advertising platforms in large part specifically because they place "attention speedbumps" in your way. If print publications were as streamlined as websites, with ads neatly presented off to the side, ads wouldn't work there either.

And so we have a decade of smart folks like Jakob teaching us exactly the right stuff to make our sites easy, but exactly the wrong strategies for making money. Amusingly, at least a part of the message is that a little disorganization is absolutely critical.