Study: netsurfers avoid flashy banner ads, prefer text ads

Clickz.com - An eye-tracking study conducted by the Nielsen/Norman Group finds Internet users avoid viewing banner ads. Text advertising is read more often than display ads, according to the research.

Banner blindness means Internet users focus on the content on a page and ignore the advertisements. This is especially true for bright, flashing ads, and other units that are not relevant to what the user is interested in reading, the researchers found.

"People are not looking at the typical blinding, graphical ads," said Nielsen Norman Group Director of Research Kara Pernice Coyne. "They are not [looking] enough time to absorb a complex ad or branding message."

There's still hope for online ads. Pernice Coyne said graphical ads with text and contrasting colors, like white text on red, is less likely to be disregarded. "They're looking at them if they're text," she said. "I hate to sound boring, but [it is best] if you can make sure your ad is something simple, text or a recognized logo, and it needs to be relevant to the page."

Pernice Coyne suggests the behavior may be due, in part, to experience with ads that direct users to sites with malicious code or other undesirable results. "There are too many unscrupulous advertisers," she said. "Users are not going to click on something they think will annoy them or hurt their system."

While the study's finding would seem to threaten the rising Internet ad spend, the Internet Advertising Bureau (IAB), at least, isn't concerned.

"There literally have been thousands of studies now on online advertising's effectiveness," said IAB President Greg Stuart. "All the evidence is in: smoking kills, online advertising works. There is no more information to be had. We can deny the information, but that's all it is: denial."

Stuart points to the IAB's cross media optimization studies (XMOS), which demonstrates the value of online advertising. While the research doesn't observe usability, it measures the medium against other channels.

"All of those studies indicate that online advertising is effective and that in most cases it is the most cost-effective medium in a marketer's mix," said Stuart.

Stuart agrees that things could work better. "I believe very strongly that we can still continue to make online advertising work even harder," he said.

Though the eye-tracking study casts doubt on banners, it supports the effectiveness of one category: search. "People do look at sponsored links on search pages and images on search pages," said Pernice Coyne. "They really look for words that match what they are searching for. If you have a strange title to the page, people will skip over it.

"It's so important what you call your links in search results. Users don't give you a lot of chances," Pernice Coyne said.

Previous reports have shown sponsored listings can succeed beyond the first position. The NNG research finds it is important to be placed among the first few.

"In search results, people look mostly at the first few links on the page," said Pernice Coyne.

Sponsored links need to closely relate to search terms to be successful. "When you look at [the sponsored links] and they're not related to the content, users look away quickly, almost like you're slapping them on the hand," said Pernice Coyne. "These types of advertisers doing this type of advertising, I feel like they're going to ruin it very quickly. People are incredibly good at blocking things out they don’t like."

The researchers also found that people read Web pages in an F-pattern, narrowing their focus as they scroll down a page of content. Pernice Coyne said readers fixate or focus on the content at the top of a page, read a little bit further down, then give up and go back to the beginning of the same or subsequent page.

Images that appear in the middle of the page, a spot for advertisements, are considered "obstacles" and annoying.

The usability study was conducted on a group of about 230 people. Participants were asked to surf the Web to their own destinations. They were also given tasks such as finding out how to tie a bowline knot and researching getting a mortgage. Participants carried out these tasks while a camera tracked the pupil as the eye engaged what was on the screen.