Beyond the Click: Which online ads are the most effective? (Part 1)

If you are responsible for developing the ad content for your website or indeed if you are in the process of selecting ad placements on behalf of an advertiser then read on.

Whenever Eyetrack III publishes findings regarding online ad effectiveness the digital industry takes note. Why? Mainly because Eyetrack are purveyors of proprietary technology which enables tracking of consumer 'eye fixations' across a webpage. Termed 'heatmaps', these colorful images provide researchers with precise aggregate representations of where consumers are looking. In other words, they can answer the question; 'do consumers actually see your ads?', taking our understanding beyond the click and into the realms of ad impact and brand awareness.

Eyetrack's latest study is based on research across 46 participants who viewed a series of mock news websites. Five website designs contained a variety of ad formats and layouts to enable comparisons of performance. The results of these studies provided some interesting observations which I can summarise as follows:

Eyetrack III Observation 1: 'People avoid looking directly at some ads while looking at others; performance depends on placement.'

No big surprise here, but the observation does address a big concern in online media, that of overcoming 'ad invisibility' or 'banner blindness'. The findings of the ad viewing percentages in the study were found to range from 14 to 68 percent. Much higher than click percentages and great news for branding, as even if the ad unit doesn't get clicked it may still have been directly viewed. Positive news for advertisers which base so much of a campaign's success on the total number of clicks.

In total, 4 ad placements were compared. The Ad locations were Top of page, Left column, right column, and bottom of page. Unsurprisingly bottom of the page came in a distant last place as its placement far 'below the fold' ensured that only 1 in 10 viewers even saw the ad. What was more surprising is that left hand column placement came top of all ads. Perhaps this may be due to its proximity to the main vertical navigation which normally sits on the left side.

Eyetrack III Observation 2: 'Visual Breaks in design serve as barriers to seeing ads'.

Borders between ads and editorial content in form of space or line breaks affected the viewing performance considerably. These simple design touches stopped participants viewing the ads completely. Whats more, the increase in the space or border between the ad and the editorial content produced even less ad viewings. A point to consider when your web designers decide to add some design elements to the overall page layout.

Eyetrack III Observation 3: 'Ads that blend into the look and feel of the page draw more eyes'

This one contradicts most traditional online ad thinking. Normally we try to contrast the ads from the surrounding editorial content in an effort to make it 'stand out' to the consumer. According to Eyetrack, ads which blended into the surrounding environment (ie. similar background color) received more eye fixations. Perhaps consumers have perfected 'banner blindness' when surfing the web, even better than we thought.

To make things even more surprising, a series of text ads produced the highest viewing percentages of all ads with 82 percent of all participants having looked directly at the ads. The study doesn't go so far as to give possible reasons for this but certainly the little used text ad, when properly placed can give publishers a strong revenue stream whilst providing advertisers with a low-cost effective ad placement.

Eyetrack III Observation 4: 'Size Matters; Half-page ads perform well'

The observation suggested that larger ads are seen before smaller ones, so no surprise there, but the study also revealed that the larger ads do not get viewed for any longer than their smaller counterparts. All banners were seen on average of between 0.6 to 1.6 seconds, not much time to make an impression.

In terms of average fixations per user, ads which were placed within the editorial content (ie. island ads) were found to be among the best performing. As were roadblock style half-page ads which dominate a page, unsurprisingly.

Eyetrack III Observation 5: 'Ads inset within article text are seen more than other ads'

Bringing observation 4 forward, further research revealed that a 300x250 pixel ad placed within the actual article content produced higher percentage of direct eye fixations than any other ad (other than those text ads again) at 60%. Even the half-page ad wasn't viewed as often.

Considering that in this region, very few publishers even offer this type of ad placement, then this becomes quite a profound finding. As with the other observations, the study continues to reveal that ads which can integrate within the content of the site provides better 'eyeballs' for the advertiser. Since these ads aren't normally on offer here, then this becomes a potentially new revenue stream for many publishers across the region.

Eyetrack III Observation 6: 'Mouseover-expand ads were viewed more than other banner ads'

Expandables which open upon user interaction performed much better than the same banners without expandability function. In fact, more people initiated the expandability than even noticed the non-expandable counterpart. Much of this had to do with the message but the study doesn't provide any conclusive reasoning for this occurrence.

Furthermore, the expandables received much higher than average viewing durations. Meaning that the participants were spending more time engaged with these ads. Great news for branding aficionados.

The observations above provide some useful insights which we in the industry should take heed. Many of the common preconceptions have been put to the test and not all of them confirmed their validity. But overall, the news from the study is positive and encouraging and the advice that we can take from this confirms that online advertising is about more than just clicks.

In the second part of this article I will focus on the remaining Eyetrack III observations and provide a summary list of conclusions to consider.