Report: Clicks and impressions don't correlate with conversions

Share this article:

Measuring ads by viewability and hover time had a stronger correlation with conversions than measuring with clicks or total impressions, said Andrew Lipsman, VP of industry analysis at comScore.

The Apr. 24 report, published by comScore and Pretarget, a provider of online ad targeting solutions, showed that there was little correlation between clicks and marketing objectives. “We've known clicks are not the best metric, but the fact that it showed no correlation underscores how perilous it might be for brand advertisers to look at a metric like that in evaluating campaign performance,” Lipsman said.

Pretarget analyzed 263 million impressions over a nine-month period across 18 advertisers in various verticals. The company used comScore's validated Campaign Essentials (vCE) tool to collect viewability and hover data, as well as a digital signal processor (DSP) to collect clicks and cookie-based conversion data.

Pretarget then conducted a correlation analysis of the data, which included gross impressions, views (defined as 75% of an ad visible within the screen above the fold or after scrolling), time the ad was viewed, hover or engagements, total hover or engagement time, clicks and conversions.

Lipsman said conversions were defined as “any definitive online action where there was a submission in the process.” This could include an online purchase or filling out a survey, depending on the specific campaign.

The research showed ad hover/interaction and viewable impressions, correlated to conversions at .49 and .35 respectively; gross impressions and clicks correlated at .17 and .01 respectively.

Lipsman said comScore has previously performed studies that also showed clicks are not the best measure for evaluating whether a marketer's campaign is effective, including the 2009 Google Natural Born Clickers study.

"This is just another data point that confirms that brand marketers need to break their addiction to the click,” Lipsman said. “Marketers should be thinking about other ways to measure the performance of their campaigns.”

This material may not be published, broadcast, rewritten or redistributed in any form without prior authorization. Your use of this website constitutes acceptance of Haymarket Media's Privacy Policy and Terms & Conditions