Beyond Clicks: Nailing Down Ad Effectiveness


hammer_smallADOTAS – For the past decade, advertising was measured by reach, frequency and Gross Rating Points (GRP). However, when online advertising came about, it had a distinct advantage — measuring users’ reaction to the ad that was presented to them.

This is how the click was born, and became synonymous with online advertising analytics. However, in the past few years, research has emerged to suggest that clicks are a very partial measure of advertising effectiveness. Now it seems that industry mavens are advocating the use of the good old offline metrics for online advertising and thus developing cross-channel measurement.

The online research firm comScore maintains that the primary effect of online ads is the exposure itself and not necessarily the interaction with it. In research titled “How Online Advertising Works: Whither The Click?”, comScore has shown that two-thirds of Internet users do not click on any display ads over the course of a month and that only 16% of Internet users account for 80% of all clicks.

Furthermore, clickers tend to be younger and less affluent than non-clickers. comScore confirmed that there is a latency effect and branding effect to online advertising, in which users arrive at the advertisers website even without clicking.

However, the research by comScore also indicates that display advertising has an effect on user behavior even at low click through rates. In the research, which included 139 display campaigns from seven verticals, comScore has shown substantial effects on traffic, sales and branding despite a lack of clicks.

According to comScore, the display campaigns yielded a 46% lift in advertiser websites visits, over a four week period. In addition, exposed users are 38% more likely to conduct an advertiser related branded keyword search, over a four week period and are 27% more likely to make a purchase online. Furthermore, exposed users are 17% more likely to make a purchase at the advertiser’s retail store.

The view that the exposure is what matters rather than the interaction was further supported by a recent research from the American Online Publishers Association (OPA). Using eye-tracking technology, the OPA tested to see whether people actually viewed the ads that were presented to them on a page that they were browsing.

In addition, the researchers used biometric testing to track the participants’ emotional response by monitoring their breath and heart rate. The biometric testing is similar to a lie detector test, and signifies a positive or negative emotional response.

The results were staggering, with 96% of participants paying attention to the ads that were presented to them during a natural browsing experience. Furthermore, 90% of participants noticed the ads during the first 10 seconds of uploading the page. In fact, on average, it took participants less than one second to notice the banner.

Another important finding is that most participants look at the ad on the page multiple times. The OPA researchers found that 67% of participants view the ad both during the first 10 seconds after loading the page and view it again during the following 10 seconds. On average, participants have fixated their eyes on the banners 15 times.

When measuring the emotional response, the OPA found that participants responded to the banners in a similar manner to other parts of the page. However, users who viewed the ad again after the first 10 seconds exhibited elevated emotional response. Users also rated the ads favourably overall, giving them a self reporting rating of 6.3 out of 9.

Another research by Microsoft, comScore and MediaMind found that engagement with an ad is sufficient to spring users to action. The research used the online engagement metric “Dwell” that measures the proportion of users that hovered their mouse over the ad as a proxy to actual ad exposure. The research found that campaigns with high Dwell generated more site traffic, enticed more brand related keyword searches and commanded more “brand engagement.”

With this insurmountable evidence that a large part of the job of online advertising is done at the exposure phase, rather than the interaction phase, measurement companies such as Nielsen are working to develop new tools for measuring reach and frequency similar to the way it is done offline. In addition to measuring online advertising, these tools will enable cross channel measurement—the ability to tally the number of people that have seen your brand regardless of media.

The new service by Nielsen, named “Online Campaign Ratings,” will present online reach, frequency and GRP data in a similar fashion to Nielsen’s television ratings. In a company statement Steve Hasker, president of media products at Nielsen acknowledged that the introduction of reach and frequency does not make online analytics redundant — “Reach alone doesn’t tell the whole story, however, and Nielsen intends to combine the ratings with cross-platform advertising effectiveness metrics to provide a complete view.”

Overall, adding reach, frequency and GRP to the traditional click-based measurement and advertising web analytics is a welcomed development. It complements the interactivity metrics with measurement of exposure that has been shown by research to be linked to ad effectiveness.

Most important of all, it will enable cross-channel offline and online measurement and make both channels speak the same language.


  1. It’s important to think about the evolution of the display ad and new formats that look to engage with consumers in new ways beyond the traditional clickthrough. For instance, at Widgetbox we’re working with publishers and advertisers to create ads that pull in interactive, social content that compel consumers to hover over the ad for more content, and don’t necessitate a clickthrough. That’s still impactful without bringing the consumer to a new webpage.

  2. Ahh, Telmar is glad to hear that Reach & Frequency has been added back into digital. While we’re a fan of new metrics (Impact, engagement, etc), we know eventually folks will want to know ‘how many people they reached and how often’

    Et voilà.


Please enter your comment!
Please enter your name here