New research claiming online advertising click-throughs may not be all they seem has been brushed off by local interactive advertising bodies.
The research, released last month, claims just 6% of the internet population accounts for 50% of all ad clicks. Further, it claims the demographic profile of this 6% is not ideal, with an average household income of less than US$40,000. The research was released in a report titled “Natural born clickers”, by media agency Starcom, and conducted with the aid of AOL’s Tacoda, and internet measurement company comScore.
Mark Evans, CEO of the Interactive Advertising Bureau, downplayed the research in an email to Computerworld, saying it “isn’t really all that startling or insightful”. He went on to analyse the implications, firstly, for direct response campaigns and, secondly, for branding campaigns.
Evans says direct response advertising is a numbers game and success is based on conversion rate, or how many people take up a given offer, not on click-throughs.
“The fact that (according to this piece) 50% of click-throughs are done by 6% of the viewers, and the fact that those viewers aren’t necessarily a fabulous demographic actually doesn’t impact on the equation,” he says.
“It would be nice, of course, if all viewers were high socio-economic, motivated buyers, but the critical factor is simply the ROI on the investment. Did the amount I paid to get the number of conversions I wanted stack up against the other advertising options that I have?”
Tony Marlow, associate research director for Nielsen Online in the Asia Pacific region, agrees, saying the idea that “click-thru can be used as a measure of success for display advertising is a pretty outdated notion”.
“You need to look at the primary objective of display campaigns, and for the majority you see across publishers like MSN it is about brand affinity and message cut-through. If you think about an ad telling you about the latest razor with even more blades — that ad isn’t trying to elicit a click, it is trying to deliver a specific message and associate their brand name with that message for future reference,” Marlow says.
In a further sign that performance measurement for online advertising is continuing to morph, CNET reported last week that Microsoft is proposing the credit for online ad conversion be “shared”, arguing the last click a consumer makes may not be the most important. That final click on a text ad could have been prompted by, say, a related video ad.
Microsoft is proposing what it calls an “engagement mapping” model for measurement. It announced it had such a tool in beta at last week’s Interactive Advertising Bureau annual meeting, held in the US.
Evans says interactive advertising offers a better ability to track ROI on advertising investment than other media.
“I know exactly how many people saw my ad (ad impressions) from the adserving service’s report. I know exactly how many people clicked on the ad (from the ad-server reports again). And I know how many of the people who visited my site via the click-through actually purchased (or whatever it was I wanted them to do),” Evans writes.
“Since I know what the cost of my advertising was, I know whether this equation is profitable in terms of the numbers I got converting.”
Marlow says the industry is moving past limited interpretations of effectiveness.
“At Nielsen Online, we have been successfully measuring the impact of online advertising across a variety of brand-centric measures, such as brand favourability, awareness and purchase propensity,” he says.
“It is our unique ability to accurately attribute changes in such metrics to a specific online campaign or even a particular element of a campaign that can show the true impact of these efforts, and allow advertisers to accurately know the level of ROI they are achieving for their brand.”
Evans says the Starcom report is not specific about the sites it is measuring and he suspects it is averaged across everything — which would hide what are likely to be significant variations between sites. He adds that interactive specialists are now pretty good at targeting campaigns to sites that provide the best conversion rates at the lowest cost.
“For instance, if I paid $100 CPM (cost per 1,000 ads) on Computerworld and $50 CPM ads on Yahoo, but Computerworld delivered 10 conversions per 1,000 ads and Yahoo only two, I would know that Computerworld is, basically, five times more efficient,” he says. (Disclosure: Evans used to work for Computerworld’s publisher).
Evans says, for this reason, agencies are moving away from simply basing their ROI calculations on CPM rates.
He says the research does not affect online brand advertising either. Rather, it says that optimising click-through rates doesn’t generate better results for brand advertising and, therefore, click-throughs are not a good metric for measuring brand advertising success. He adds that neither of these two points is “in the slightest way new”.
Success should be based on other metrics such as ad recall and brand awareness and perception, using traditional pre- and post-campaign techniques: surveys. And the online world offers new opportunities for conducting these surveys, says Evans. New online advertising modes are making the impact of online brand advertising “dramatic”.
Marlow says methods, such as Nielsen Online’s “tagging”, mean advertisers can “know beyond doubt” who has been exposed to an ad, and who remembers it.
“So, what makes this really interesting is that despite low levels of overt ad recall ad impact still tends to be high. Put simply, we know that just being exposed to an ad is enough to generate a significant positive brand-shift, and, while I don’t think it’s any kind of subliminal effect, we know it doesn’t matter that the ad is not consciously recalled, it still did its job and did it very well.
“I think that is a very powerful understanding for marketers and content publishers to have.”