Marketing and Research Consulting for a Brave New World
Subscribe via RSS

We are measuring the wrong thing when it comes to digital display advertising.  We are still focused on a DR model of digital advertising leading us to place our energies on measuring the click when digital advertising is, in fact, ADVERTISING!

Let me illustrate with simple arithmetic why the click is a red herring measure of digital advertising effectiveness.  Click rates, on average, are around 0.1% (one tenth of one percent) on a good day. However, Gian Fulgoni, President of Comscore who was the lead author on an article (JAR, 2009) entitled “Whither the click?” showed an average lift of 1.4% percentage points (off of a base of 2.1%), test vs. control in visitation to the advertiser’s home page in the first week after ad exposure.  That suggests that the total effect of digital advertising (view through plus click through) is about 10 times greater than can be accounted for by the click!

Let’s create some sample math from Gian’s article to understand this.

  • Propensity to visit the home page without an ad: 2.1%
  • Say, 0.1% click on the ad and assume that takes them to the home page, so propensity to visit the home page after clicking is 100%
  • This implies that the propensity to visit the home page if you do not click but are exposed to an ad is approximately 3.4% (an increase of 1.3%/99.9%)

Okay, so now let’s see where the leverage is.

If you can double the click through rate to .2% (hey doubling ad effects is really good), the increase in effect is about 1.4 to 1.5 or 7%.  However, if you can double the view-through effect (which presumably is like good old advertising effectiveness), the impact is close to double because over 90% of the lift in home page visitation was coming from the branding effect of display advertising in the first place.

Clearly, the leverage is in understanding and amplifying the effect on behavior of those exposed but who do not click on the ad.

When I look at dashboards of advertising effectiveness like from Double-click, click rate is most prominent.  If there is a conversion metric, it is still the percent of those who clicked that then matched to a conversion pixel.  The digital analytics infrastructure is not geared to report on the pathway by which 90% of digital advertising effect occurs!

To address this, advertisers must define what a conversion is (I believe that most do not, especially when sales occur offline) and then put the conversion pixel into the conclusion of that action.  It is not always a sale and perhaps should NOT be the sale but some behavior that is a highly correlated marker for sales.  It can be downloading a coupon, signing up for a database, requesting a price quote, sharing an asset, etc.  Then we must start tracking conversions based on exposures not only clicks. Important words for digital…behavior…highly correlated to sales.

None of this is beyond the realm of possibility because digital is all about behavior and gives us the opportunity to track this.  We CAN fully account for the effectiveness of digital display; Advertisers just need to start requiring the ecosystem and their chosen partners to start measuring the right thing, and start doing it in real time for digital advertising to fulfill its promise.

Tags: , , , , ,


One Response to “Why advertisers do not measure digital advertising effectiveness correctly”

  1. We’re also in agreement!

    Click-through-rates (CTRs) are not the right measure. One of our research directors recently suggested that relying on one-size-fits-all measures such as CTRs to understand how well a digital ad performs would be like relying on website visits to measure the impact of a print ad that happens to include a web address. We can’t throw out everything we already know about how successful ads operate when we look at digital advertising media. All ads succeed by grabbing and holding the viewer’s attention, focusing her on the brand name and motivating her to take action.

    Digital ads aren’t designed just to elicit digital responses, such as clicking through and making online purchases. An Audi digital display ad can pique my interest in their newest models without ever convincing me to click through to their website. No digital experience can ever simulate the analogue experience of actually test driving a new car, smelling the leather, and holding down the clutch to go into fifth.

    In fact, many digital display ads do not even include links to websites. As a case in point, a recent ad for Claussen pickles made my mouth water. There was no link to a Claussen website so I could learn more about their pickling process or to share pickle recipes. The call to action was to find their pickles at my grocery store. You know, the old fashioned way. The way almost everyone still buys pickles.

    Something that has contributed to confusion about measuring advertising performance in the digital space is the myriad of shapes and flavors digital ads can take. The digital space is less a medium than it is a set of media in its own right. But it’s not a space that can’t be understood. At Ameritest, we’ve developed a taxonomy to classify the digital world. This helps us choose the right tools to test digital ads based on their design and objectives.

    While CTRs are valuable to collect for ads that intend to generate click-through responses, they should only be one measure in a battery of diagnostics. And copy testers already have those diagnostics. We do, after all, understand how advertising works, Flash or no Flash.

    To learn more about our digital taxonomy, contact