Display Marketing ROI – The Placebo Experiment, Part 1

In 2011’s online display advertising environment there are more potential KPI’s and metrics than any one human could keep track of and evaluate. Why? Because much like TV before it, display advertising whether it be a banner or video – can be more about BRANDING than it is about taking action, and determining BRAND influence is usually dependent on a variety of somewhat subjective variables.

While some business models are based on meeting strict CPA’s pulled from online sales, there are a variety of other models that require digital marketers to allocate some of display’s ROI performance to non-purchase actions. One vertical this stands true is in the automobile industry. Companies like Chevy, Ford, Toyota, Honda, etc. can’t base their display marketing ROI on online sales numbers alone. The consumer isn’t fundamentally wired yet to make their car purchases online. Online marketers that deal with the same consumer intent issues are looking for measurements that share ROI based on other factors that deal with engagement and exposure (time-on-site), clicks (site visits), visitor funnels (influencing path analysis for web users), and more.

One in particular that interests me while working with display is the volume and percentage share of unique web traffic that was exposed to a display ad during the run of a campaign. Let’s say you ran a display campaign for one month (assume you had the necessary tags in place) and measure unique visits to your site that were also exposed to any one of your display ads. Not only unique click-thru traffic generated by the ad, but also users that were exposed at any time during their browsing sessions.

You realize that a total of 23,000 unique visitors to “yoursite.com” had been exposed to a display ad throughout the campaign and that equals 40% of your unique visitor traffic in that time period. Some marketers would end the analysis  there and say “40% of our unique traffic is exposed to our display campaign so it has a significant impact on bringing unique visitors to our site.” An average marketer would be satisfied with this conclusion. While their stat may be true, there are a variety of other factors that have to be taken into consideration before proving that the stat is valid.

1. What % of your digital marketing budget is attributed to display?

If you are spending 60%+ of your digital marketing budget on display then it should account for more than 40% of your unique web traffic signifying that effectiveness may not be strong for your brand when it comes to display. On the other end of the spectrum, let’s say you allocate 10% of your spend to display and 40% of your unique visitors are exposed to the campaign. Now, you can potentially argue that display is a prominent mover of unique traffic to your site.

2. Timing vs. Volume

Search is obviously the primary driver of unique traffic to your site with PR coverage coming in right behind it. If you run a display campaign for your brand during a time when you receive elevated press coverage it could be difficult for your display campaign to drive a large % of exposed unique traffic because you are getting an unusually high number of unique visits at the time. To combat this, run a display campaign during peak press weeks/months and run a display campaign during the down months and compare and contrast the % of unique users exposed.

3. What is the control % of unique users exposed to a display campaign when I run X amount of impressions?

This was the head-scratcher for me when trying to determine the true effect display had on the volume of unique visitors to “yoursite.com”. It is also the most difficult to comprehend. In simple terms, I was trying to validate the statement “okay…40% of our traffic was exposed to our display campaign but that doesn’t mean that all of those visitors saw our ad, it just means they happened to be surfing a page that displayed it (pun intended).”  What % of those users could I say were actually influenced by the ad and eventually came to the site either by clicking-thru the display ad or during a later browsing session? Measuring the clicks of the display ad is easy…allocate clicked % and subtract from the overall. But measuring the leftover users is not that simple. Then it came to me… let’s run a blank advertisement – sounds crazy right?

Read Part 2 to discover how you can effectively do this…

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: