The Oracle Data Cloud blog highlights the latest data-driven insights and trends in digital marketing and ad tech.

  • December 3, 2015

Don’t get skewed by data averages

This week's blog is contributed by Mike Schumacher, Senior Director: Data Science, Oracle Data Cloud.

Scrutinize your publisher’s impression frequency per cookie for a simple, quick improvement on your ROI.

Imagine you are a Media Planner in San Diego who sp­ends your free time running along the Pacific Coast and enjoying outdoor BBQs year ‘round. But you receive a fantastic job opportunity in Washington, DC. You love the offer, but giving up San Diego’s beautiful weather with an average daily high of 69 degrees? Not so much.

After doing your research, you discover that DC’s average high is oh-so-close at 68 degrees (a fact*). You take the job and then spend a tumultuous year in DC with 100+ degree summer heat followed by winter snowstorms. The lesson: an average can be very misleading!

All too often, we allow the convenience of simplicity to obscure the truth in the data. This misleading “average” can be particularly troublesome when evaluating impression frequency per user.

Most campaigns demonstrate reasonable average impressions per user, but show dramatically skewed underlying distributions.

The following two campaigns both average 10 impressions per user:


Tale of Two Averages Tale of Two Averages

In the first example, 87% of exposed cookies saw between 7 and 13 impressions (+/- 3 of the average). In the second example, less than 1% of households saw between 7 and 13 impressions. Clearly, these are radically different executions with very different performance expectations. Despite identical average frequencies per cookie.

Here at Oracle Data Cloud, we wanted to understand how often this situation occurs and took a sample of recent digital campaigns to evaluate their impression frequency distributions. We measured each campaign’s average impression frequency and built a +/- 3 impression range around it.

Only 33% of the typical campaign’s exposed audience resided within that target range. Even more dramatic, on average, over 44% of impressions were delivered to only 5% of the exposed audience (extreme skew).

Skip the average, ask for a distribution chart, and make analytics work for you.

A strong audience, publisher, and creative can’t overcome flawed campaign execution. Flawless campaign execution is a big part of generating strong ROI. By understanding impression distribution per cookie report with your publisher or ad-server, you can ensure there are no misallocations of marketing budget buried behind the simplicity of an average. Asking the questions can put you on the road to a strong ROI. And potentially earn yourself a nice run on the beach.

Stay up to date with all the latest in data-driven news by following @OracleDataCloud on Twitter and Facebook! Need data-related answers for your next marketing campaign or client partner? Contact The Data Hotline today. (What's The Data Hotline?) 

Photo: Shutterstock

Be the first to comment

Comments ( 0 )
Please enter your name.Please provide a valid email address.Please enter a comment.CAPTCHA challenge response provided was incorrect. Please try again.