Much of the talk about data is vague - a list of "cans," "wills" and "shoulds." Econsultancy offers a new report today - Increased ROI - A Statistical Examination of Ad Optimization - that deals in hard figures.

Does display ad optimization work? If it does, what volume is required to balance out the time and trouble? This report, from Digital Vision Winner Julia Nalven, answers those questions in detailed but straightforward language.

All that data gives digital its edge over traditional advertising, but are marketers truly leveraging the massive amounts of information they have about their audience? The right data analysis techniques can help, as does knowing how to use the results to improve performance.

Taking figures from real ad campaigns, this report (free at the Bronze membership level and above) outlines a method for optimizing online campaigns based on mathematical analysis and brand-lift data, answering such questions as:

  • What are the best metrics for understanding ad-campaign performance?
  • How can we use data to learn what works (and, conversely, what to avoid) in brand advertising?
  • How can an online advertiser allocate and reallocate ad impressions to maximize ROI?

A typical campaign, for instance, appears in a number of publications and features several creative units. A/B digital testing can tell us, in real time, how effective the campaign is and what types of ads work. It's up to the advertiser to take that data and allocate (or reallocate) ad impressions over the course of a campaign so that the best message reaches the most people. Indeed, after reviewing more than 2,000 online brand-advertising campaigns, Nalven discovered that even campaigns with as few as 2 million impressions can benefit from thoughtful data analysis in order to optimize the allocation of impressions.

Cap frequency to reach more people with the same ad budget

Ad agencies and publishers who are paid-per-click would like to believe that the more ads a person sees, the more likely she'll be influenced by the campaign. Not so, says Nalven. Reviewing brand-lift figures from numerous campaigns, she found that there's a magic number for influencing a viewer. After a person has seen, for example, ten display banners promoting Hershey bars, that person will be influenced to buy a Hershey bar at the grocery next time. There's no need after the tenth display to show the ad to that person again.

By limiting how often an ad is shown to a consumer, the marketer can reach a wider audience and thus increase the total number of people influenced without increasing ad spend. 

The ultimate goal for digital marketers is understanding how to reach consumers and how to influence as many of them as possible. When an avertiser or agency understands the various audiences out there and how they spend their time online, it's possible to maximize ad spend as well ROI. Given the amount of data available today, there's no reason not to have the best ad campaign possible.

For the full story, read the report here: Increased ROI - A Statistical Examination of Ad Optimization.

Cielo Lutino

Published 4 October, 2012 by Cielo Lutino

Cielo Lutino is a writer and producer for Econsultancy and other organizations. 

32 more posts from this author

You might be interested in

Comments (1)


Matt Lovell, Head of Customer Data, Insight & Analytics at Eurostar International Ltd.

Interesting document which is defintiely a different way of approaching things to what most agencies / advertisers are using currently.

One question though - what in the world is a Vote in the example? Am not entirely sure where that comes in to it but certainly on the creatiove front, it looks to me as though based on the responses versus impressions, Creative C is actually under performing yet because it has less votes, is being upweighted.

I also feel this doesn't take into account whether users actually see an ad (let alone whether its for long enough to actually have any recall of this ad) which makes this sort of analysis much more cloudy.

almost 6 years ago

Save or Cancel

Enjoying this article?

Get more just like this, delivered to your inbox.

Keep up to date with the latest analysis, inspiration and learning from the Econsultancy blog with our free Digital Pulse newsletter. You will receive a hand-picked digest of the latest and greatest articles, as well as snippets of new market data, best practice guides and trends research.