Dynamic Logic offers up proof that frequency capping can improve the results of your campaign.

Now that marketers are embracing digital media, its versatility and targeting capabilities, one of the more important and differentiating features of digital marketing is an ad-serving option known as "Frequency Capping."

Frequency capping is the restriction on the amount of times a specific person (or technically their browser cookie) is shown a particular online ad, and is often cited as a way to avoid "banner burnout": the point at which visitors are being overexposed to an ad and its impact begins to decline.

Frequency capping improves reach, increases the number of people impacted and, importantly, significantly increases the cost-effectiveness of online media buying. For example, if 1,000,000 impressions are served with a frequency cap of 1, they will be delivered to 1,000,000 people, whereas a typical campaign, with an average frequency of 2.5, will reach only 400,000 people with 1,000,000 impressions.

The chart below -- a frequency curve based on real data -- shows increases in brand awareness and purchase intent as exposure frequency increases.

Brand Impact Versus Frequency

We know that the first impression provides the greatest impact, and each subsequent exposure has declining incremental impact. So, more exposure is a good thing, right?

Not so fast. As exposure frequency increases, the reach decreases. The chart below shows the reach associated with 1,000,000 impressions at different average frequency levels.

Reach Versus Frequency for a Fixed Number of Impressions (1 Million)

If impact increases with increasing frequency, but reach declines, what is the best way to examine the trade-offs? One way is to try to estimate the number of people who would be impacted at different average frequency levels.

We will show that frequency capping, where the number of times a person sees an ad is capped at a pre-specified amount, is more cost effective than the average campaign in which it is not uncommon for some people to see the same ad 10, 20 or even 100 times.

Let's look at an example, using real data, where we examine both an awareness metric (Aided Brand Awareness) and a persuasion metric (Purchase Intent). Suppose you purchased 1,000,000 impressions to be delivered in one week on a popular sports site. If the site is typical, you might expect that those 1,000,000 ad impressions will be delivered to approximately 400,000 people (an average of 2.5 times per person). Some site visitors will be delivered the ad once, some twice, et cetera.

Based on Dynamic Logic MarketNorms sport site data from the past two years, if 400,000 people are exposed to a campaign with a typical exposure frequency, we would expect roughly 3.0 percent or 12,000 people to become aware of the advertised brand who otherwise would not have been. Regarding Purchase Intent, we'd expect an incremental 1.6 percent or 6,400 to report being likely or very likely to purchase. So, using real data from 145,000 survey respondents across 176 online campaigns that ran on sports sites in the past two years, we've just estimated the number of people impacted by 1,000,000 impressions.

Let's go through this exercise assuming those same 1,000,000 impressions are delivered with a frequency cap of 1 (compared to the above scenario where no caps are in affect). Based on our data, the average impact per person is noticeably less: For Aided Brand Awareness, just 1.8 percent rather than 2.4 percent become aware of the brand who otherwise would not have been and 1.3 percent report becoming likely or very likely to purchase.

With frequency capping, the impact per person is less, but let's examine how many people would be impacted. Those 1,000,000 impressions you purchased will be viewed by 1,000,000 people or 2.5 times as many people as the uncapped campaigns. And, although the yield on a per 100 person basis is lower, the total number of people impacted is greater: for Brand Awareness, 1.8 percent of 1,000,000 translates to 18,000 people persuaded and for Purchase Intent, 1.3 percent of 1,000,000 translates to 13,000 people impacted. Those numbers are 50 percent and 103 percent greater for Brand Awareness and Purchase Intent, respectively, than estimates from equivalent upcapped campaigns.

So, frequency capping with a cap of 1 is more cost effective than a typical campaign with average frequency of 2.5. The chart below shows the estimated number of people impacted (with respect to Brand Awareness and Purchase Intent) by 1,000,000 impressions at different frequency levels. It clearly demonstrates that you change more people's attitudes about your brand by limiting the number of exposures.

People Impacted by Frequency, for a Fixed Number of Impressions (1 Million)

Thus, frequency capping is a highly recommended way to improve reach, increase the number of people impacted and, importantly, increase the cost effectiveness of an online media buy.

So, what's the catch? Why aren't all campaigns frequency capped?

Frequency capping requires more ad inventory. Therefore, while in theory it is best to do this, in practice, sites will be crunched for inventory if everyone is drawn to frequency capping. Furthermore, as capping becomes more popular, sites will begin charging a premium for capped ad delivery. But, as of now, frequency capping is an efficient way to increase reach and impact. So, if you want to stretch your online dollar further, the time to try frequency capping is now.  

Ken Mallon is VP, ad effectiveness consulting at Dynamic Logic, a Millward Brown Company. Read full bio.