Rethinking Low Frequency Caps

Many veterans of the online advertising industry are starting to acknowledge that their single-minded focus on direct response and click-through rate (CTR) during the formation of this industry was a costly mistake.

The cult of click-through has hurt publishers because they have been driven toward punitive cost-per-action pricing models and devalued inventory. Agencies have had trouble justifying their client’s Web investments. And advertisers have been hobbled by metrics that don’t correspond to their online objectives.

Fortunately, many people are waking up to the fact that the 99.5 percent of online advertising that isn’t clicked on has measurable value. But despite that recent enlightenment, vestiges of the direct response fixation remain. One of them is putting a low cap on the number of advertising impressions that any one person will see.

Calculating advertising reach and frequency is a basic part of offline media planning and buying. Though determining reach on the Internet is tricky, ad serving systems allow frequency to be controlled relatively easily. This power is frequently put to use, with many planners (and networks) capping frequency at three exposures.

Here’s one reason why: In 1996 DoubleClick published an influential study showing an inverse relationship between frequency and click-through. In other words, the more someone is exposed to a Web advertisement, the less likely he or she is to click on it. CTR dropped precipitously after the first exposure and plateaued at four.

This was enough to convince many media planners to set low frequency caps on their online media buys. Theory had it that if people don’t click on an ad the first few times they see it, they probably never will. Based on the evidence, it’s logical; if click-through is the sole measurement of success, then enhancing CTR by limiting exposure to unique ads makes sense.

But since 1996 the primary objective for most online advertisers has shifted from click-through to branding. Traditional advertisers are using the Internet to influence the way people think and feel about their brands. Nonetheless, the practice of setting low frequency caps is still widespread.

For most campaigns, the practice is misguided. Using normative data from dozens of online advertising campaigns, Dynamic Logic tested the relationship between branding effectiveness and frequency. We found that online advertising’s ability to influence brand awareness increased with each subsequent exposure to an advertisement. In fact, the branding effectiveness of one exposure to an ad nearly doubled after four or more exposures.

This makes perfect sense. For a message to hit home, you need to be exposed to it a number of times. Traditional advertisers know this well — that’s why you rarely see a television commercial only once.

But branding-only campaigns aren’t the only ones that might benefit from lifting frequency caps. Indeed, most campaigns have dual objectives: branding and direct response. And the role of frequency is an interesting lesson in the interplay between these two goals.

I talked to several experienced media planners, some through the popular Online Advertising Industry Old Timers list, regarding what they have learned about frequency’s role. Contradicting the DoubleClick study, several told me that they found a sharp increase in response after 9-11 ad exposures. One person told me that he found this to be the case most often for brands with low initial awareness, suggesting that multiple exposures have a branding effect that eventually culminates in a click and a sale.

In other words, higher frequencies correlate with branding, and branding can lead to higher response. If this is true, then low frequency caps can be counterproductive for all but the narrowest of direct response initiatives.

Of course, optimal frequency will depend on the type of product advertised, the level of awareness and consideration, and any interplay with offline media tactics. And capping at a certain level might always be appropriate — something is wrong if people are seeing your ad 600 times.

But like many tactics, low frequency caps should also be rethought in light of new evidence and online advertising’s evolving role.

Related reading

Vector set of user interface login and account registration form design. Isolated on blue background.