Data and Frequency Capping Can Drive Better Media Buying: Q&A

Posted on by Richard H. Levey

Media planners have traditionally relied on measurements such as clicks and unique visitors to guide their online buys, says Adometry vice president of products John Dietz. But a combination of shifts in the metrics used, as well as frequency capping within and across ad networks, can significantly boost their programs’ effectiveness. Dietz discusses his theories and research with Chief Marketer.

CHIEF MARKETER: Why aren’t clicks and unique visitors appropriate metrics for media buyers?

DIETZ: A click is an intermediate event. A lot of time [what marketers want] is revenue, or lifetime value, or something reflecting branding and brand recognition. Something like clicks is not a great proxy for performance.

CHIEF MARKETER: So what should marketers seek?

DIETZ: Conversions—some kind of onsite action which they use as a stand-in for revenue. For someone who is taking online orders, it is easy. If you can match online activity with offline transactions, that is ideal as well. If you can’t, you find a reasonable proxy.

CHIEF MARKETER: Media buyers focus on getting traffic to that point where they can be converted in the most cost-effective way. What can you recommend?

DIETZ: There’s frequency capping, in which media buyers limit the number of times a prospect is exposed to an online ad. We’ve done a lot of research internally, and there is a diminishing return on frequency. Is it really helpful to have someone see the same ad 20 times when they have already seen it 15? This is one of the areas where a [display space buyer] can make a big difference without making a lot of effort – finding a way to put a cap on that.

Let’s say I am trying to drive conversions, and there’s a jump between those who convert after seeing an ad nine times and those who see it ten times. Analyzing frequency capping lets me calculate an incremental cost to each of those extra conversions. If I am paying a site an extra $10000 to get from nine exposures to ten [per tracked visitor], and am going from 400 to 500 conversions, that’s $100 a conversion. There is an incremental cost to net conversion gain, and I can compare to cost and see if it is worth it. That is how we display it in our product.

CHIEF MARKETER: When you consider frequency capping, do you include messages shown through different channels?

DIETZ: That’s a challenge from a marketing standpoint. When you apply a frequency cap – even just with online display ads – you can’t even build a frequency cap across, say, Google, Yahoo and MSN. A media buyer can go to Yahoo and say “I don’t want prospects to see the ad more than four times a day.” It doesn’t mean they haven’t seen the ad six times on Google.

It is something we do try to address in our attribution modeling. We take into account overlap between sites and channels. For the average user, there is 20% overlap between Google and Yahoo, and we take that into account when we generate placement recommendations. I don’t know that there is a whole lot more than that [that can be done]. You can only practically cap within a single publisher. For sites with a large amount of overlap we might make a lower frequency cap suggestion.

I will call someone stupid for not using some analytics system. There are analytics systems such as DoubleClick or Google Analytics that will tell you something about frequency capping [across networks].

CHIEF MARKETER: Assuming a media buyer can negotiate an advantageous enough rate, what’s the harm to saturating a market?

DIETZ: We have actually done analysis of retargeting ads [ads for products viewed on sites that follow a web surfer around the Internet]. Initially, the shape of the curve is that as frequency goes up, the propensity to convert goes up. But eventually it dips back down. Maybe people get annoyed. The data show that at a really high frequency the effect becomes negative. At that point it is not about wasting money, it is about giving someone a worse view of your brand.

If you have a high overlap between, say, one publisher or network, and we see a lot of overlap between your email campaign and users who are proceeding on that site or network, we can take that into account.

CHIEF MARKETER: How much targeting to individual browsers is possible?

DIETZ: Frequency capping is more about average. You can’t say “I want this person to see it more than this person.”


Related Posts

Chief Marketer Videos

by Chief Marketer Staff

In our latest Marketers on Fire LinkedIn Live, Anywhere Real Estate CMO Esther-Mireya Tejeda discusses consumer targeting strategies, the evolution of the CMO role and advice for aspiring C-suite marketers.


Call for entries now open