Few things in business are harder to win and keep than customer loyalty.
Online publishers often take for granted that their most loyal users’ frequent visits are an implied agreement that they will accept the same offers several times per session. This strains the tenuous bond between reader and publisher, and it tends to punish advertisers who bankroll the site.
Heavy ad deliveries may spur a quick rise in awareness, and may even stimulate sales, but soon enough they can result in high brand negatives. As for direct response rates, well, bombardment is usually an effective way to sink battleships, and it does the same to DR metrics.
Frequency is important to media planners, but so is limiting audience exposures to high volume ad campaigns in order to build effective reach over time. Ad server limitations typically restrict the publisher to single-campaign, daily frequency caps, so the frequent user may see fewer ads in a day, but he or she will see them every day, ad nauseum. So what’s a publisher to do?
One of the lesser-known benefits of behavioral targeting is its ability to help publishers manage the rate at which an advertiser can reach target audiences over time. There are at least two ways to do this: (1) recurring visit frequency segments and (2) real-time targeting.
Targeting segments of repeat visitors
With this method, the publisher creates segments of both repeat visitors and new visitors, and then targets them differently from each other and from one day to the next. Any segment you can create can be targeted through the ad server, so it’s a matter of determining how many segments of repeat visitors to create and for what time period.
Your behavioral targeting service can isolate loyal users and instruct the ad server to deliver -- or not -- based on their frequency of visits and on their prior exposures to multiple campaigns.
Let’s say we know the percentage of a website’s visitors on March 1 who are likely to visit four or more times each week this month. We’ve calculated that by capping them at three impressions per week for all of the advertiser’s campaigns -- regardless of their number of sessions per week -- we can deliver the advertiser’s goal of reaching 80 percent of the target audience ten times or less by March 25.
This is a far cry from the common scenario in which frequent visitors might have to endure one hundred impressions in order for the publisher to fulfill the order.
For a site with, say, more than 250,000 unique visitors per day, we might want to create audience segments that are automatically recalculated and exported to the ad server every day. This gives us greater control over the ad server’s frequency capping ability. Sites with less traffic may be better off doing this every nth day or weekly.
Real-time targeting
If your behavioral targeting service offers real-time targeting, you can use this feature to accomplish the same goal with less effort. Real-time targeting features continuous profile updating that makes it possible for the behavioral targeting service to tell the ad server instantly whether to deliver the ad -- because the visitor has not seen it three times this week yet, for example -- or to substitute another ad to avoid going over the limit.
Whether you use recurring segments or real-time targeting, when a visitor hits the prescribed limit of exposures -- in our example, three views of all of the advertiser’s campaigns in one week, regardless of number of sessions -- the server withholds what would have been the fourth impression and serves a different ad instead. As the campaign progresses, the publisher and advertiser can agree to dial exposures up or down according to adjustable segmentation criteria and requirements of the campaign.
All segmentation and targeting are configured at the start of the campaign, and reports are set to recur every day to monitor actual versus forecast. Once you’re in motion, it becomes routine.
Advertiser and publisher should negotiate ahead of time the method -- recurring segments or real-time targeting -- as well as the frequency caps and target audience cume, based on careful consideration of all factors, such as availability of targetable impressions by time period. Advertisers shouldn’t demand that eager publishers produce a smooth audience reach curve when everyone knows that meeting the campaign objectives will require oversaturating the site’s loyal visitors.
The result of following these techniques? More satisfied users who respond better, and advertisers who are more likely to return next month for more opportunities to reach a greater share of an audience that they value.
That’s the loyalty payoff in behavioral targeting: more relevant messages, less often, with better results and repeat business between publisher and advertiser.
Bennett Zucker is Vice President of Customer Success for TACODA, The Audience Company, in New York City.
