Untangling the Gordian Knot of Campaign Tracking, Part 4

When it comes to online campaign tracking, there’s a new buzz phrase: “attribution management.” In our business you know it’s a trend when you’re asked the same question twice in the same week. At the moment, the question seems to be: “How do I attribute sales to different marketing channels in a more sophisticated way?”

Increasingly, advertisers know the way they’re currently tracking campaigns and the return on marketing investment are sub-optimal. They know the current paradigm of “last click or impression gets the sale” means they’re potentially making poor decisions when it comes to fully evaluating the role of different online media in the marketing mix. There is the classic phrase in marketing: “I know that half my advertising spend is wasted, I just don’t know which half.” In the online era of almost total accountability, it seems we still struggle with this fundamental paradox.

The problem is simple. In the majority of cases, online advertisers use their campaign management or Web analytic systems to determine how they attribute sales (or other conversion events) to different online marketing channels. The current paradigm is that the channel that delivers the last click or the last impression before the visitor converts gets credited with the sale. If someone sees a banner ad, then does a search, clicks a sponsored ad, and then visits the Web site and converts, the investment in the sponsored search link gets full credit for that conversion.

Based on those results, you may take the view to reduce the investment in banner advertising and increase investment in search marketing, as that’s the channel that seems to be getting results. That’s all very well and good, but what about the fact the customer saw a number of banner ads over a period of time which raises their awareness of the brand and prompted them to be more susceptible to clicking on the sponsored link when they carried out a search?

The two different media channels do a different job; one raises awareness and consideration, the other generates a direct response. But only the direct response mechanism is credited with the “sale.” The danger is you then turn off investment in banner and display advertising because it doesn’t seem to do anything. The next thing you notice is the number of search generated conversions seems to be going down. We live in a multi-channel, multi-media world, and we can’t manage these things in silos.

The solution? That’s a touch more problematic. The key thing for advertisers to recognise is the current paradigm is sub-optimal. Whilst they may be focussing efforts on optimizing their marketing investment within a channel through, say, sophisticated bid management strategies, they may not be optimizing their investment across channels because they don’t have complete visibility on the dependencies of one channel on another.

The next question, then, is how does an advertiser recognize and quantify the indirect effects of some channels? The major challenge here is many common technologies used in campaign tracking and analysis aren’t up to the task. The default setting is generally to credit the last click or impression with the conversion event. In some cases, it’s possible with ad serving technologies to look at the “halo” effects of display advertising on search activity, but this isn’t a standard report and requires the display and search campaigns be managed from the same tool. In some Web analytic systems, it’s possible to define some basic attribution rules other than “last click wins,” such as the ability to spread credit across all channels a visitor touches before they buy. However, as we’ve discussed before, Web analytic systems cannot take into account the impression effects of display advertising.

According to a recent Jupiter report on this issue, the advertiser solution appears to be to turn to their agencies. Some agencies have developed analytic systems that operate off ad serving data. These systems allow for more flexible and sophisticated attribution management, but lock the advertiser into a specific approach. An alternative may be to bring the data in-house, but that requires the advertiser to handle and manipulate large volumes of data they may not be very comfortable with. Another possibility is to use a third party to collect, manage, and report on the data on your behalf.

They say recognizing a problem is half way to a solution. When it comes to attribution management, the trend is that advertisers are realizing the decisions they make on the basis of the data they have might be less than perfect. The challenge they face is the present solutions are somewhat limited. Hopefully, we’ll see technology providers in the ad serving, campaign management, and Web analytics spaces work to provide better ways of understanding the complexities of the online marketing mix.

Related reading