The great thing about investing in stocks is that it’s very easy to know how well you’re performing: if the value of your investment goes up, you’re doing well; if it goes down, you’re doing poorly.
In online marketing, this clarity of outcome doesn’t exist. If you run a $100K advertising campaign and your sales double, can you claim with certainty that your ad campaign drove those incremental sales? Even though Real-Time Bidding (RTB) is supposed to make digital advertising hyper-efficient and optimizable, ambiguity around the attribution of credit (or blame) for a campaign’s outcome remains.
To address this ambiguity, two main questions need to be answered by marketers.
1) Which attribution methodologies should I use? The three most common answers include:
- Last-Click Attribution: Here you’re giving credit to the last impression that resulted in a click prior to a conversion. This is the classic form of online attribution (think AdWords).
- View-Through / Last-Touch Attribution: Instead of relying on just a click, credit is given to the last impression that was served to a user prior to a conversion.
- Fractional Attribution: Rather than giving credit to any individual impression, Fractional Attribution algorithmically gives credit across all of the impressions that were served to a user prior to a conversion.
Click-based attribution initially got the most adoption, primarily because of AdWords. However, over the last 3-5 years, marketers have increasingly been adopting last touch or fractional as their attribution methodology. We’re now at a point where most marketers agree that some form of non-click based attribution makes sense. This acceptance comes from both a qualitative perspective (think of billboards and TV where there’s no option to click), and a quantitative perspective (numerous studies, white papers and campaign tests that have statistically proven the virtue of non-click attribution).
However, moving away from click-based attribution brings about many questions. How should credit be allocated outside of clicks? Is post-view right for all campaign types — from branding to direct response (DR) to affiliate? How does viewability tie into this? Should post-view credit be given if an ad wasn’t actually seen (the answer is no)? There aren’t perfect answers to these questions but they all merit conversation.
Now that you’ve made the decision to move away from click based attribution, it’s time to answer the second question that arises:
2) How do I coordinate my marketing team, agency, vendors and organization to effectively implement and judge post-view attribution?
You might think this isn’t a difficult challenge — after all, an attribution methodology has been chosen so that’s that. It’ll be used, measured and optimized just like anything else. But, in reality, this ideal outcome is far from reality. The case of a company jumping into non-click-based attribution only to be knocked back into it three to nine months down the line happens more often than you might think.
Have you tried / successfully moved away from click based attribution? What problems did you face?