As marketers continue to grapple with assessing every channel and marketing touchpoint, it’s important to note — attribution data needs context to be fully actionable. Whether you’re using rule-based attribution where all credit is assigned to a channel depending on its position in the conversion path, or data-driven attribution, where credit is fairly attributed to channels depending on how much it increased the probability of conversion, cost data adds this necessary context.
This is best explained through an example.
Let’s assume you’re already using an attribution solution (say, Google Attribution 360). You’ve gone through the vendor selection, implemented all the necessary tags for tracking, validated the model, and are ready to start using your insights.
The table you see in your attribution platform looks something like below:
Based on what you see here, you would probably think that the campaign that performed the best is Campaign A. It brought in 3 times as many conversions and revenue as Campaign B, and 15 times as many conversions and revenue as Campaign C! So looking at this, we would likely recommend shifting more of our advertising budget towards Campaign A.
But we aren’t seeing the complete picture. If you instead looked at the table with the associated campaign costs, you’d come to a different conclusion.
While Campaign A drove a large number of conversions, it came at a much higher cost than the other campaigns, and so had an Attributed Return on Ad Spend (attributed revenue/cost) of only 37 cents. Meanwhile, the other campaigns have much better ROAS: Campaign B had a $6 ROAS and Campaign C had a $3.50 ROAS. So now, we would recommend more advertising budget shifting to Campaign B.
From this example, it’s clear that to fully use our attribution findings we need to look at the complete picture. When you are set up with your attribution solution, whether it be first touch or last touch in Adobe Analytics or Google Analytics, or a dedicated data driven attribution solution like Google Attribution 360, keep this in mind and make sure your data is as comprehensive as possible.
