Learning Hub | E-Commerce
Attribution Modeling: Which Touchpoint is Your Star Player?
May 20, 2021 | Jon Teodoro
In team sports, it’s usually very clear who gets credit for the score: the person who touches the ball last.
Not so when it comes to converting consumers.
Who’s to say that an Instagram ad viewed by someone weeks ago isn’t more responsible for their purchase than the organic search that later helps them find that item?
That’s where attribution modeling comes in. Marketers need a way to dependably trace the purchase back to a consumer touchpoint in order to optimize their campaigns.
Attribution modeling offers flexibility through five primary models, which means you can give credit to different points in the chain depending on your preference.
Participation trophies are not a bad thing
Let’s go back to team sports. Every play—pass, block, rebound, deflection, free kick, at-bat, etc.—contributes to that moment when the team scores.
They all wind up on the stat sheet, too. Cumulatively, they paint a picture of how the team is functioning as a whole. Individually, these stats allow for a very specific understanding of areas the team may need to work on and where they’re excelling.
Attribution modeling lets you pick out the key players in your digital marketing and ecommerce strategy to assess their performance. It allows you to see conversions touchpoint by touchpoint, or play by play.
These are the five primary attribution models that marketers use to understand conversion. There’s something to be said for all.
First Interaction: The Moment You Gain A Fan
This model leans on that first moment of exposure to your brand. If a consumer finds you on Facebook, Facebook gets the credit for a future sale. If they find you through Adwords, Adwords gets the win, even if there are multiple touchpoints that follow before they buy your product.
Because this model leaves out additional customer interactions before the final purchase, it’s most often used for quick conversion cycles. It’s a good way to measure how to build new awareness for your brand.
Last Interaction: Last Touch Gets All the Credit
This one’s a no-brainer. If you catch the ball in the end zone, it’s your touchdown. If a customer gets targeted with an Instagram ad and clicks shop now to make a purchase, that ad gets all the credit.
In the case that there aren’t too many touchpoints, Last Interaction is an easy model to use and understand. But similarly to First Interaction, this model ignores all the other touchpoints leading up to the final click. It puts all the weight behind the final play.
Last Non-Direct Click: The Assist
What is the very last thing the customer did before they came to your site and made a purchase? Did they see a banner ad or maybe sponsored email content? That touchpoint gets full credit.
While actively going to your ecomm site to buy something is clearly the final play, it’s the “assist” leading up to that action that gets the nod here. Direct traffic does not get credit.
Last Non-Direct Click helps identify the strength of different tactics or channels outside of your ecomm store, but it’s another model that’s limited somewhat by attributing full credit to a single consumer touchpoint.
Linear: Equal Credit To The Whole Team
You might call this one the “Ted Lasso” of attribution models. Meaning, markets choose this model like a coach who chooses to reward everyone equally. Each contribution to the final purchase is valued on the same level.
Linear attribution takes into account the range of touchpoints that are in play, offering a much more holistic view of your strategy than the three models above.
It’s the ideal model for those who want to understand different touchpoints in relation to each other and see the customer’s entire journey from first click to last. The downside here is that you don’t have an eye into which channels are performing better than others.
Time-Decay: Respect to the Team, But Spoils to the Scorer
Time-Decay attribution is basically a reflection of how we experience fandom: we love our teams, but we especially love the players who win it all.
This model understands that there’s value in each customer interaction but assigns the most value to the final touchpoint. A sliding scale of credit, as it were.
It can take time to win customers over before they’re willing to buy from your brand—especially if the item is on the pricey side—and this model acknowledges that process while emphasizing the touchpoint closest to the moment of purchase.
One Winning Season Isn’t Always Like the Next
If attribution models are an inexact science, they help marketers test and learn using different goals. They are varied and adaptable. Different campaigns may call for different models, and the ability to shift models to meet the moment should be a priority for any marketing team.
No game is won the exact same way as another.