With Adobe’s recent acquisition of Marketo and Marketo’s acquisition of Bizible, the topic of attribution is quickly breaching past the early adopters and crossing into the minds of the early and late majority. What is it? Should you care?

To catch you up quickly, marketing attribution provides a solution to the 100-year old John Wannamaker quote, “half of my advertising is wasted…I just don’t know which half.”

In short, it identifies the user touch points that lead to a purchase, and assigning a value (i.e. “credit”) to each of those touch points.

This is BIG because most marketing up until recent has been first-touch attribution, as in the 1st marketing touch point has gotten all the credit for turning someone into a customer.

Think, “1st come, 1st credit.”

As you can imagine, this neglects the reality of almost all customer journeys since there were likely dozens of touch points (ads, emails, billboards, a conversation) that ultimately lead to the purchase.

Hence, this has spawned a whopping 10+ other models of how to fairly give credit to all marketing efforts, including last-touch attribution.

Here, think “last come, 1st credit.” Credit is given to the final touchpoint that ultimately got the customer to make the move and purchase.

The power of this approach is that very little time passes from the last touch to the “conversion,” but once again neglects all other marketing that went into “warming them up.”

So what’s a company to do?

Multi-touch attribution.

Considered the most “democratic, ” multi-touch attribution looks at the entire conversion journey from start to finish and gives equal credit to all touch points.

There are a few variations of this (time decay, W-shaped, full-path) but I believe the most effective is linear where credit is evenly applied all touch points.

Care to make the switch yourself? Go to your Google Analytics, click the Conversions tab and then the Model Comparison Tool to pick “Linear” from the attribution model pool.

That said, everyone’s different. Which attribution model have you been using? Do you have your own preference? Please share this with a friend or colleague if you find this valuable.

~JK