Why You Should Be Considering a Multi-touch Attribution Model, if You aren’t Already

Last-touch attribution is so last decade. The last engagement isn’t always the most significant, and there are so many newer models that take the whole customer journey into consideration. That’s why you need to learn about multi-touch attribution. 

You know how “you don’t know what you don’t know?” That’s especially true when it comes to your digital marketing campaigns. If you’re promoting content in social, investing in search, and targeting audiences with multi-channel display campaigns, how do you know what’s driving consumers to convert? Do you know which clicks has the greatest influence on that conversion?

Last-click attribution definitely isn’t the way; neither is first-touch attribution. Sure, both models tell you the first or last thing a person clicked on before they took action, but they don’t give you a full picture of what influenced them to finally do it. They may have seen or interacted with your ads six times before finally filling out that form. Multi-touch attribution is the only way to see how all those interactions factored into bringing you that new lead or customer – and which interactions mattered most.

There’s more than one type of multi-touch attribution, though, so let’s take a closer look at attribution modeling and some of the most common ways to approach it. 

• Linear attribution model: the simplest model, linear attribution gives equal credit to every touchpoint. It’s multi-touch, but it’s not the most accurate way to measure campaign effectiveness. You’ll see everything, and you’ll see which channel delivers the most conversions, but now how the other touch points factored in. You might be missing out on critical information that could impact how you’re investing your campaign dollars.
• Time-decay attribution model: In this model, each marketing touchpoint is considered, but receives less credit based on when it occurred. So, the first touchpoint receives the least credit, but the one that occurred just prior to conversion receives the most. Each point in between is credited according to when it occurred, from first to list. Many marketers prefer this model because, as analytics expert Avinash Kaushik puts it, “…it does seem to make sense that the further back a media touch point is … the less credit it should get. After all, if the touch points were magnificent, why did they not convert?” 
• U-shaped (or position-based) attribution model: A blend of first-touch and last-touch attribution, the U-shaped model assigns 40 percent of the credit to the first touch and 40 percent to the last touch. The remaining 20 percent are divided evenly amongst all the touches in between. Top web influencer, Neil Patel is a fan of this one. “This makes sense when you think about it. After all, the first touch point is the one that acquired the customer, and the last touch point is the one that converted them.”
• W-shaped attribution model: The W-shaped model takes five touch points into consideration, giving 30 percent of the credit to each of the first, third and final marketing touch points. The remaining 10 percent is split between the second and fourth touch points – hence, the “W” shape. While this model does give some credit to multiple touches, it may overlook the midpoint of the journey and the valuable interactions that may happen there. 
• Data-driven or custom attribution models: For some marketers, none of the models listed above will feel like an ideal fit. In that case, a data-driven or custom model may be the best option. Google defines it like this: a data-driven model “distributes credit for the conversion based on your past data for this conversion action. It's different from the other models, in that it uses your account's data to calculate the actual contribution of each interaction across the conversion path.” 

In other words, this model is unique to your campaigns – but you have to have accumulated enough data about your campaign performance for it to work. So, it’s really for more seasoned marketers and/or those with bigger advertising budgets. 

Ultimately, attribution modeling is not a one-size-fits-all affair. Different marketers have different preferences – and campaigns will have varying requirements. What’s right for one marketer may not be right for another, and whichever attribution model you choose, it must fit your needs and goals. In general, you’re probably hoping to find out which elements of your campaigns are most effective, and where you should be investing more of your budget and energy. That may require experimenting with a few different models. Over time, you’ll figure out which one works best for your needs – and that will help you drive more ROI for every campaign dollar you invest.