THATBlog

Attribution modeling is a complex topic, which is likely why most marketers favor a last non-direct click model - it's one of the simplest (and it's the standard model used for non-multi-channel funnels reports in Google Analytics). However, simple (and standard) might not be what's best for your business. It's important to consider your attribution model carefully to be sure you're making smart, data-driven marketing decisions about how to spend your budget.

Scratching your head about what an attribution model is? Here's how Google defines it:

An attribution model is the rule, or set of rules, that determines how credit for sales and conversions is assigned to touchpoints in conversion paths.

As for what a touchpoint is, SurveyMonkey explains:

Customer touchpoints are your brand's points of customer contact, from start to finish.

In the digital marketing world, a touchpoint can be anything from an email to a social media post or pay-per-click ad.

Make sense? If so, then let's take a look at some of the more popular attribution models out there and explore the benefits of each, so you can select the right marketing attribution model for your business.

Last-Interaction Attribution Modeling

Often referred to as last-click attribution modeling, last-interaction attribution modeling focuses on the last touchpoint encountered on a customer's conversion path. In other words, the channel directly responsible for getting the sale receives 100% credit.

This is the standard attribution model used in most web analytics tools. However, it is the least logical when you think about a real-life customer's conversion path. 

For example, a real-life customer's journey might look something like this:

Social Network > Direct > Organic Search > Referral > Direct

As you can see, the real-life customer encountered multiple touchpoints that created his/her path to conversion. 

In last-interaction attribution modeling, the direct channel would get all the credit. Silly, right? Especially when social media, organic search and referral were all involved in influencing the customer to convert.

Use this model only if you have no other attribution models available in your web analytics toolkit. However, with today's technology, the odds of that are slim to none.

Last Non-Direct Click Attribution Modeling

Last non-direct click attribution, Google Analytics' default model used for non-multi-channel funnels reports, gives full conversion credit to the last campaign prior to the conversion. Meaning, this model excludes direct traffic. So, in the scenario below, all the conversion credit would go to referral:

Social Network > Direct > Organic Search > Referral > Direct

Again, this attribution model seems a bit skewed. Clearly social and organic played a part in influencing the customer to covert. And why omit direct traffic? Why disregard a marketer's ability to deliver brand recognition?

Despite these flaws, last non-direct click attribution modeling has merit. For example, if your business has a super-quick conversion cycle (think online clothing sales), your customers are likely the most convinced (or the least convinced) by the last marketing offer they saw, so adopting last non-direct click attribution could work well for your data-driven marketing.

Last AdWords Click Attribution Modeling

Considered Google's sweetheart attribution model, last AdWords click attribution should be used to gauge the effectiveness of Google AdWords alone. That's because, regardless of at what point the click occurred on the customer's path to conversion, Google AdWords is given 100% credit.

First-Interaction Attribution Modeling

This model gives all the conversion credit to the first touchpoint. Use first-interaction attribution modeling to determine what channel initially drove a customer to your website, but has low conversion rates.

You can also use this attribution model when your brand is new and focused heavily on building awareness in the market. However, even when building a brand from scratch, you should look for ways to optimize landing pages to get more conversions.

Linear Attribution Modeling

Linear attribution modeling seems a bit more logical, as it allocates credit evenly across all touchpoints. This makes sense, because all the touchpoints along the customer journey contributed to the conversion. However, the flaw in this logic is whether or not all the touchpoints contributed equally.

Therefore, you should only use this attribution model if you're looking to measure overall channel performance as part of your data-driven marketing strategy.

Time Decay Attribution Modeling

In time decay attribution modeling, more credit is given to the touchpoints closest to the conversion point, but some credit is given to all the preceding touchpoints. Basically, this model uses the idea of exponential decay, assigning more value to the more recent interactions. 

Incorporating time into the equation seems like common sense. As marketers, we measure performance over specific periods of time. And if a particular touchpoint occurred early on, why didn't it convert?

This model is ideal if your business has a long conversion cycle. It also fits when you believe that each touchpoint is important, but that impressions fade over time.

Position-Based Attribution Modeling

If you value both the first impression and the last impression the most, then position-based attribution modeling is for you. This model assigns 40% of the conversion credit to the first and last touchpoint along the customer conversion path, and assigns the remaining 20% to all the other touchpoints. 

Use position-based attribution when you want to find out which channels bring in the most new customers and which channels convert at higher rates.

In closing, each of the more popular methods for attribution modeling exists for a reason. What's important is identifying the reason why you are choosing to use said model to assess your digital marketing efforts. By evaluating your customer conversion cycle and the goals of your business, you can select and employ the right attribution model for data-driven marketing that optimizes your budget.