What Is an Attribution Model and What Kind Should You Use?
Attribution models for tracking digital marketing results
If you want to drive results for your marketing, you must know how your campaigns perform. Campaign monitoring tells you which areas need more work and which ones are going strong.
One way to see how your campaigns are performing is to look at key performance indicators (KPIs) like conversion rate. But those raw metrics don’t always give you the full picture, which is why you should use attribution modeling.
What is an attribution model, though, and why should you use one? Keep reading to find out and see an overview of different attribution model types. Then subscribe to Revenue Weekly — our email newsletter — for more pro digital marketing tips!
Hear from HydroWorx, who saw a 236% increase in organic sessions with WebFX services.
WebFX is a partner businesses trust.
Hear from HydroWorx, who saw a 236% increase in organic sessions with WebFX services.Watch the Video Testimonial
What is an attribution model?
A marketing attribution model is a tool for helping you determine which of your marketing efforts receive credit for the conversions they help to drive.
Customers rarely convert after encountering a single marketing material. More often than not, they’ll run across several of your campaigns before deciding to convert. So, the question at that point is, which of those campaigns receive credit for that conversion, and how is that credit divided?
That’s the question attribution modeling helps you answer. Different models give credit to different strategies depending on how early or late users encounter them, and each model has its own advantages and disadvantages.
Why use attribution modeling?
You might be wondering why marketing attribution models are so useful. The answer is that they let you see which campaigns benefit you the most.
Let’s say a user first encounters your business by running across one of your paid ads. Later, they find one of your blog posts in organic search results. Then they receive an email from you. Finally, they directly visit your website to make a purchase.
In the above scenario, three different strategies played into converting the user. Does that mean they all share equal credit for the conversion? Or was it mainly one specific strategy that drove the purchase?
You can use marketing attribution models to help you answer that question. For instance, maybe you think the first marketing interaction is the most important. In that case, your attribution model for the above example would give the most credit to the paid ad.
You can then look at the broader trends that come from those observations to determine which strategies help your company the most.
Types of marketing attribution models
There’s no single attribution model that your business should use. On the contrary, several different models exist on the market, with each looking at attributions from a different perspective.
Here are some of the most common marketing attribution models to consider.
1. Last Interaction model
In the Last Interaction attribution model, the focus centers on the very last marketing material the user encounters before converting.
So, let’s say someone encountered three marketing materials, and converted after the third. With the Last Interaction model, only the third and final marketing material receives credit for that conversion.
This model ignores any influence the earlier materials had. Since the final interaction was the one that directly drove the conversion, it’s the one that receives the credit.
2. Last Non-Direct Click model
The Last Non-Direct Click model is similar to the Last Interaction model. The key difference lies in the exclusion of direct traffic.
Direct traffic refers to any instance in which a user goes straight to your website rather than going through one of your marketing campaigns. The Last Non-Direct Click model disregards any direct traffic since that traffic isn’t tied to a particular marketing campaign.
So, maybe someone reads a blog post, sees a social media ad, and then directly visits your site to convert. In that instance, the direct visit to your website was technically the last encounter. But in this model, you would disregard that, meaning the second-to-last interaction — the social media ad — would receive all the credit.
3. First Interaction model
The First Interaction model is essentially the opposite of the Last Interaction model, as the names suggest. In this model, the first marketing material a user encounters receives the credit for the conversion.
So, let’s say someone encounters a blog post, an email, and a paid ad, in that order. The blog post receives all the credit if you use the First Interaction model.
The idea behind this model is that the first interaction is the one that established the initial connection between the user and your company, and the subsequent materials were simply nurturing the lead that the first one created.
Therefore, according to this model, the first interaction is the primary contributor.
4. Linear model
The Linear attribution model distributes credit evenly across all interactions. All the other models we’ve looked at so far assign 100% of the credit to a single interaction, but the Linear model takes a different approach.
If users go through five different campaigns before they convert, the Linear model gives all five campaigns equal credit. The idea is that since they all contributed in some way, they all share the credit.
This model is useful because it avoids ignoring the contributions of the campaigns in the middle. Since this model gives equal credit to all campaigns, some campaigns may receive credit though they didn’t impact the customers decision to convert.
5. Position-Based model
The Position-Based model is a combination of all the previous models on this list. It splits the credit primarily between the first and last interactions, with the remaining interactions in between receiving smaller amounts.
Generally, the split is 40-20-40 — 40% of the credit goes to the first interaction, 40% goes to the last interaction, and the remaining 20% is divided evenly among the remaining ones.
This way you can acknowledge the role of the first interaction in beginning the customer journey, the role of the last one in inciting the conversion, and any contributions that the other interactions made.
6. Time Decay model
The Time Decay attribution model takes a similar approach to the Position-Based model, distributing credit in differing amounts to all the interactions. Essentially, the closer an interaction is to the conversion, the more credit it gets.
So, let’s say a user goes through four interactions: A blog post, a paid ad, a social media post, and an email.
As the last interaction before the conversion, the email would get the most credit. The social media post would get slightly less, and then less again for the paid ad. Finally, the blog post would receive the least credit, being the furthest from the conversion.
7. Custom models
The six attribution models listed above are the primary ones used in the marketing world. However, you’re not limited solely to those options. If you feel that something else would work best for your business, develop a customer model for your business.
To start, you might choose to use more than one of the above models together. Maybe you use the First Interaction model to evaluate which of your campaigns work best for top-of-funnel marketing, and you simultaneously use the Last Interaction model to see which ones work best for bottom-of-funnel marketing.
You can also create entirely new marketing attribution models if you prefer to divide credit in a way not shown in any of the already-existing models.
Where can you view and track attribution models?
There are a few different ways to track your campaigns and apply your attribution models. One of the best places is Google Analytics. In addition to tracking traffic to your website, you can also apply attribution models to more see which campaigns drive the most conversions.
Alternatively, you can choose from the many third-party tools that come with an attribution modeling feature. Some examples include:
“Working with WebFX, everything is very organized and strategic. We’re only a year into our partnership, and our expectations have definitely been met.”
93% of WebFX customers are extremely satisfied with their digital marketing results.
“Working with WebFX, everything is very organized and strategic. We’re only a year into our partnership, and our expectations have definitely been met.”View More Client Testimonials
WebFX can boost your marketing attribution strategy
Need a hand creating and applying your attribution models? WebFX would love to help! We’ve been optimizing digital marketing campaigns for over 25 years, so we know how to track results and drive conversions.
With our marketing analytics services, you can gain deep insights from your marketing data and use it to improve your campaigns’ performance. Our services will ultimately allow you to boost the revenue your company earns to spark growth online.