Select the right attribution model for your business

By Brenna Lofquist, Senior Marketing Consultant / Client Services Operations at Heinz Marketing

Attribution models always seem to be a hot topic. What’s the best attribution model? How do you implement it? Do you need an attribution tool/platform? I would imagine most companies have asked these questions at some point. If it were easy everyone would be doing it, right? Well let’s see if we can make it easier, at least.

First lets briefly chat about the different models.

Types of Attribution Models

First-touch

• Gives all credit to the first touchpoint that the customer interacts with before buying

Last-touch

• Gives all credit to the last touchpoint that the customer interacts with before buying

Multi-touch

• Distributes credit to multiple touchpoints throughout the customer’s journey
• There are a handful of attribution models that can be considered multi-touch, including:
  • Linear: Gives equal credit to all touchpoints that a customer interacts with before buying
  • U-shaped: There’s a few versions of this model floating around which include:
    • Equal credit to the first-touch and lead-creation touchpoint
    • More credit to the first- and last-touchpoints, and less credit to the middle touchpoints
    • Equal credit to the first-and last-touchpoints
  • W-shaped: Counts three milestones; first-touch, last-touch, and opportunity creation (OC) where each receives 30% of the credit and the remaining 10% is distributed evenly to the rest of the touchpoints
  • Full Path: Includes four milestone touchpoints; first-touch, last-touch, OC, and Close Won (CW), where each are given 22.5% of the credit and the remaining 10% is distributed evenly to the rest of the touchpoints

Check out this great Marketo Measure resource for a visual of a few of the above attribution models.

Self-Reported

• Relies on the customer or prospect to self-report how they heard about a business

Custom

• Custom attribution models range from hybrid models to a combination of models
• Custom models allow organizations to choose which touchpoints or custom stages to include in the model
• Organizations can also control the percentages given to different touchpoints

How to determine the right model for your business

The next step is to figure out which model is right for your business goals and objectives. Follow the steps below to get started.

1. Audit your marketing efforts and channels to see how many touchpoints you have across your funnel
2. Set clear and specific goals such as increasing conversions, revenue, or ROI
3. Map your customer journey and identify the stages, channels, and touchpoints that are most relevant for your target audience
4. Incorporate lead scoring to assign values to leads based on their behavior, engagement, and demographics
5. Invest in the right tools that can collect, integrate, and analyze data from multiple sources and platforms

And as always keep it simple! You can always optimize and improve your attribution model however, I would always recommend starting simple before unnecessarily overcomplicating it right out of the gate. Figuring out an attribution model for your business is no easy feat, it will be a process. If you think this is one of those set it and forget it type things, you are sorely wrong.

A few things to look out for

Nothing is perfect, so there are a few things to keep in mind to make sure your attribution model is successful.

• Make sure you’re able to extract actionable insights. Setting up an attribution model is one thing but you still need to be able to analyze the results to make decisions, and if you can’t then there’s a problem
• Watch for duplicate counts or miscounts – these happen more frequently than you’d think
• Most models are unable to account for immeasurable or offline touchpoints such as word of mouth, communities, etc. (although self-reported attribution does a better job of this)
• Certain models offer very limited optimization ability or can be misguiding if you’re relying on a single touchpoint

So, which model will you choose? Let us know in the comments which attribution model you’ve had success with!