Attribution is a crucial part of any marketing strategy. It helps you understand which elements of your marketing mix are driving the most impact and which ones are not.
Multi-touch attribution takes this a step further by providing a holistic view of how different touchpoints or interactions with your brand lead to the ultimate conversion.
What is multi-touch attribution?
Multi-touch attribution (MTA) is a technique marketers use to assess the efficacy of their marketing assets and campaigns. It involves tracking and measuring the influence of each marketing component a customer engages with before purchasing or taking any desired action.
Traditional attribution models tend to credit only the first or last touchpoint before a conversion, ignoring all other interactions that may have played a role in influencing the customer’s decision. MTA takes into account all touchpoints across various channels, giving a more accurate understanding of the entire customer journey.
How does multi-touch attribution work?
When you’re evaluating the ROI of a marketing campaign, you’re going to look at money in vs. money back. However, as we know, customer journeys are not always straightforward and linear.
With MTA, you’re able to assign credit for conversions across multiple touchpoints based on predetermined rules or algorithms. In the context of ROI, that means your revenue attribution algorithm will assign a percentage of the revenue generated to each touchpoint, based on its influence or impact in the customer’s journey.
You can either set the amount of credit for each touchpoint manually, based on your own understanding and experience, or you can use complex data-driven algorithms to calculate the weight of each touchpoint in the conversion process.
Or, you can use a model like W-shaped attribution, where credit is attributed to the first and last touchpoints, as well as any touchpoints in between at a predetermined percentage.
Types of multi-touch attribution models
There are several models marketers use for multi-touch attribution, each with a different way of assigning value to touchpoints.
- Linear model: Assigns equal credit to all touchpoints.
- Time decay model: Gives more credit to touchpoints closer in time to the conversion.
- U-shaped or position-based model: Assigns 40% credit to the first and last touchpoints, with the remaining 20% credit distributed evenly among the other touchpoints.
- W-shaped model: Similar to the U-shaped model, it also gives significant credit to the touchpoint where a lead is generated.
- Custom or algorithmic model: Uses algorithms and machine learning to assign credit based on data and patterns specific to a particular business.
The more complex your business and sales process, the further down this list you’ll go. Linear and time decay models work well for DTC brands, e-commerce stores, and basic subscription-based services (B2C SaaS). However, if you have a longer and more complex sales cycle (e.g., B2B SaaS) with multiple touchpoints, a W-shaped or algorithmic attribution model is the only one to consider.
MTA in practice: a real-world example
Let’s consider a real-world example to illustrate how multi-touch attribution works. Let’s say an online clothing store is launching a marketing campaign for a new line of summer dresses.
Here’s what the customer journey might look like:
- Initial awareness (paid social media ad): A customer, Sarah, first sees an ad for the dresses on a social media platform. She clicks on the ad and browses the website but doesn’t make a purchase.
- Follow-up engagement (email marketing): A week later, Sarah receives an email from the store showcasing their summer collection, including the dresses. She opens the email, clicks through to the website, but again leaves without buying.
- Retargeting (display ad): Later, while reading a blog, Sarah sees a display ad for the same dresses as part of a retargeting effort. This time, she clicks on the ad, returns to the store’s website, and finally makes a purchase.
When we’re applying multi-touch attribution, it’ll look different depending on which method you apply.
Let’s apply the different multi-touch attribution models to the $100 order in the online clothing store example. Each model divides the $100 differently among the touchpoints (social media ad, email, and display ad).
Linear model
This model assumes each step was equally important in Sarah’s decision-making process. So, in our example with three touchpoints, each touchpoint would be attributed $33.33 ($100 divided by 3).
Time decay model
More credit is given to touchpoints closer to the purchase. Let’s say the model gives 50% weight to the last touchpoint (display ad), 30% to the second last (email), and 20% to the first (social media ad). The distribution would be:
- Social media ad: $20 (20% of $100)
- Email: $30 (30% of $100)
- Display ad: $50 (50% of $100)
U-shaped model
This model typically assigns 40% of the credit to both the first and last touchpoint and the remaining 20% to the middle touchpoint. The attribution would look like:
- Social media ad: $40 (40% of $100)
- Email: $20 (20% of $100)
- Display ad: $40 (40% of $100)
In the above example, the store learns that while social media is good for initial awareness, their email marketing effectively nurtures interest, and display ads are crucial in closing sales. Based on these insights, the store could allocate more budget to retargeting ads while still utilizing social media for awareness.
Final thoughts
As you can see, multi-touch attribution is a powerful tool for understanding the effectiveness of your marketing efforts and making data-driven decisions to improve ROI. By assigning credit to each touchpoint in a customer’s journey, you can gain valuable insights into which channels and strategies are most effective at driving conversions.
