As a marketer, you know that the customer journey is no longer a straight line from awareness to purchase. In today‘s digital world, people interact with your brand across multiple touchpoints – they might see your ad on social media, research your products on your website, sign up for your email list, compare reviews on third-party sites, and more before finally converting.
So how do you determine which of those touchpoints had the greatest impact on driving the sale? That‘s the key question multi-touch attribution seeks to answer. Let‘s dive into exactly what multi-touch attribution is, why it‘s important, and how you can start leveraging it in your marketing strategy.
What is Multi-Touch Attribution?
First, let‘s define what we mean by multi-touch attribution. Here‘s the simplest way to explain it:
Multi-touch attribution is the process of assigning fractional credit to each marketing touchpoint along the customer journey that led to a conversion.
Rather than giving all the credit to a single interaction, like the first or last touchpoint before the sale, multi-touch attribution recognizes that multiple touchpoints likely influenced the customer‘s decision.
Image Source: CallRail
Think of it like a basketball game. While the player who makes the final shot might get the glory, that basket resulted from the team‘s collective effort in passing, screening, and moving the ball down the court. Multi-touch attribution is like the box score for your marketing – showing how your various "players" or touchpoints each contributed to the win.
Multi-Touch vs Single-Touch Attribution
You might be familiar with single-touch attribution models like first-touch or last-touch. As the names suggest, these give 100% of the conversion credit to either the first or last interaction before the conversion.
While these models are easy to understand and implement, they don‘t reflect the nuances of the modern customer journey. With people engaging with brands across so many touchpoints, it‘s overly simplistic to assign all the value to a single one and ignore everything that happened in between.
Imagine a customer journey that looks like this:
- Saw your display ad on a news website
- Later searched for your product category on Google and clicked on your search ad
- Visited your website directly a few days later
- Clicked a retargeting ad on social media
- Returned directly to your site to make a purchase
In this case, first-touch would give all the credit to the display ad, while last-touch would attribute the conversion entirely to the direct visit right before purchase. Neither tells the complete story.
Multi-touch attribution aims to solve for this by assigning fractional credit to multiple relevant touchpoints along the path to conversion. Exactly how it splits up that credit depends on the specific multi-touch model you use.
Common Multi-Touch Attribution Models and Pros & Cons
There are several common ways multi-touch attribution models assign credit to touchpoints. Let‘s break each one down, along with their pros and cons.
Linear Attribution
- Assigns equal credit to each touchpoint
- Pros: Easy to understand and implement, gives a high-level view of the customer journey
- Cons: Doesn‘t account for the potentially outsized importance of certain touchpoints like the first or last interaction
Time Decay Attribution
- Assigns more credit to touchpoints closer in time to the conversion
- Pros: Reflects the idea that interactions right before conversion likely had more direct influence
- Cons: Might undervalue early touchpoints that generated initial awareness and interest
U-Shaped Attribution
- Assigns 40% credit each to the first and last touchpoints, with the remaining 20% split between the middle interactions
- Pros: Recognizes the key roles of the first touchpoint in starting the journey and the last in closing the deal
- Cons: Might oversimplify the importance of middle touchpoints
W-Shaped Attribution
- Variation of U-shaped that also gives extra 30% credit to the touchpoint that generated the lead, with first touch 30%, last touch 30%, lead creation 30%, and 10% for other touches
- Pros: Useful for businesses with a defined lead generation stage in the funnel
- Cons: Isn‘t applicable to every business model, may not be enough middle touchpoints getting credit
Custom or Algorithmic Attribution
- Uses machine learning and historical data to create a custom model for assigning credit
- Pros: Leverages real data to create the most accurate model for your unique customer journey
- Cons: Requires significant data and resources to implement
Here‘s a visual of how credit gets assigned in each model:
Image Source: ListenUp
While there‘s no one-size-fits-all solution, the key is to choose the model that best reflects your customer journey and business goals.
Implementing Multi-Touch Attribution
Now that you understand the different models, how do you actually put multi-touch attribution into practice? The high-level process looks like this:
- Collect data on all touchpoints (ad platforms, website, CRM, etc)
- Connect touchpoints to unique user journeys
- Apply your chosen attribution model to assign credit
- Measure impact and ROI using cost data
- Evaluate reports and extract actionable insights
Let‘s dive deeper into a few of these key steps.
Collecting Marketing Touchpoint Data
The foundation of multi-touch attribution is having robust data on all your marketing touchpoints. This can include:
- Ad impressions and clicks from platforms like Google, Facebook, LinkedIn
- Website interactions tracked by Google Analytics or Adobe Analytics
- Email engagement from your marketing automation or ESP
- Social media interactions and referral traffic
- Offline touchpoints like events, direct mail, or sales calls
The key is having consistent methods for tracking the source of each interaction. Using standardized UTM parameters for your campaigns, setting up proper conversion tracking, and having unique identifiers like user IDs to tie actions to specific individual users is critical.
Identity Resolution
One of the biggest challenges in multi-touch attribution is identity resolution – connecting all those disparate touchpoints across different devices and channels to a single user‘s journey. Techniques like deterministic and probabilistic matching can help unite interactions that might be coming from a user‘s phone, laptop, and offline actions.
The goal is to build a complete picture of the unique paths users took to converting, even if those paths zigzagged across multiple sessions and surfaces. Having a system for creating persistent user profiles is key to enabling this.
Turning Data Into Action
Collecting the data and running it through an attribution model is just the first step – the real value comes from leveraging those insights to optimize your marketing strategy. Multi-touch attribution tools typically provide reporting dashboards and visualizations to help you answer questions like:
- Which channels and campaigns are driving the most conversions?
- What does the typical path to conversion look like?
- Which touchpoints tend to play an "assist" role vs a "closer" role?
- Where are there points of friction or drop-off in the customer journey?
- What‘s the ROI of different channels based on attributed conversion value?
By looking not just at individual touchpoints but at the common sequences and patterns of interactions, you can start to glean deeper insights into how your marketing efforts are working together. This can inform decisions around budget allocation, campaign planning, audience targeting, and messaging strategy.
Real-World Multi-Touch Attribution Examples
To help illustrate how companies are using multi-touch attribution, let‘s look at a couple examples.
Airbnb
Airbnb has a notoriously complex customer journey – people might start by just browsing listings, coming back multiple times before creating an account, saving properties to wish lists, and finally booking. And those interactions could span mobile app, mobile web, and desktop.
Using multi-touch attribution, Airbnb is able to connect those cross-platform touchpoints and identify key assist interactions that get users to book, like saving a wish list. They can run incrementality tests to see how channels impact bookings and optimize budget accordingly. (Source)
Grubhub
Food delivery service Grubhub found that existing attribution solutions didn‘t work for their business model – diners would often open the app, browse around, and come back later to place an order. The "last click" approach was giving affiliates too much credit and not enough to upper funnel efforts.
Using a custom multi-touch model in their internal BI system, Grubhub was able to more accurately distribute attribution credit. For example, they found that emails were a key driver of first orders and repeat purchases – insights they used to optimize their spend and increase revenue. (Source)
The Future of Multi-Touch Attribution
While multi-touch attribution has come a long way, it‘s still an evolving field with plenty of room for innovation. Some key trends and predictions for the future:
- Increased adoption of data-driven attribution using machine learning models
- Heavier use of incrementality testing and experimentation to validate models
- Advances in cross-device tracking and identity resolution
- Integration of offline and online attribution
- Shift from session-based to person-based attribution
- More focus on predictive and prescriptive analytics to guide strategy
- Use of multi-touch attribution for customer lifetime value, not just single conversions
As the customer journey continues to fragment across devices and channels, the importance of having an advanced multi-touch attribution approach will only grow. Brands that are able to crack the code on understanding the nuances of their conversion paths and using those learnings to fine-tune their marketing will have a major edge.
Getting Started With Multi-Touch Attribution
If you‘re convinced of the value of multi-touch attribution but not sure where to start, here‘s a quick roadmap:
- Audit your current attribution practices and identify gaps in tracking, data integration, and analysis
- Outline your key business questions and use cases for attribution
- Evaluate different attribution tools and models
- Start with a simple model like linear or U-shaped before advancing to more complex algorithmic approaches
- Ensure tracking is in place across all relevant marketing channels
- Test your model on a segment of data and validate results
- Roll out to more campaigns and teams, evangelizing insights and best practices
The key is to crawl before you walk – get the basics in place, prove value with real results, and then scale your practice from there. Don‘t let perfect be the enemy of good when it comes to attribution.
Conclusion
We‘ve covered a lot of ground in this guide to multi-touch attribution – from the fundamentals of what it is and why it matters, to the nuances of different models, to real-world examples and implementation tips. Hopefully you‘re coming away with a solid understanding of the concepts and some concrete next steps to begin applying them in your own marketing.
As a parting thought, it‘s worth emphasizing that multi-touch attribution is ultimately a means to an end – driving better marketing decisions and business outcomes. The fancy models and shiny dashboards only matter if you‘re able to extract actionable insights and optimizations from them. Don‘t lose sight of the big picture.
The marketers who will thrive in the years to come will be the ones who can harness the power of data to deeply understand the customer journey and relentlessly improve the path to conversion. Multi-touch attribution is an essential tool in that pursuit – one that‘s well worth investing the time and resources to master.
