What is multi-touch attribution?
Multi-touch attribution is the process of tracking and assigning fractional credit to marketing touchpoints along the path to conversion. Unlike multi-channel reporting, multi-touch attribution reporting is not limited to reporting on a single dimension and can encompass several dimensions, such as channel (e.g. source, medium, campaign), content (e.g. landing pages, offer content), and even device parameters (e.g. device type, browser, location.)
Marketing and sales teams use multi-touch attribution data to better understand both the chronology and the type of interactions that precede and influence conversions. They then use multi-touch reporting to optimize the conversion paths of prospects.
What is the difference between multi-touch and multi-channel attribution?
They’re quite similar; however, multi-touch attribution is touchpoint based (and much more comprehensive). Multi-channel attribution is used for understanding one touchpoint dimension (channel). While many people refer to them interchangeably, there are important distinctions in terms of reporting capability and assigning credit.
In a strictly multi-channel attribution model, channels like Paid Search, Social, Organic Search, Web Referral, etc. get fractional (or weighted) credit for the conversion–regardless of the chronology, content, or types of interactions that preceded the conversion. A multi-touch attribution model is able to report on channels but it would include much more data on those individual touchpoints.
It’s important to realize that while most multi-touch attribution models involve multiple channels, some may only incorporate one channel. A good example of this is when outbound sales teams want to understand how outbound phone calls (touchpoints from a single channel) influence sales conversions. A multi-touch attribution model would assign fractional credit to each outbound phone call for a sales conversion. Data collected from those phone calls (salesperson, timestamp, duration, keywords mentioned, etc.) would be used to optimize the outbound phone call channel.
In order to get started with multi-touch attribution modeling and reporting, it’s necessary to be able to tie an identity to each marketing touchpoint. In marketing attribution modeling this process is called identity resolution.
What is identity resolution and why is it necessary for multi-touch attribution?
Identity resolution is the process of assigning, capturing, and joining datasets based on a unique identifier. Without an identity resolution system, marketers have no way to associate a website visit (touchpoint one) made by Sally on Monday with a subsequent inbound phone call (touchpoint two) by Sally on Friday.
Most attribution marketers use an existing analytics or other marketing tool’s unique identifier for tying together multiple touchpoints from the same person. For example, Google Analytics assigns a Client-ID to each website visitor. This GA Client-ID is then retrieved and captured by other applications on a website that capture leads (e.g. Unbounce form submissions with a hidden landing page form field, CallRail inbound phone calls with custom cookie capture, etc.)
Essentially, identity resolution is the narrative glue for multi-touch attribution.
How can marketers use multi-touch attribution data?
Let’s assume you’re using multi-touch attribution methods to understand the types of content on your website prospects interact with before a conversion event. You begin noticing that there is a relationship between the chronology of content interactions and types of content. You determine that the chronology and content interactions break down into three stages of the prospective customer’s journey.
Using Google Ads ( formerly AdWords) as an example, you could very well have three different campaigns for the same product that represent different stages of the prospective customer’s journey. The first is general research, the second is competitor/specific market research and the third is purchasing intent.
While a single touch attribution model (first touch/click or last touch/click) would only give credit to one stage of the customer journey (the research stage or purchasing intent stage), your multi-touch attribution model would then assign fractional credit to each Ads campaign.
This kind of information gives you clearer insight into how your keywords are working for you in your account. Using the default last-click model you’d think that money spent on clicks in the prior two campaigns were wasted while using a multi-touch model you realize that they were crucial in contributing to the final conversion.
How to track multi-touch attribution
In our blog on multi-channel attribution, we discussed the different attribution models, each of which track the different touchpoints on the customer journey differently.
The time decay model in Google Ads (Google AdWords), for example, increasingly gives more credit to touchpoints that happen closer to the final conversion event, based on a predetermined “half-life.” If for example, you set your half-life at 14 days, interactions that happened 14 days prior to the final conversion event, would receive half credit.
As competition steadily increases it becomes more and more important to track each interaction along the customer journey as they contribute to conversions. As you become more scientific about touchpoint data analysis, you can extract insights from all the deep metrics that are trackable on each touchpoint.
Whether it’s a form filled out on your website, an app download, a call to your business, or even a customer purchase, knowing which interactions, as well as the order they happened, actually preceded the conversion event will transform the way you market and help prove your ROI.
- First Touch Model
- Qualified Lead Model (Last Touch)
- Lead Creation Model
- Last Non-Direct Click Attribution
- Last < most important touch > attribution model
- Linear Attribution
- Time Decay Attribution
- U-Shaped Attribution
- W-Shaped Attribution
- Z-Shaped Attribution