May 23, 2019 - By Ginna Hall, Senior Content Writer, Nielsen
Multi-touch attribution is an advanced way of measuring marketing effectiveness that helps you understand what’s working and what’s not.
For example, say your company sells running shoes. Your customer may see your ads on ESPN.com and Snapchat, get an email from you, see an ad on YouTube, do an online search and come to your site—all before buying a pair. How do you know what worked?
Multi-touch attribution analyzes your results and assigns credit to every touchpoint along the cross-channel consumer journey so you know what influenced the sale and what didn’t.
This measurement approach has multiple methods and models. Knowing which model is right for your business depends on what you want to do with output.
Read on to learn the fundamentals of multi-touch attribution, the differences in methods and models, and how to choose the right model for your business.
Multi-touch attribution eliminates biases by algorithmically allocating credit to every element of every touchpoint in the consumer journey, across marketing and advertising channels and tactics, according to its influence on driving a conversion event.
Multi-touch attribution is more sophisticated than traditional, rules-based attribution such as first- and last-touch. These rules-based methods give all of the credit to the first or last marketing touchpoint before a consumer converts through a purchase, download, or any other conversion event.
By applying rules, these methods fail to accurately measure the contribution of every touchpoint in the consumer journey, causing marketers to make decisions based on skewed data.
Leveraging individual, user-level data across addressable channels such as direct mail, online display and paid search, multi-touch attribution calculates and assigns credit for a KPI event to the marketing touchpoints along the consumer journey that influenced the outcome.
These insights into past performance can be used to inform more efficient and effective planning and optimization decisions for future campaigns, as well as those already in flight.
There are many different types of multi-touch attribution models and the sophistication of each model can differ dramatically.
Rules-based methods are subjective, relying on marketers to define the rules of how credit is allocated to one or more points in the consumer journey.
Another approach—algorithmic methodology—is based on objective, statistical modeling and machine-learning techniques. The output of an algorithmic model can be used to predict outcomes to help marketers plan or optimize future marketing efforts. But even algorithmic models vary in terms of their modeling sophistication and granularity of predictions.
No matter the method, all multi-touch attribution models help marketers determine the impact of their marketing investments at granular levels by assigning credit for a KPI event (conversion, lead, etc.) to one or more touchpoints.
Here’s an overview of rules-based and algorithmic attribution models:
These two models assign 100% of the credit for a sale to a single touchpoint.
In a last-touch model, the last touchpoint in the consumer journey receives 100% of the credit for the KPI event. Due to its history as the default marketing measurement methodology, this model is often used as a baseline to compare other multi-touch attribution models.
In a first-touch model, the first touchpoint in the consumer journey receives 100% of the credit for the KPI event. This model is often used to measure marketing efforts that are intended to drive awareness by reaching new consumers for the first time.
These three models assign credit across the customer journey according to a rule set in advance.
In an even-weighting model, also known as a linear model, all touchpoints in the consumer journey receive equal credit for the KPI event. This model is often used to measure marketing efforts that have long consideration cycles, making it important to reinforce the message many times throughout the consumer journey. It is also a good starting point for marketers that are new to multi-touch attribution and do not have an existing understanding of the typical consumer journey across channels.
In a position-based model, also known as a bathtub or u-shaped model, the first and last touchpoints in the consumer journey receive a higher percentage of credit than the touchpoints in between. This model is often used when marketers want to favor the “Opener” and “Closer” touchpoints more heavily than the “Supporter” touchpoints in the middle of the consumer journey.
In a time-decay model, the percentage of credit gradually builds while leading up to the last touchpoint in the consumer journey. This model is often used to measure marketing efforts that have short consideration cycles, such as promotional campaigns.
There are two main types of algorithmic multi-touch attribution: fractional and incremental.
Fractional attribution employs machine-learning to calculate and assign fractional credit for a given success metric to the influential marketing touchpoints and dimensions (campaign, placement, publisher, creative, offer, etc.) along the consumer journey, and to inform future marketing spend allocations.
Multi-dimensional techniques model every dimension of every touchpoint before assigning fractional credit. Not only can marketers use this model to objectively measure marketing performance, but they can also use scenario planning to predict outcomes at the most tactical levels of their marketing and media plans.
This allows you to understand optimal budget allocation at the channel and sub-channel levels, but also the most granular levels like keyword, placement and creative. Models are rebuilt as often as daily, so marketers can make optimization decisions while campaigns are still in flight.
Incrementality is the measure of the lift that advertising spend provides to the conversion rate. Incrementality is best used to determine not only where to allocate budget, but also how much budget to allocate.
Using incremental attribution, you can determine which campaigns, creatives, placements, etc. are performing best, and whether they should be investing in them at all. Incrementality also accounts for other factors, such as pre-existing brand awareness and demographic factors while fractional attribution does not.
At a time when marketers are increasingly held accountable for demonstrating marketing success, using attribution is better than siloed channel measurement or no measurement at all.
Any of these models will arm you with insight into the consumer journey that you didn’t have before. Ultimately, deciding which multi-touch attribution solution is right for you comes down to your goals, business requirements and how you want to use the output to improve the effectiveness of your marketing.
Download this ebook to learn more about attribution: Untangling Attribution’s Web of Confusion: A Primer for Marketers.