One of the keys to unlocking the full value of your investment in digital marketing lies in mastering attribution, which can be summed up as the science of determining which media, tactics and channels drive revenue for your business.
To expand on this definition, attribution is about identifying which customer actions across your various touchpoints, channels and media lead to the outcome you desire—whether that is a click-through, a view or purchase.
At its most sophisticated, multi-touch attribution (MTA) enables marketers to optimize spend and impact using advanced algorithms and machine learning (ML) to allocate weightings to specific media, channels and tactics that will influence a desired customer outcome. That means investment can be redirected to the channels (e.g. web, paid search, social media, mobile) that have the greatest impact for your business and your customers.
Here’s a quick summary of the more main-stream approaches to MTA along with their pros and cons:
First Touch-Last Touch: You attribute the customer conversion to the first or last marketing event or channel with which they engaged. It’s tidy and simple but masks the true complexity of how various channels played a part in getting the customer to perform the action you desired. It is a suitable approach for marketers who are just beginning their journey towards true multi-touch attribution and brands who are running simple demand generation and conversion campaigns.
Even weighting: For some marketers, giving every interaction equal (or even) weighting for a conversion is their chosen approach. Like First/Last-Touch, it’s simple in practice and implementation, but this overly simplified approach misappropriates credit which can result in an associated misallocation of spend. It’s often a place to start but is short-lived as marketers gain experience with the results and become better equipped for more advanced strategies.
Position-based: This approach attributes an equal weighting to the first and last events. It’s relatively easy and cost-effective to implement and is less skewed than first or last touch attribution. Yet it may underplay the importance of events between the last and first engagement. It’s well suited to marketing executions and campaigns where the impact of first and last event is usually high.
Time Decay: In this approach, a higher weighting is assigned to subsequent events in the customer journey, giving the most credit to the final event and the least to the first. This approach addresses the entire customer journey and is good for campaigns with short consideration cycles, for example promotions. Getting it right is time-consuming and there is a risk of underestimating the importance of events at the beginning of the journey or, top of the funnel.
Agent-based: Perhaps the most precise and sophisticated approach, algorithms simulate the actions and interactions of consumers to assess the effects of causal factors (marketing) on their behaviours (purchasing) and the system as a whole (market). It’s complex and expensive to implement, and demands that you have a mature technology stack and experience in multi-channel attribution.
When tuned and running, it can be a powerful tool for understanding the customer journey across large or simultaneous multi and omni-channel campaigns, especially where there is significant value tied to micro-conversion events. While it offers accurate data for optimisation, it does demand that your team has a high calibre of statistical and technical skill.
Mastering multi-touch attribution has never been more important, with brands under pressure to get the most bang for their marketing buck while delivering more relevant and personalised customer experiences. Getting it right is challenging, but the benefits are significant. Brands that have not yet experimented with it can start on small campaigns with a basic multi-touch attribution approach, then evolve to more sophisticated models and tools as they become more comfortable and experienced.