Attribution — the process of assigning sales outcomes to one or more channels – can seem less like science and more like Harry Potter wizardry at times. But marketing muggles have analytic choices that can help connect attribution to the digital activity that’s driving sales.
Attribution has traditionally been difficult for marketers to measure well. Marketers wanted to understand the ROI from digital marketing media, but early efforts were rudimentary, and typically based attribution on the last click. Last click assumes that the final referral source tracked by an analytics solution is the only influence on the conversion. For example, a click from a tweet that leads a reader to a website page means that Twitter is the referral source in an analytic solution. This assumption, however, ignores other influences in play before, or at the time of, the click.
What has changed since those early times is the increased visibility of those influences relative to cross-channel activity. This is due to the increasing use of platforms. Platforms are an ecosystem of digital media with which customers interact. Marketing and technology luminaries have long highlighted the value platforms bring to a brand. Scott Brinker, creator of the Chief Martec blog, and author of Hacking Marketing, has noted the strategic importance of a platform:
A2. A platform is not a linearly extrapolation of “another kind of product.” It’s a qualitatively different mindset: product companies are measured on the success of their product; platform companies are measured on the success of their ecosystem. #STinsights
— Scott Brinker (@chiefmartec) March 29, 2018
So how does a marketer approach attribution in this ecosystem?
The most straightforward way of developing a meaningful attribution platform is to apply a “dual sphere” measurement environment. Each environment is meant to address a general activity objective – one for activity in a medium, another to help capture the overall marketing ecosystem. I call them spheres to emphasis that these environments can overlap, but that they should reveal the overall relationship of platform activity to business metrics.
The first sphere is meant to capture activity in each part of a marketing ecosystem. The end result is dashboard metrics related a given channel. The types of reports can vary by platform, ranging from measurements within the manager interface for paid ads (like Google Ad Manager) to social media analytics (such as Pinterest Analytics or Facebook Insights). These reports analyze performance without the context of a website.
The second sphere addresses the platform, by combining data from across the ecosystem, and then generting visualizations that provide an overall view relative to the business metrics. In days past, that meant web analytics solutions like Google and Adobe. But over the years, other dashboard solutions have come into play, offering more statistical detail. The dashboards run the gamut from online cloud solutions like Tableau, Google Data Studio, and Neo4j, to data science models built in either Python or R programming.
Using these tools can be a gateway for deeper understanding of revenue drivers, and even a gateway for planning machine learning initiatives.
These platforms allow for regression capability, ranking different channels as significant contributors to conversion goals. A regression can not only help marketers organize marketing resources to improve the ecosystem, but it can also reveal which resources should receive investment based on sustained relationships. For example, a regression shows a statistically consistent relationship between the media ecosystem and a dependent variable — which is usually sales or market share in some models. Marketers can then use that regression as guidance for establishing resource priorities. Other kinds of useful data patterns, such as determining a sustainable growth trend, or a sales churn, can be predicted based on the model.
Marketing attribution analysis was in its infancy when social media first took hold. But the world now has a lot of history and experience with social media, while tech companies have added new media for customer engagement — apps, chatbots, virtual reality, and voice assistant devices. These media overlap in activity to drive sales. Applying a dual sphere measurement strategy to marketing attribution can better display value from that overlap, and can map customers’ cross channel activity.
The end result when it comes to accurate attribution is at least an improvement in the wizardry.
