Tips for Managing Attribution in an Omnichannel World, Part 1
There's no question that marketing has become infinitely more complex for retailers over the past several years. The number of potential customer touchpoints has exploded, as consumers shift how they spend their time from nondigital (TV, print) to digital channels (web, mobile, social). The upside to this complexity is that retailers have a huge opportunity to engage with customers on a much richer and more personalized basis. The downside is that it's become even more complex for retailers to measure marketing effectiveness.
As a result of these challenges, there's a lot of focus currently in the technology market on attribution solutions (see Google's recent purchase of Adometry, AOL's acquisition of Convertro). Fundamentally these solutions attempt to apply advanced analytics to massive amounts of data in order to allocate "credit" to each customer interaction in driving conversion. Deployed properly, these capabilities can help companies more precisely measure the return on their marketing investment and optimize their marketing spend across channels.
However, retailers face a unique challenge when it comes to effective attribution. Even as they significantly increase the volume of digital marketing, the vast majority of conversions still take place in-store. Let's consider the following scenario: A consumer receives an email from a retailer announcing new shoe styles for the spring season. The consumer clicks through the email to the landing page, browses the products on the page, and then leaves the retailer's website. Later that day, the consumer is spending time on one of her regular news sites and is exposed to remarketing from the retailer. The next day she decides to go to one of the retailer's store locations and makes a purchase.
In this case, although the digital marketing investment made by the retailer had a significant impact in driving the in-store conversion, these digital touchpoints won't receive any credit because they can't be tied to the in-store purchase. This is a significant challenge, as the true impact from digital marketing isn't being fully measured and therefore the level of investment can't be optimized.
And yet, this scenario is becoming more common as consumers expect to engage with retailers seamlessly across channels and digital advertising takes on ever greater importance. According to research conducted by Deloitte Digital, digital across all platforms — desktops/laptops, tablets, smartphones — influenced 36 percent of in-store purchases in 2013 vs. 14 percent in 2012. This trend shows no signs of slowing.
So what can retailers do to ensure that cross-channel marketing investment is being accurately measured and optimized? Ultimately the answer will involve a combination of data, technology and people/process. The following questions should be considered by retailers as they start down the path of true omnichannel attribution:
- Is customer data being gathered at the point of sale in-store? What's the incentive for customers to provide information like an email address?
- Assuming a process is in place to gather customer information at the point of sale, do in-store associates have the proper customer service skills and training to effectively do this?
- Is a technology platform in place to integrate in-store data with data from other customer touchpoints (e.g., email, direct mail, web, mobile) and tie those touchpoints into "event streams" associated with individual customers?
- Is a reporting infrastructure in place to track marketing effectiveness at both a detailed and macro level, as well as respond to changes?
The benefits to retailers of getting omnichannel attribution right are clear: marketing dollars can be deployed where they'll drive maximum results in terms of online and in-store revenue and margin. However, the challenges are equally clear. So where should you begin? In part two of this series, I'll lay out a road map and a set of recommendations for retailers considering an omnichannel attribution initiative.