I’ve learned something funny about analytics over the past few years out in the field. After nearly every discussion, the customer or prospect says, “This is great, and we definitely need it — but what do we do with the results?”
The value of analytics is almost taken for granted at this point. Companies urgently need to become data-driven just to keep up with the market. So, naturally, managers and executives want to explore opportunities to leverage analytics to improve their business. But it’s that last bit — improving the business — that we often fail to think through from the beginning.
In his blog post “Clarify Marketing Impacts with Attributions and Marketing Mix Modeling,” Gartner contributor Chris Pemberton outlines five steps for successful marketing modeling. The first and fifth steps Pemberton describes — “define the goal” and “use the results,” respectively — are the ones through which companies often want to fast-forward. But in my experience, these are indeed cornerstones to successful engagements.
Let’s consider these two steps in the context of a typical attribution project. Since we released our Attribution Guided Analytics Interface a few months ago, I’ve recently been involved in several such projects.
Define the goal. “Begin with the end in mind,” Pemberton writes. Do you plan to use attribution so you can drop a few visuals into an executive deck to convince your management team that you are doing it right? Or are you planning to realign marketing investments quarterly or even weekly based on the results of one or more attribution models? Or do you plan to incorporate results from attribution models into management performance incentives?
Use the results. The ability to use these models “separates success from failure,” according to Pemberton. If you’ve ever tried to justify your investment in analytics to executives, or if you’ve ever presented to a room full of analytics professionals, you’re undoubtedly familiar with the question, “So what?” Pemberton suggests assigning someone upfront to capture findings and log the changes you make as a result of your analyses.