Cleaning Up the Data Crumbs
Overlapping data is all around us. Consumers are leaving breadcrumbs of data all over their shopping trails with each purchasing decision they make. Analysts follow close behind, collecting all of the crumbs and reassembling the purchase paths to a sale.
Here’s the challenge: There are many silos involved in today’s marketing world, and each collects its own data. Each also wants its share of the credit for the sale. If we add up the shares the various silos claim, it always totals more than 100 percent.
Let’s examine why this happens, look at an example and consider a solution.
Why Overlapping Data Happens
There are many data points that influence a purchase in the ever-expanding marketing world. The macro segments that overlap for delivering content are television, print, Internet, retail stores and word of mouth. Inside each of these macro segments are more data points. In fact, the Internet alone creates attribution paralysis; the analytics of the Internet platforms do not line up with analytics for each of the components within the Internet platforms.
Data scientists do not have a single, agreed-upon set of rules for making sense of all of the overlapping data. It’s like the Wild West out there, where the biggest gun rules! That means there is a negotiation happening and the most persuasive person at the table wins. Analysts are left in the middle inhaling the smoke from all the dueling parties. The challenge is how to settle on what actually drives marketing success, rather than award credit to the loudest voice.
What Overlapping Data Looks Like
As an example, let’s compare some data I recently encountered with a client. Using the same keyword, time band and a last-touch attribution model, both a Google AdWords partner and an organic search partner reported on their view of the results.
The Google AdWords partner claimed the search reference code — Air Humidifier AB 123 — from its paid search landing page was clicked on and the partner did not see that transaction inside Google Analytics. This meant that its efforts represented the last touch. The organic search partner saw that a click on its organic search page had come through as a Last Referrer URL and reported that as the last click.
The rub comes when there is a lack of common rules for which the reference code is defined as “last-touch.” For the same sale, the client was paying 100 percent for both the Google AdWords cost and for getting the organic keyword to rise to the top of the organic search results page.
The amount of money being invested in marketing channels today, together with the need for some kind of ROI model, is a pressing concern for all brands. The data itself provides an overwhelming amount of visibility, but somehow we need to make sense of it all. As the gatekeepers of the data, analysts are increasingly finding themselves embroiled in this challenge.
Analysts know better than anyone that there is more than one way to report on a set of data. This means the interpretation starts with them. Sometimes analysts have a clear strategic direction set by the brand itself or management. Other times, analysts are fully trusted to report on data as they believe is best. Here is the first crossroads: “Best” can be in the best interests of the brand itself, the manager the analyst reports to, or the analyst him or herself.
Once the sales data is delivered, the reporting process is reviewed by bean counters, where a second filter is applied to relevant expenses. This is where the ROI of the past and the future is determined. There are prejudices and choices at this stage, as well. The same overlap reality exists on the expense side. For example, the amount of overhead expenses that needs to be included in any marketing ROI judgment is an open question.
Finally, management makes the call for where to invest its limited marketing budget. The various choices for this stage are well beyond the scope of this column. Suffice it to say that management relies heavily on how data is reported to them.
At the end of the day, the value of a dollar invested in marketing has to be measured against the value that dollar puts in the bank. Analysts’ reporting serves as the foundation for this decision.
Solving the Problem
There are two keys to solving the challenge of overlapping data from the analyst’s point of view. The first is data integrity, and that is the job of the analyst. Verifying data is a cornerstone of everything that flows from the data itself. Regardless of any reporting views bias or business rule constraints, the reporting of data eventually informs budget decisions.
The second key is a process of negotiation. Once the data is ready for interpretation, only a series of negotiations between the different tactical teams can reveal actionable insights.
It is critical to weigh the influence of various elements relative to the key behavior that drives brand success. These insights inform the allocation of marketing budgets across all touchpoints and establish achievable short-term milestones and long-term goals.
Multichannel data analysts face unprecedented challenges today. There are many choices to be made every day. A clear path forward needs to be negotiated by the entire team to be successful.