How to Prioritize Analysis

Prioritizing analysis can be messy and political. When multiple teams clamor for limited analysts, the decision of what work will and won't be done usually doesn't leave anybody happy. After trying different approaches to take the politics out of prioritization, I realized that we were approaching the problem backward. We started with the assumption that as a support group the analytics team has to be reactive. The solution is for the analytics team to be proactive in setting the agenda for analysis.

Trying to Take Politics Out of Analysis

When we focus on "depoliticizing" prioritization we actually create more problems. At Dell, for example, we went through several iterations of deciding on a single metric (or set of metrics) that everybody could agree was important. Sometimes the metric was revenue. Sometimes it was customer satisfaction. This approach de-emphasized things that were important but only indirectly influenced the chosen metric. It led us to do silly things like focus on things directly touching purchases (like the checkout process) to the exclusion of things that had an important but indirect influence (like content strategy).

Likewise, leaving the prioritization up to our stakeholders has consequences that don't help our customers. I have seen teams divide their bandwidth equally among a set of stakeholders. I have also seen teams that collate lists of their stakeholders' priorities and then sort them by a common metric or by how many teams list the same priority. These approaches emphasize business goals over customer needs and put the analytics team in a powerless, reactive position.

Set the Agenda

These approaches overlook the real potential of an analytics team to set the agenda by laying out a plan that answers the overarching business questions. Embedded in all our stakeholders' requests are fundamental questions about the customer. If we can answer those bigger questions, the ad hoc requests resolve themselves. Analytics teams should lay out long-term plans that address the big questions that concern everybody. For example, map the customer decision process.

The power of a bigger plan like this is that it's easy to explain as a narrative. It prioritizes itself. Every stakeholder can see why it's important to know when our customers have which questions and how they expect to get answers. Getting to that level of understanding will require a phased approach that feels like a natural progression to stakeholders instead of like a contrived, arbitrary sequence.

It also puts ad hoc requests into perspective. It's easier to weigh whether a new request is more valuable than the bigger plan. With a long-term plan there are clear costs to refocusing on a short-term question, since it will delay the entire plan and set back the recommendations that will help everybody. It also provides some consistency on what the focus will be over time.

Once a company has a solid understanding of the fundamental business questions, prioritization of analysis can happen naturally because it will be obvious where the biggest gaps are. Those gaps, where the company creates dissonance between customer expectations and the company's messaging, should be analyzed first.

But who will decide where the biggest gaps are? The team with all the data will—the analytics team. It can set the agenda for where to focus the most resources. And the rest of the company is usually relieved to have them do it.

No comments:

Post a Comment