“Analytics” is a popular buzz term today, and companies across sectors view the discipline as a new avenue for creating innovation and seizing competitive advantage. In order for these goals to be successful, however, it’s critical that organizations understand how to operationalize the insights that emerge from data analysis.
So what exactly does this mean? I recently wrote about this in my ITWorld blog but to summarize—operationalizing insights means aligning the behavior of all employees that are responsible for the desired outcome. This involves three critical elements:
- Behavior: Ensuring that all employees act in a way that delivers the outcome desired from the insight
- Metrics: Measuring and monitoring the impact of employees’ actions
- Response: Giving employees timely, actionable feedback on their behavior and its outcomes
Looking at the call center environment as an example, let’s say that analysts determine that customer satisfaction is highest when calls last two minutes. Any interactions that come in above or below the two-minute mark significantly lower customer satisfaction scores.
A knee-jerk reaction to this conclusion would be to standardize all calls to two minutes. However, this approach will produce the opposite effect in many situations. For example, some employees may just hang up at two minutes to adhere to the policy, even if they have not addressed the caller’s issue. It goes without saying that this will drive overall satisfaction down.
This example underscores the fundamental difference between insight and operationalization, and the reasons why analytics often fail to deliver economic value:
- An insight cannot be reliably operationalized as simply a written or verbal policy. Without constant measuring and monitoring, the feedback loop doesn’t exist and the expected behavior cannot be standardized. As a result, the expected outcomes will have either unintended consequences (i.e., hanging up) or a high degree of variation.
- Insights reveal facts but, by themselves, tell us nothing about the correct incentives and measurements to change the behavior of people in the direction of the desired outcome. In the call center example above, a policy mandating that all calls be two minutes can simply result in agents hanging up. While this meets the objective, it also causes more harm than good. As a result, the measure has to include numerous metrics in addition to timing—satisfaction for each call, average agent satisfaction, peer group comparisons, etc.
- Because operationalizing insights aims to standardize outcomes by changing behavior, it’s best done by applications and not through policy. The latter can be ambiguous and easily misinterpreted. Applications, by comparison, provide precise measurements and real-time feedback to employees as they are making decisions.
To be effective, apps aimed at operationalizing insights must be able to provide on-the-job measurement and decision guidance. In the case of the call center, this could be the duration of each call, the satisfaction with each call, and whether overall satisfaction is trending up or down compared to peers.
So with the above in mind, are you truly deriving the most value from your analytics? Do you foresee any challenges operationalizing insights in your organization? I’d love to hear your thoughts in the comments below.