Why data analytics’ small "aha" moments can have a big impact

A company engages in data analytics and eagerly awaits the results, expecting a big “aha” conclusion about what’s really causing a particular problem or situation. But when the results align exactly with what people had suspected, there is sometimes disappointment. Business leaders start to wonder why they spent so much time and money on analytics when the results only tell them what they had intuited all along.

Recently Varicent – a Sales Performance Management software provider – worked with a technology company that had high hopes for data analytics to uncover why they were experiencing higher-than-expected employee turnover. The analytics concluded that turnover was caused by low-performing employees leaving the company to do something else – just as company leaders had thought. In their disappointment that the data analytics results weren’t unexpected or dramatic, they called the findings “underwhelming.”

The data analysis had, however, done its job: The primary driver was identified and confirmed, and now the company had concrete evidence on which to devise and implement solutions.

Recalibrating expectations is crucial as more business leaders try to become data-driven in their decision making. Often, they expect a big reveal that makes people say, “Wow – we never saw that one.”

Sometimes data analytics does deliver those "wow moments" – but usually not. Analytics are just as valuable when confirming hypotheses, which are otherwise guesses.

What data analytics can do

Data analytics are growing in popularity among business leaders. In a survey by Harvard Business Review Analytic Services, nearly 80% of survey respondents projected that customer analytics would be important to the overall performance of their organisation. More than half had already seen improvement in customer retention and loyalty because of analytics.

"Nearly 80% of survey respondents projected that customer analytics would be important to the overall performance of their organisation."

To tap the potential of data analytics, it’s important to understand how they’re used and what problems they can address. Data analytics can be grouped into two basic categories. The first is “transactional” analysis, which uses data in mathematical expressions or other business rules. For example, data analyses can be used to project the spending habits of customers. Coupled with profit margins on various products, such data-driven spending projections help companies forecast their profits with greater accuracy.

The second type of analytics is when companies want to discover the underlying story of why something happened or, conversely, what might happen with a different set of variables. This is “generative” data analysis – a rich area to be explored in search of “hidden nuggets” of information or insights. With generative analysis, the objective is discovering drivers and contributing factors – sometimes surprising, but often confirming.

"Any business leader engaging in generative analytics should be willing to accept results that are largely in line with expectations"

In fact, any business leader engaging in generative analytics should be willing to accept results that are largely in line with expectations. People who are closest to the situation (e.g., salespeople who frequently interact with and receive feedback from customers or hiring managers who know the people who are leaving the company) may have great insights. But it takes data to provide concrete confirmation, especially since frequently, intuition is not correct.

Why confirmation matters

When analytics prove what previously had only been intuited, it’s an opportunity invest based on real intelligence. Over time, a portfolio of evidence can be built that shows what works and to what degree. Building evidence isn’t solely about finding new and hidden trends, but rather serving as a foundation upon which future decisions can be made.

In the case of the employee turnover problem, data-driven confirmation that low performers accounted for the bulk of the people who left the company pointed to steps that could be taken. For example, the company could improve its recruiting, screening and hiring processes. In addition, the company may want to invest in higher performers to ensure this valuable talent is retained.

Once the data analysis confirms what was believed to be true, then analytics resources can be deployed elsewhere. Across any company, there are numerous areas in which generative analytics, in particular, can identify key drivers and provide confirmation (along with the occasional surprise). No matter how surprising the outcome, solid evidence should never be a disappointment.