Big data can revolutionize decision-making within an organization; but without great analytics, it can be a huge waste of time. This “big data trap” often gives organizations a false sense of confidence in the information they are gathering, resulting in wasted time, and gathering “bad data” that has the potential to create a negative impact on the business. Larry Freed, president and CEO of ForeSee, will present his ideas on predictive analytics at this year’s CEO2CEO Leadership Summit on December 9 at the NYSE.
As a CEO, how can you steer your organization clear of this trap? Here are four factors to consider in order to insure your big data is good data:
Capture the right data: Many organizations tend to focus on increasing the volume of one type of data captured, but they should really be focusing on gathering the right kind of information.
For example, when an advertising revenue-driven content website completed an overhaul of its navigation functionality, it immediately saw a drop in revenue. The behavioral data indicated that visitors were spending much less time on the site than before and therefore weren’t clicking through ads at the same volume – this was immediately perceived as a negative result. However, the customer experience data showed that site visitors were actually finding what they were looking for much faster than before, left with a higher level of satisfaction and a much higher likelihood to return, all of which predicted higher future revenues. The CEO was persuaded to let the change stand, and they saw revenues increase over time to higher levels than ever. If they’d relied only on behavioral data, they would have sacrificed long-term gain for short-term results.
Demand quality data: Levels of acceptable measurement error vary greatly by the situation and the accuracy and precision level of the data needs to fit the business decision you are making. For example, as a society we have very low tolerance for a measurement error when it comes to vehicle safety but we have a much higher tolerance when weathermen predict next week’s weather. When planning time and resources for a big data project, be sure to consider whether or not you are going to get the right accuracy and precision for the project.
Set proper metrics: It is important to use metrics that are predictive and actionable. A limited, one-question metric like net promoter score (NPS) that has not evolved over time will not give an organization the insights and action ability it needs to be successful. Applying the right system of analytics and the right metrics will make big data truly valuable.
Define the relationship: We also don’t want to fall into the trap of assuming that a correlated relationship is causation, just because we have a lot of data. Consider an example from a B2B software company that recently launched a new website. After the launch, the conversion and lead generation rates dropped dramatically, and the CEO panicked. At first, it seemed that the new site had caused the rates to drop but they began to dig through the data. In the end, they found that the new SEO campaign had begun driving tremendous traffic to the site, causing the number of visitors to spike and, therefore, the conversion rates to drop. It would have been easy to assume the new website caused the dropped rates, but by analyzing the various data available, they were able to correctly identify the cause.
Big data and great analytics will give you the opportunity to gain the competitive advantage that all companies are seeking.
Larry Freed is the president and CEO of ForeSee, a customer experience analytics firm that measures satisfaction and delivers powerful insights on where to prioritize improvements for maximum impact. He is the author of Innovating Analytics (Wiley), published earlier this Fall.