The biggest mistakes Companies make when using business analytics

Many companies use advanced analytical tools, but they are far from reaping the maximum benefits. At the same time, analytics can permeate the entire company, optimize marketing campaigns, understand customer preferences and behavior, and find new sales opportunities.

Commenting on InformationWeek, John Edwards focuses on the main mistakes that companies make when using analytics. It does not mean specific software, types of products (business analytics, data warehouses, etc.), or technology as such, but rather matters of organization and processes.

According to him, the main mistake is the insufficient integration of the analytical department into the very operation of the company, its sales department. The analytical department then turns into a “service” that handles requests, much like someone else is in charge of printers, for example.

However, this is wrong. Analysts tend to be highly qualified people. They have to come up with their own proposals, be at the center of the sales department, and design projects related to business transformation. They should be given space and asked for a proactive approach.

The second problem is related to the first. The sales department often wants analysts to just confirm their previously adopted opinion or even justify a decision already made. This does not mean that preliminary hypotheses should not be worked with and tested, but the decision itself should follow (if it is, of course, possible for time, etc. reasons) only after the analytics. In addition, analysts should be free to ask their own questions and come up with answers from the data that no one originally wanted.

Companies should also use analytics to a greater extent for “internal audit”, i.e. to assess how effectively their processes or various departments work. If such tools are available, they need not be limited to analyzing customer behavior or financial flows, but constantly expanding the deployment of these technologies; especially when digitization provides much more data (even types of data) than before.

Another question is in what form analytical departments present their results to others (visualization and easy comprehensibility of the output should be a matter of course). In part, there may be some kind of data overload and fascination, when in an effort to use all the possibilities and functions.

A large number of other interesting findings, suggestions and hypotheses will occur, but the main issue will be solved (actually the opposite extreme than the problems mentioned above). The project is delayed and will not be implemented in the originally planned time.

Finally, analytical outputs cannot be of better quality than input data. This results in the often mentioned recommendations on unified data management, access to data, if possible in real time and without data forces, where all departments of the company work with the same set of data. Data management may require additional investment. However, it requires such support if the analyst is to deliver the expected results.