One directive is worth more than a thousand graphs

Delivering visual insights from unstructured data is just half the work. The most crucial part of any successful analytics story comes after the data has been crunched. Sometimes, we tend to forget why we train our AI and machine learning algorithms to present us with fancy graphs.

If you can think of it, chances are there is a dataset on it. Data has been called the new oil again and again, but it’s no less true. Just like oil, data houses tremendous potential and value. While the comparison works, it has its limits. Anyone with access to a developed oil field can see the money pouring in. With data, monetization isn’t quite so easy. Real value only emerges after thorough analysis, and even then there are obstacles to overcome.

While data analysis is getting more advanced every day, it’s also becoming more accessible. Any business user today can get access to beautiful informative dashboards showing graphs and predictions about the present and the future of the company. Those insights are interesting for sure, but in and of themselves they don’t create added value. The dashboard is just a stop down the road to data monetization, not the destination.

Too much data, too few experts

During the and& festival in Leuven a couple of months ago, a number of speakers identified this issue. Kristof De Mey, technology expert, noticed there is an abundance of data today, but there is a shortage of people who know what to do with it. Sports scientist Andrew Jones agreed, and so did Chris Van Hoof. As an engineer involved in medical research, he has noticed the value of data analytics in the battle against stress. To his surprise, he discovered the most difficult part is always in communicating insights from devices or medical personnel to the patients themselves.

The real added value is found in the communication. Jones quickly realized that during his work with world class athletes. They don’t care for dashboards full of graphical visualizations of relevant data. They might even start to overthink what they’re doing. An athlete needs a very tangible set of directions, and nothing more.

That’s an interesting insight, because data about the athlete can only provide added value once it reaches him. The scientist looking at the graphs won’t win the next race. It’s his job to look beyond the broader insights, and to discover straightforward directives. Maybe drinking slightly more during a marathon will allow a runner to shave five seconds off his best time, and that might be the difference between first and fifth place. Why this new drinking affects his time, isn’t relevant to him.

With patients suffering from stress, the same applies. Showing them a comprehensive visualization of all the data relevant to their stress level will probably do more harm than good. A single directive such as ‘try to take some time off next week’, will generate a far
better result.

From insight to value

Both examples illustrate the often overlooked last step in any successful data analytics project: actually translating insights into very tangible action plans. When looking at data in a corporate environment, the same can be said. A data scientist needs to translate available data into usable information about, for instance, overhead costs or growth predictions. Those graphs might show a profit decrease in the coming years. By just looking at the analyzed data, you might discover a certain department, while working hard, isn’t performing very efficiently.

This insight is theoretically very valuable, but won’t increase profit until someone turns it into a useful directive. “Throw out the old scanner, buy a new one that can scan invoices in bulk, and move to a cloud solution to make our accountants more productive.” At this point, you’re actually making a difference. Data might have taught you that the accounting department was costing too much, and it might have indicated the increased customer base bogging down the already inefficient department, but only when you translate these insights into a straightforward plan, value is created.

The big challenge for organizations is in finding people who can translate raw insights into sound business decisions. This is what De Mey meant when he lamented the lack of qualified data scientists. Today, gathering insights from raw data is the easy part. Ever smarter software, powered by the cloud, can quickly discover correlations and trends in available datasets. Far too often, the results of this analysis fail to reach people who actually know what they mean for their business. So ask yourself the following question: who is looking at the analytics dashboard in your organization? And what is he or she doing with the newly gained insights?