Have you heard that data is the new oil? If not, then the chances are great that you will read this in an article somewhere—tomorrow or the day after. And, if data is the new oil, then the way to make money from it is by exploiting it—in as many ways possible. (Borrowing from businessman and author Stephen Covey, the habit of being obsessed with analytics is part of my continuing series on the seven “habits” that successful IoT projects have in common.)
How do you exploit data, you may ask? Well, just as in the oil industry, exploitation comes down to how good your refinery capabilities are. Just as oil has to be refined to get valuable products out of it, such as gasoline and jet fuel, data has to be refined into insights. And, just as in the old days of the oil business, the rush is on.
For this reason, many companies are snapping up data scientists. Those who can’t get a data scientist to join them try to hire consultancy firms to help. Consultancy firms that specialize in data science are exploding in number—and this is causing some problems.
Not Enough Data Scientists
For a start, there are not enough data scientists to quench the current thirst and appetite, driving salaries through the roof. If they get poached away from you, your domain/company knowledge could be gone tomorrow. I heard someone remark that today’s data scientist is what the programmer was in the 1980s. If that is true, the worst is still to come. (It is interesting to see that data science consultancy firms are establishing strongholds around university cities with a strong pedigree for analytics. Cluj in Romania is one such place, where tens of thousands of fresh graduates address the need for new hires.)
What do you do, however, if you don’t have the means to start centers of excellence in exotic places to get those data scientists? There are solutions coming that can help fight fire with fire.
The Advance of Self-Service Analytics
One of the most exciting trends in the market is the advance of self-service analytics, bringing the cup of wisdom to the many. One of the examples I encountered recently was the acquisition of a self-service analytical tool, TrendMiner, by my own company. It has capabilities that enable self-service analysis of time series data. That means knowledge workers, such as process engineers, can visualize any time series data they have access to and then ask the tool—through predefined analytical models and techniques such as machine learning—for correlations and similarities. The whole idea is to find indicators and early warnings that can be used to prevent major anomalies.
Just last week I was called by a colleague who had his first experience of using it for a day, as well as a knowledge worker in a pharmaceutical company, who said that within a few hours, they had found correlations between events they never thought were related.
The Future of Data Science
This strengthens my belief that the future and the way forward is to make the data science work easy. Similar to how we simplified programming by going from binary to third- and fourth-generation languages, we will see simplification of data science tools.
That is what democratization of data is all about. If data science is about asking questions, then it will not be about who can program the smartest answer, but who can ask the smartest question. An obsession with analytics pays off.
The future and the way forward is to make the data science work easy. Similar to how we simplified programming by going from binary to third- and fourth-generation languages, we will see simplification of data science tools.