The writer is Daniel Djupsjöbacka, Head of Data Science at Gambit (part of Atea).
Currently, 90% of all data there is has been created in the last two years. In two years’ time, there will be ten times more data stored compared to today. This explosion of information is called “Big Data”. It fundamentally changes our environment.
To make use of all this information a new profession has emerged, the Data Scientist. Being a Data Scientist myself, I will write about the lens that the Data Scientist looks through when observing the world.
Surrounded by data and insights
As a kid, I always had a small notepad with me when our family was out driving. Each page was headed by a car brand, and as we drove through the landscape, I focused on taking notes on the number of car brands we encountered. If it was a Volvo, the page with Volvos got a score. At times, when we parked, I walked through the parking lot, carefully collecting the datapoints on car brand quantity.
I compared the frequency of brands across parking lots. Toyota, a cheaper brand then, indicated less wealth or status among the owners. German cars the other way around. From the datapoints I collected, I could interpret something about the group of people that had parked their cars on the lot.
This is still a habit of mine. Parking lots are generally sized optimally with respect to the business they serve. The number of cars does not necessarily say that much about how well the business is doing, rather is it the percentage of parking spaces used that tells the story. And since the profit margin of a business outpaces the growth of revenue, it’s exciting to infer business profitability from the percentage of used lots outside a company’s parking place.
More data than ever – why are results missing?
When the data scientist looks at the world, he digs into the data and tries to uncover insights from that data. The data is the lens, through which reality is observed.
However, as the recent report Five Rules for Fixing AI in Business from Boston Consulting Group points out, many AI projects struggle to deliver expected business value. Why is that? The amount of data is vastly increasing, and data is collected in ever-growing ways. Analytic methods are improving. How come many projects still fail?
Part of the answer lies in the specific expertise of the people involved in the development of AI. In organizations, artificial intelligence and machine learning are often part of the technology or IT strategy.
However, data and AI are not only technological domains. Business processes are complex, and better forecasts are of no value to the business if they aren’t attached to actual decision making.
For successful projects we need Data Scientists with strong business knowledge and business experts who understand what they can achieve with AI. That is how data can become the lens through which business decisions are made, and it’s not until then that data becomes truly enjoyable and useful.