Quadrupling profit through analytics: Data science with business value in mind

The initial hype around data science has all but died. Why is it that some companies hired great data scientists, gave them all the resources and very interesting and impactful problems, but failed to realize value out of them? And more importantly, how do we integrate analytics in business in a way that generates value?

TBC Bank went though many pitfalls while connecting analytics to business value. The company found that the two biggest obstacles were viewing data scientists as unicorns who can work and deliver in isolation and not giving data science projects clearly defined success/failure measures.

Mikheil Nadareishvili explains how TBC Bank reorganized its analytics to avoid these two problems. Firstly, it connected data scientists to data engineers at an organizational level so that data scientists would be free to focus on building models. Then it created a “translator” role, whose sole responsibility was to make sure data scientists understood business problems and that business understood what analytics could deliver for them. Finally, the company devised a structural way of measuring success of analytics that was deeply connected to business value.

As a result of transformation, TBC Bank has seen a huge increase in business value generated out of analytics, widespread adoption and respect for analytics throughout the organization, and a surge in employee satisfaction for its staff. In one case, the company was able to increase the profitability of one of the most popular loan products four-fold.