Making data science projects profitable
Big data and machine learning are all the rage these days, so companies are throwing money into data science projects – often for zero return. Why? Invariably because they fail to connect the data science function to the business at large. But this can be addressed.
An MIT SMR piece has identified the five most common mistakes made by businesses when they launch data science projects. The authors illustrate these pitfalls, and most importantly, how to avoid them, with examples drawn from the private banking sector in India. And the solution really boils down to being (business-)smart about data science. Managers tended to be so dazzled by its rocket-science aura, or so infatuated with mysterious algorithms conjured by their “analytics wizards” that they throw themselves headlong into data science projects without ensuring that they actually fit with the current business context, priorities and processes. Yet managers need to truly integrate new tools from data science in the business as they would any other new tool: making sure it is relevant, and worth the extra investment; applying at the right time and putting it in the most appropriate hands; testing the tools and their delivery channel and so on. At the end of the day, the focus should be on business, not buzzwords.
“Why So Many Data Science Projects Fail to Deliver”
(Mayur P. Joshi, Ning Su, Robert D. Austin, and Anand K. Sundaram, MIT Sloan Management Review, March 2021).
© Copyright Business Digest - All rights reserved