> data-driven approach to our investments This is as absurd as saying "data-driven approach for finding out successful humans". I am kind of aggravated to see drivel like this from VC's. This is similar to algorithm trading on stock market and not funding deserving startups. This 'data-drive' garbage is how you end up with more grocery delivery, food delivery and pet walking startups.
But yes, these days I think "data-driven" is really just a convoluted way of saying "we approximate the informed decisions of others without specifically understanding what we're doing". This is ML in a nut-shell after all. Is it possible to use ML and data science to find insights and draw conclusions that you then base your activities off of? Sure, but most are just looking for a magic black box that makes decisions for them specifically because they don't want to go to the trouble of understanding X. Once you abstract things away to that extent, you're inevitably training your algorithms to do what others do in an informed way, because where else would the training data come from?