All these predictions about how AI is going to replace analysts! I think it could be one of the best times to be one. Hear me out…
For a long time, the instinct in analytics has been the same: if you want to understand something, quantify it. Numbers were the language, validly so, given how easy they are to handle, and we’ve built entire careers analysing them to make better business decisions. But AI has now opened up something we could never really touch before: the qualitative data. I’m talking about the text. Feedback, conversations, opinions. The “soft” data!
Every company that’s been around more than 6 or 7 years is sitting on a mountain of it. Complaints. Reviews. Support calls. Employee feedback. Every bug ever logged. Every product decision ever documented. Years and years of words, just sitting in some wiki or ticketing tool, never once treated as a dataset.
We couldn’t really do much with it before. You’d run a keyword search, maybe build a clunky sentiment score, and call it a day. But now, an LLM can actually read a complaint and tell you what it’s really about. Not whether it contains the word “late”, but what the person is actually unhappy with, how valid the complaint is, the potential damage to the business, the level of frustration, even the likelihood of it turning into a public complaint that costs you the brand’s image. That’s a different thing entirely.
And it doesn’t stop at text either. A support call recording, a product photo a customer attaches to a complaint, a screen recording of a bug! That’s all analysable at scale now, and that changes everything.
Plainly put, I’m just really excited about this.
And with that excitement, here’s where I land: an analyst’s job was never really about numbers. It was always about solving problems with whatever data you could get your hands on. Numbers just happened to be the only kind we could actually work with, for a long time. Now words are workable too, and that’s not a new skill to learn. It’s just a bigger playground for a skill we already have!
First posted on LinkedIn.