I’ve been doing a lot of market and user research lately, and I kept running into the same problem: the research itself wasn’t particularly difficult, but there were a ridiculous number of small steps around it.
For one project, I had to check competitor websites, product pages, Reddit discussions, a few industry reports, and some customer feedback. I’d have 15–20 browser tabs open, copy useful bits into a doc, clean up the notes, group similar findings, then go back through everything and try to answer the actual question I cared about. The annoying part was that AI tools could help with almost every individual step, but I was still the one doing the “glue work” between them.
That’s where I’ve started questioning the usual AI workflow. If I ask “What are the main trends in this market?”, I get an answer. But that doesn’t mean the research has actually been done. I still need to decide what sources to look at, collect the information, give it to the model, notice what’s missing, ask follow-up questions, and eventually turn the whole thing into something I can actually use. For a long research task, the bottleneck isn’t really generating text, but keeping the whole process moving.
I need a tool to research a market, gather information from different sources, compare competitors, identify recurring themes, and turn the findings into a structured report, rather than just another chat response. Of course it’s definitely not “press a button and replace the researcher.” But I wish myself spending less time reviewing the findings and copying information between tabs.
We’ve spent a lot of time asking whether models can answer questions better. Now I’m more interested in whether they can take something messy like “figure out what’s happening in this market and give me something I can make a decision from” and actually carry the whole task through.
[link] [comments]