There’s probably a version of this conversation happening in every insights team right now. AI-moderated interviews are faster & cheaper than traditional IDIs. Self-guided digital diaries eliminate the problem of scheduling interviews. Automated platforms can synthesize themes across hundreds of responses in the time it takes to transcribe one. The case for efficiency is real, and for some research questions in specific contexts, the tools deliver genuine value.

Over the last few years, the insights industry has experienced a number of shifts. Providers that pretend otherwise, holding on to business as usual, are losing credibility with the insights buyers who know it best. Surveys are cheaper than ever. Panels are commoditized. AI is compressing timelines that used to require weeks of analyst (or moderator) time into hours. The basic infrastructure of research is now accessible at a fraction of what it cost even a few years ago.

Far from being doom & gloom, this is genuinely good news for organizations that need fast, directional answers to well-defined questions. It’s difficult news for organizations whose value proposition is built primarily around access to that infrastructure.

And the honest version of this conversation starts by acknowledging both things at once.

Most insights professionals have a version of this story. A research project comes together wonderfully: good sample, rigorous fieldwork with truly interesting learnings, and a clean narrative in the final deck. The presentation goes great. The team nods in the right places & seems to take good notes. There’s engaged discussion.

And then momentum stalls. The learnings get filed away. The decisions they were supposed to inform move ahead based on instinct & opinion. The research, weeks of work, real budget and genuine effort all quietly collect (digital) dust.

If you work in client-side insights, you may recognize a common dilemma. Your internal clients (product, marketing, brand, etc.) come to you with a business problem that demands research. You work with them to translate that problem into a research brief & RFP. You send it to a handful of potential vendors, evaluate proposals, and award the work.

In print, this looks like a clean process. In practice, we often see that it quietly sets research up to fall short. Not because the work isn’t good, but because the brief was never designed to produce success. It was designed to produce a report.