I remember the floppy disk: the MD’s room had the office’s one internet-connected computer, so emailing London meant carrying the disk over. It sounds like a museum exhibit now. That disk had more responsibility than some junior account handlers.
Before a campaign came research: quantitative, qualitative, customer groups, behaviour habits. Then hours trying to connect a woman in a rural village with a bar of soap. Coffee, cigarettes, and someone inevitably saying, “What if…?” just as the dots began to meet.
Today Claude Dashboards can turn plain-language questions about connected company data into charts and queries. A view of sales or sign-ups can arrive before we’ve made the coffee. Handy. I don’t miss every delay, or printing 80 pages for a meeting.
But those sessions weren’t valuable just because they took ages. They made us sit with the question. One person spotted a price barrier; another remembered something their mum said; someone joined soap to status and the neighbour’s washing line. The insight rarely appeared in a spreadsheet wearing a name badge. It came from disagreeing, getting distracted, then returning with a connection nobody knew to request.
AI makes it easier to explore the evidence. The risk is mistaking a tidy answer for the whole story. A dashboard can show what’s in the data we connected. We still have to decide if it’s the right question, what to doubt, and why an odd detail won’t leave us alone.
So where’s the journey? In the conversation after the dashboard appears: when we stop admiring tidy blue bars and ask, “Hang on. What’s going on in that village?” Maybe it’s moved from collecting clues to arguing about them, then following the odd detail that won’t sit still. The numbers can arrive before the kettle boils. The connection may take another pot.