Most consumer brands can tell you their repeat rate. Very few can tell you why customers stop reordering, which ones are about to go, or what to change first. That's the work. Your order history, your cancellation reasons, your reviews and survey responses — turned into a prioritised plan your team can act on this quarter.
Take a brand doing £5M a year losing 8% of its customer base every month. That's roughly £400,000 of revenue walking out the door annually, and with acquisition costs up 222% in eight years, replacing it gets more expensive every quarter.
Most brands aren't ready to commission a full diagnostic from someone they've just met. So pick the single question that's bothering you most and I'll answer that one first. Narrow scope, fixed price, one CSV export from you.
Across DTC, 12–30% of customers are gone after the first cycle. Almost nobody has that split by acquisition channel. You'll find out which channels are buying you customers who never come back.
You collect cancellation reasons. Nobody has read all of them. I categorise a sample and quantify what's actually driving people out — and which reasons are growing.
Not every churned customer is worth chasing. Some cohorts cost more to reacquire than they'll ever return. This tells you which subscribers to defend and which to let go.
When one question isn't enough. These are the full pieces of work — the complete picture, the systems that keep it running, and the monthly intelligence that replaces an in-house hire.
You send me your order history and cancellation records. I come back with the honest picture: who stops reordering, when in the lifecycle, from which segments, and why. Repeat rate, time to second order, one-and-done customers, retention by acquisition channel. Not a dashboard — a written answer to the question your board keeps asking, ranked by what to fix first.
Cancellation reasons, support tickets, reviews, NPS verbatims and survey open-text — thousands of rows nobody has time to read. I categorise the lot, quantify the themes, and tell you what customers are actually saying versus what you assume they're saying. Usually the gap is the whole story.
The reason most insight work dies is that somebody has to redo it manually every month. I configure a repeatable pipeline inside your existing stack — monday.com, Sheets or Looker, whichever you already run — so data comes in, gets categorised, and a report goes out without you rebuilding it from scratch each time. Handed over with playbooks so your team owns it.
Your insight function, without the £65k hire. Every month: refreshed cohort analysis, new feedback categorised, churn drivers tracked against last month, and a call where we decide what to do about it. Most brands land here after a diagnostic and stay.
A free 30-minute call. You tell me what you can already see and what you can't. I tell you honestly whether there's a real question here worth paying to answer.
A CSV export, read-only access, or a shared folder. Usually under an hour of your team's time. NDA signed before anything moves.
Analysis, categorisation, cohort modelling. One mid-point check so the findings aren't a surprise and the questions stay the ones you care about.
Written deck, ranked recommendations, and a call where we agree what changes first. You get the working files, not just the summary.
High-frequency reorder categories, where a customer is meant to come back every 30 to 60 days and the whole model breaks if they don't. Big enough to have real data. Small enough that nobody in-house is already answering these questions.
Thirty minutes. You show me what you can see, I'll tell you what your data could answer and whether it's worth paying to find out. If it isn't, I'll say so.