RH Dépannage is a vehicle recovery business in Lyon: two trucks, a phone that rings at terrible hours, and — like every trade that's been going a few years — a customer list nobody had touched since it was exported from the old system.
The situation
1,842 past customers and old enquiries, sitting in a spreadsheet. Every recovery firm has this file. Nobody works it, because working it is a fortnight of admin: dedupe it, work out who's still local, write something that doesn't sound like spam, send it in batches, log who replied, book the work. The maths never works when the person doing it is also the person out on a recovery.
Meanwhile winter — the season when tired batteries and bald tyres become call-outs — was eight weeks away.
The instruction, verbatim
“Go through the old list and get us some pre-winter vehicle checks booked. Don't annoy anyone we've helped recently.”
One sentence, typed by the owner, with the rules already standing:
The plan it came back with
Datify read the list against its own records first, split what was left — past breakdown customers ahead of one-off enquiries — and held the priced plan for approval. The owner read it, changed nothing, and pressed approve. Total time invested: one coffee.
What actually happened
Sends went out paced over eight days — a short, plain message about a pre-winter check, no discount. Replies landed on the right customer record, responders dropped out of the follow-up automatically, and the two people who asked to be left alone were marked so it stays true forever.
What changed
Twenty-three checks booked into the diary from a file that had been dead weight for two years — at RH Dépannage's call-out rates, a five-figure return on a list they already owned, for roughly an hour of the owner's attention across the fortnight.
The interesting part isn't the 23. The list is a maintained asset now: 104 conversations logged, dead addresses gone, and next autumn the same instruction takes thirty seconds, because the reactivation rate is already part of the system's arithmetic.
One instruction, a plan read before it ran, and a ledger that never rounds up. 23 jobs booked from a dead list — and the same loop runs again next month without being asked twice.