Data and AI · Luxury pre-owned watches
What is this exact watch worth today? Now there is one number.
One trusted figure for every watch a UK luxury pre-owned watch dealer buys, trades and sells.
- Client
- A UK luxury pre-owned watch dealer
- Sector
- Luxury retail
- Services
- Data & Intelligence, AI & Automation
- Role
- Product lead and build, with a collaborator on data monitoring
- Status
- Live
01
The problem
Price evidence for a specific reference, dial, bracelet, year, condition and box and papers was scattered across marketplaces, dealer sites, auctions and trade chat groups.
The dealer needed one figure to buy, trade and sell at, plus an instant valuation on their website that never exposes margins.
What we did
- 01
One market database
Aggregated 10+ market sources into one database, refreshed every day.
- 02
A pricing engine
Buy, trade and sell figures for every variant, adjusted for year, condition and box and papers using measured data.
- 03
AI reading the trade
AI reads trade messages and listing photos, for example the bracelet type, into a review queue before anything is used.
- 04
An accuracy loop
Every pricing change is scored against dealer-verified prices.
- 05
A website valuation tool
Shows one rounded range, and declines to answer when the data is thin or stale.
03
The results.
Two separate checks: 13 real dealer offers, and 97 dealer-verified prices.
Median error against 13 real dealer offers, down from −12.6%.
Offers within ±10%, up from 3 of 13.
Median error against 97 dealer-verified prices.
References covered, across 12 brands.
What made the difference
Letting the tool say no. A valuation that declines when the evidence is thin is one the dealer can stand behind, and the accuracy loop shows whether each change actually helped.
Start a project
Got a similar problem?
Tell us what is not working. You will get a plain answer on whether we can help, and what it would take.