Industry
Retail and e-commerce
Most retail AI spend goes on the storefront. The money is in the stockroom, and it has been all along.
Where we stand
The industry sells retailers personalisation engines and chatbots because those demo well. For a business of 20 to 250 people, the largest recoverable number is almost always working capital sitting in stock that turns twice a year, and the second largest is the checkout step nobody has instrumented. Both are unglamorous, both are measurable within a quarter, and neither needs a foundation model to start. We would rather fix your reorder cycle than put a chat widget on your homepage — and if you already have the stock right, we will say the forecasting work is not worth doing.
Three problems we see repeatedly
Not an exhaustive list of what can go wrong — the three we see over and over, with what they cost and the shape of the fix.
The reorder decision runs on a Tally export, last year's numbers and somebody's instinct about the weather.
What it costs: Typically three to four months of cover on lines that turn twice a year — cash that is the real constraint on opening the next location.
The fix: Reconcile the sales history into one product record, forecast per SKU with a size curve, and hand back a reorder sheet with its reasoning attached.
Online and in-store keep separate product records, and the SKU codes drifted apart years ago.
What it costs: Every stock question needs a manual reconciliation first, so nobody asks the questions that would be useful weekly.
The fix: One product identity across both, with the drift resolved once and a nightly job that stops it recurring.
Nobody knows which checkout step loses the basket, because the checkout has never been instrumented.
What it costs: Two-thirds basket abandonment is common, and the assumed cause is price roughly as often as it is a form that breaks on a phone.
The fix: Instrument before rebuilding. Watch it for two weeks, then rebuild only the steps the data indicts.
Not yet published
We publish a vertical once two client stories stand behind it. This one has fewer, so it is listed honestly rather than dressed up.
We work in retail & e-commerce but we will not claim depth here yet. An empty vertical advertising itself as a speciality is worse than an absent one, so this hub stays out of the main navigation until the proof exists.
What we have done so far:
Three days of stock guesswork, gone
A two-floor clothing retailer replaced a Monday morning reorder meeting with a forecast read from its own till data.
31% target reduction in stock-outs across 400 SKUs
120 staffThe carts that used to just leave
A grocery group's online checkout was losing more baskets than it converted. Rebuilding it recovered nearly a quarter of them.
23% target recovery of abandoned baskets within eight weeks
You can still scope work with us. You will be told exactly which parts have precedent behind them and which would be a first.
Scope retail & e-commerce work
Tell us what is slow and what it is costing. You get the arc it belongs to, a published price band, and the closest story we have — or a straight answer that we do not have one.
Scope retail work with us