Surface localized product demand for assortment
Shoppers ask store staff for products you don't carry in that location, and most of those requests never reach the team that sets the range. NEXT reads where customers express that demand — in-store feedback, service chats, reviews, and local online searches — and groups the repeated requests by store. You get a short brief showing which products customers want, in which stores, how often, and what the missed sales are worth.
Most of this demand is invisible in your sales reports. You can't sell what isn't on the shelf, so a product customers keep asking for shows up as zero — not as demand. The request lives in a staff member's memory, a one-off email to the buyer, or nowhere at all.
What the demand brief looks like
Example output based on grouped in-store, service, and review feedback. Numbers are illustrative.
Product cluster
Oat-based and lactose-free chilled drinks, currently out of range in the South-West region
Where the demand shows up
11 stores, concentrated in three urban catchments with younger demographics
What customers say
"Third time I've asked here — you have the almond one but never the oat. I end up going to the shop down the road for it."
"Staff said they get asked for this every week, but it's not on their order list."
How often
Roughly 140 logged requests over eight weeks, plus matching local online searches for two named brands you don't stock regionally
Commercial exposure
An estimated £18K–£24K in monthly basket value walking out the door, based on the requested items plus typical attached purchases
Signal strength
Strong and consistent in the three urban catchments; thin and mixed in the six rural stores, where the same products show little repeat demand
The brief is ready before the regional range review, not reconstructed from memory in the meeting.
How NEXT detects this
NEXT reads the places customers express product demand — in-store feedback logged by staff, service conversations, reviews tied to a location, and local search interest. It keeps a running record of what customers ask for at each store, so a one-off request and a repeating pattern look different. When requests for the same out-of-range product cluster across a set of stores, NEXT groups them, estimates the missed-sales exposure, and writes an assortment recommendation routed to merchandising. The recommendation lands where the merchandising team already plans range. Retail Ops and merchandising still decide whether to stock it, where, and at what depth.
Why local demand surfaces late today
The request a customer makes at the till is the clearest demand signal you get. It is also the one your systems are worst at capturing.
Sales data only shows what sold. A product that isn't ranged in a store sells zero there, which reads as no demand — even when staff get asked for it every week. The signal is real; the report can't see it.
So you rely on people noticing. A store manager mentions it in a stand-down. A field manager emails the buyer. By the time it reaches merchandising, the wording is gone and only a vague "a few customers asked about oat drinks" is left. Each handoff strips a layer of detail until the request is too soft to act on.
The two tools meant to fix this don't. Open a dashboard and it shows what already sold, not what customers asked for and couldn't buy. Ask an AI assistant and you get the loudest recent complaint, not the pattern across eleven stores over two months. Neither comes looking for you.
NEXT pushes the demand to the team that sets the range, instead of waiting for a manager to escalate it or a buyer to ask the right question.
How this compares to the tools you already know
Approach | Where the evidence lives | What Retail Ops does at decision time |
|---|---|---|
Sales-data range review | Transactions for products already stocked | Infers demand from what sold; out-of-range demand is invisible |
Manager escalations and emails | Scattered across inboxes and stand-down notes | Reconstructs the request from memory, often after the range is set |
NEXT | A running record of requests grouped by store and product | Opens a brief that already shows the product, the stores, the frequency, and the exposure |
What changes for Retail Ops
Today, when a buyer asks "is there real demand for this in the South-West?", you go hunting. You ping a few store managers, you wait, you get three different answers, and you still can't say whether it's eleven stores or one loud one.
With the brief attached, that question is answered before the range review. You can see the product, the specific stores, the request count, and what it's worth — and you can see where the demand is thin, so you don't over-range a rural store on the back of urban signal.
The conversation changes. Instead of "a couple of people asked about oat drinks," merchandising opens with "140 requests across these eleven stores, mostly three catchments, worth roughly £20K a month in basket value." The range debate shifts from whether the demand is real to which stores and what depth.
You still make the call. NEXT brings the demand to the decision; it doesn't decide what goes on the shelf.
NEXT already supports retail and field teams at companies like Action and Rituals in connecting customer feedback from stores, service, and reviews to operational decisions.
Downstream effects
Range reviews start from local demand, not regional averages. A product that under-indexes regionally can still be a clear winner in three catchments — and the brief shows that split instead of burying it.
Field managers stop being the relay. The demand reaches merchandising without depending on someone remembering to escalate it, which frees the daily stand-down for execution rather than request-logging.
Uptake is tracked. Once a recommendation is stocked, NEXT keeps reading the same stores, so you can see whether the requests convert to sales or whether the demand was softer than it looked.
Where the human stays in control
NEXT groups requests and estimates exposure; it does not change a planogram or place an order. You set how many requests, over what window, across how many stores, before a cluster becomes a recommendation. You can require a human to review clusters before they are routed to merchandising, or let well-supported ones through and hold thin ones.
This is configuration work — request thresholds, store groupings, the catchments you treat as comparable — not approval work on every signal. Merchandising still owns the range; Retail Ops still owns whether the recommendation fits the operating reality of those stores.
What to configure first
The brief is only as good as where you let NEXT read. If staff rarely log in-store requests, the signal leans on reviews and search, which skews toward the loudest stores. Decide which sources count and make sure in-store request capture is consistent across the estate.
Set the cluster threshold to your replenishment reality — the number of stores and requests that justifies a range change differs for chilled food versus seasonal goods. Group stores by catchment, not just by region, so urban demand isn't diluted by rural stores that will never sell the item. And agree where the recommendation lands so merchandising sees it inside the planning cycle, not after the range is locked.
Where this breaks down
Thin or inconsistent in-store logging
If only a handful of stores log requests well, NEXT sees demand where capture is good, not where demand is highest. The fix is operational: consistent request logging, not more tuning.
Treating regional signal as local
Strong demand in three urban catchments does not mean range it across the region. If you ignore the thin-signal caveat and over-range, you carry stock that doesn't move and the brief gets blamed for a decision it flagged against.
Substitution noise
Some requests are for a product you already stock under a different name or pack size. Without that mapped, NEXT may cluster demand you're already partly meeting. Maintain the product synonyms so a real gap isn't confused with a labelling one.
Acting before uptake is read
A recommendation stocked and never checked tells you nothing. If you don't track whether requests convert once the product lands, you can't tell a real assortment gap from a vocal minority.
FAQ
How is this different from a sales report?
A sales report shows what sold in products you already stock. It cannot show demand for something that isn't on the shelf — that reads as zero. NEXT captures the requests customers make for out-of-range products, groups them by store, and estimates the missed sales, so you see demand the transaction data structurally can't.
Does NEXT change the range automatically?
No. NEXT clusters requests, estimates the exposure, and routes a recommendation to merchandising. It does not place orders, change planograms, or alter range. Whether to stock a product, in which stores, and at what depth stays a merchandising and Retail Ops decision.
How many requests before something becomes a recommendation?
You set that. The threshold is the number of requests, the time window, and the number of stores you consider meaningful — and it should match the category. Chilled food might justify a change on fewer, faster signals than seasonal goods. NEXT applies the threshold you configure rather than a fixed rule.
What if the demand is only in a few stores?
That is often the point. The brief shows where demand concentrates and where it's thin, so you can range to specific catchments instead of the whole region. The signal-strength note exists precisely so you don't over-range stores that won't sell the item.
Can we tell whether stocking it actually worked?
Yes. After a recommendation is stocked, NEXT keeps reading the same stores and tracks whether the requests convert to sales. That uptake check tells you whether the demand was real or whether a vocal few overstated it — and it informs the next range review.