Saturday lunch, a food hall with fourteen stalls trading. One vendor’s fryer goes down to half capacity and their queue quietly triples — nobody at the front desk knows until a customer complains they’ve been waiting twenty-five minutes for something that usually takes eight. Thirteen other stalls are running fine. The operator just had no way to see that one wasn’t.
Why walking the floor every twenty minutes doesn’t fix it
Physically checking on every stall is the obvious response, and it doesn’t scale. A market big enough to need active management is a market too big to babysit stall-by-stall, and the moment you’re not standing at a particular counter, you’re blind to it again. This isn’t a staffing gap — it’s that each stall is genuinely its own operation, running its own pace, with no shared view back to whoever’s meant to be running the market as a whole.
The instinct to fix this by tightening control — standardising every stall’s process, putting the operator’s own staff behind every till — usually backfires. Vendors joined a shared market to run their own stall, not hand it over. The real fix isn’t more control. It’s visibility that doesn’t require it.
What actually needs to happen
Give every stall its own real, working POS
Each vendor runs a proper POS for their own stall — fast enough for their own queue, with their own reports and their own numbers. Vendors keep running their stall exactly as they choose; nothing about visibility for the operator should mean rewriting how any one stall works.
Let one QR carry the whole market’s orders
Customers scan a single code, order from any stall in the market, and pay once — but each vendor sees only their own tickets, exactly as if the customer had walked straight to their counter. That single ordering layer is also what gives the operator a live, market-wide read on where orders are actually landing, without a single vendor losing control of their own queue.
Report per-stall and in aggregate, in the same place
This is the actual fix for the fryer-going-down scenario: the operator sees aggregate volume and per-vendor status side by side, so a stall that’s suddenly taking twice as long to clear tickets is visible immediately — not after a customer complaint, and not because someone happened to walk past. Vendors’ financial data stays strictly theirs; the operator sees enough to manage the market as a whole, and nothing more.
Recognise the customer, market-wide
A visitor who ate at three different stalls across two visits should register as one loyal customer to the market — not three anonymous transactions. A shared CRM and loyalty layer across every stall means the operator can promote the market as a destination, not just hope each vendor happens to remember a regular.
Give customers the same visibility you just gave yourself
The same blind spot that hides a backed-up stall from the operator also hides basic information from customers browsing the market — which stalls are gluten-free, halal or vegan, without walking the whole floor asking. Once every stall declares its own allergen and dietary tags, customers filter the entire market at once, and “does anywhere here do vegan?” stops being a question anyone has to answer in person.
Keep vendors comfortable enough to actually share data
Visibility only works if vendors trust it won’t be used against them. Strict data isolation — each vendor sees only their own orders and customers, the operator sees aggregate plus per-vendor for fees and reconciliation — is what makes stalls willing to run on a shared system in the first place. And if a vendor eventually leaves the market, their menu, customer history and data leave with them. No lock-in is what keeps the arrangement honest.
Control was never about the till
The operator’s job was never to run every stall — it’s to know what’s happening across all of them at once. Give each vendor their own working system and give the operator a shared window into all of them, and you get a market that runs like one venue without a single stallholder feeling managed.