Labour gets the attention. Ordering quietly does just as much damage.
Ask most operators where their margin is under pressure and they’ll talk about wages first — rightly, since it’s the biggest line item and the most visible one. But COGS and supplier ordering tend to leak margin just as consistently, just more quietly, because nobody’s watching it with the same scrutiny.
Nobody holds a crisis meeting over ordering 3% too much stock. It just shows up, month after month, as a COGS percentage that’s slightly worse than it should be — explained away as “food costs are just up everywhere right now.”
Sometimes that’s true. Often, it isn’t.
It matters more this year than most. Coffee, beef, and lamb have all seen sharp price rises over the past twelve months, and wholesalers have been public about margins on a cup of coffee shrinking to almost nothing at current prices. When the underlying cost of what you’re ordering is moving that fast, a guesswork buffer that used to cost you 2% now costs meaningfully more — the same bad habit, at a higher price.
The Guesswork Baked Into Most Ordering
Most ordering decisions come down to the same input: what moved last week, adjusted for what a manager expects this week. That’s not guesswork — it’s a reasonable estimate, built without a demand forecast to check it against.
That’s not a criticism — it’s the only option most operators have without a reliable demand forecast to order against. But it means ordering decisions are being made the same way rostering decisions used to be made before forecasting existed: as an estimate, padded with a safety margin, corrected only after the fact when the waste bin or the COGS report tells the real story.
The safety margin is the expensive part. Order a little heavy “just in case” often enough, across enough weeks, and it adds up to a COGS percentage that’s structurally higher than it needs to be — not because of one bad decision, but because of dozens of reasonable, cautious ones.
What Ties Ordering to Demand Actually Changes
The fix isn’t ordering less. It’s ordering against something more reliable than last week’s usage and this week’s instinct.
AI supplier governance ties ordering budgets directly to the same demand forecast driving your rostering — so what gets ordered reflects what’s actually coming, service period by service period, rather than a pattern from a week that may not repeat.
That doesn’t eliminate judgment. A venue manager who knows a local event is happening still factors that in. But it replaces the guesswork baseline with a forecast-driven one — so the starting point for every order is closer to right, and the adjustments on top of it are smaller and more deliberate.
Why This Compounds Faster Than Operators Expect
A 1–2% improvement in COGS doesn’t sound dramatic on its own. But COGS typically sits at a similar order of magnitude to labour as a percentage of revenue — which means a small, consistent improvement here compounds the same way a labour improvement does.
Across a single venue doing meaningful weekly turnover, 1–2% in COGS adds up over 50 weeks. Across a multi-venue group, replicated at every site, it becomes one of the larger and least-discussed sources of margin improvement available — precisely because it’s rarely tracked with the same discipline as labour.
The Waste Problem Nobody Line-Items
Waste is where ordering guesswork becomes visible, eventually — but usually well after the cost has already been incurred. Stock that doesn’t move gets thrown out, marked down, or absorbed quietly into a “shrinkage” line that nobody investigates too closely, because investigating it properly takes time most operators don’t have.
The honest fix isn’t a stricter waste policy. It’s ordering closer to actual demand in the first place, so there’s simply less excess to waste. When ordering is tied to a forecast rather than a habit, the waste conversation becomes much smaller — not because anyone got stricter about it, but because there’s less of it to manage.
Where This Fits Alongside Labour
Margin control that only looks at labour is only half the picture. The operators consistently reaching 8–16%+ margins are managing both sides of the cost equation with the same discipline — labour tracked against a live benchmark, and ordering tracked against a live demand forecast, with EBITDA pulling both together into one continuous profit picture.
Treating COGS as an afterthought to the labour conversation means leaving one of the two biggest levers in the business essentially unmanaged. The same logic extends to the rest of your fixed and semi-fixed spend — see hospitality overhead cost control for how the same visibility applies beyond ordering.
A Smaller Problem Than It Looks
Supplier ordering feels like a hard problem to fix because it’s been treated as an art rather than a data problem for so long. It isn’t. It’s a forecasting problem, same as rostering — and it responds to the same fix: a reliable demand signal, applied consistently, with the guesswork buffer removed.
See how forecast-driven ordering works for your venues. Book a demo at viability.io/book-a-demo.
Viability.io ties supplier ordering to live demand forecasts — helping Australian multi-venue operators cut waste and reduce COGS without guesswork.
Also in this series: