Short answer
When food cost percentage differs between restaurants with the same menu and the same suppliers, the difference is usually in how each kitchen orders, preps, portions and records waste. Those habits are not written down anywhere. They live with the head chef and the people who do the ordering.
The stock report shows which site is high. To find out why, ask the people in each kitchen to walk you through a normal week of ordering and prep, then compare sites. If that is your question, you can ask it now.
What the food cost report tells you, and where it stops
A food cost report compares what was bought and used with what was sold. It is good at three things:
- showing which site is above target;
- showing which categories drive the gap, such as protein or dairy;
- showing whether the gap is steady or seasonal.
It stops at the kitchen door. Two sites can show the same three-point variance on protein for entirely different reasons: one over-orders for the weekend, the other portions by eye. The report cannot tell them apart, and the usual response, a reminder about portion control sent to every site, treats them as the same.
Where variance usually comes from
| Area | What differs between sites | What to ask |
|---|---|---|
| Ordering | Who orders, when, and from what information | "Walk me through how you placed last Thursday's order." |
| Par levels | Whether they exist, and who updates them | "How do you decide how much to keep in?" |
| Receiving | Whether deliveries are weighed and checked | "What happened the last time a delivery was short or wrong?" |
| Prep | Batch sizes, and prep for covers that do not come | "How do you decide how much to prep on a Tuesday?" |
| Portioning | Scales, scoops, or judgement | "How does a new cook learn the portion for this dish?" |
| Waste | Whether it is recorded, and what happens with the record | "What went in the bin yesterday, and was it written down?" |
| Menu and specials | Use of trim and surplus | "What do you do with what is left on a Sunday night?" |
| Staff meals and comps | Counted or invisible | "How are staff meals and comped dishes recorded?" |
Each question asks about a recent, specific moment. "Do you follow the ordering procedure?" gets a yes. "Walk me through last Thursday's order" gets the procedure as it actually runs.
Who to ask
- The head chef at every site, including the ones doing well. What the best site does differently is the most valuable finding.
- Whoever places the orders, if it is someone else. It often is.
- A sous chef or senior cook per site, who sees prep and portioning daily.
- The general manager, for how the numbers are discussed.
Make it clear this is about method. If people think the exercise is a hunt for someone to blame, they will describe the manual.
A fictional example: six sites, one menu
A fictional casual dining group runs six restaurants with one menu and one supplier list. Food cost runs from 27% at the best site to 34% at the worst. Head office suspects theft at the two highest sites.
Before installing cameras, the operations director asks eleven people across the six kitchens one question: "How does your kitchen order and prepare for a normal week, and what gets in the way?"
- At the two lowest-cost sites, the head chef orders daily from a count done after service.
- At three sites, ordering is done twice a week from memory, by whoever is on.
- Five of the eleven say par levels were set at opening and never changed.
- Four describe prepping for weekend volume on Thursday, then discarding on Monday.
- Nobody outside the two best sites records waste.
The pattern suggests that food cost follows the person and the method of ordering far more than the site or the menu. The group can test it cheaply: move the three twice-weekly sites to a daily count for a month and watch the variance.
The example is invented. Eleven conversations do not rule out theft. They do offer an explanation that fits all six sites, and a test that costs nothing.
From what you hear to what you change
Summarise by practice, with a count:
| Practice | Sites | Next step |
|---|---|---|
| Orders from a count after service | 2 of 6 | Document it with those chefs |
| Orders from memory twice a week | 3 of 6 | Trial the daily count for four weeks |
| Pars unchanged since opening | 5 of 11 people | Reset pars with each chef |
| Waste not recorded | 4 of 6 | One simple sheet, reviewed weekly |
Let the chefs from the best sites explain their method to the others. A method described by a peer travels better than a policy from head office.
The same approach in other multi-site businesses
The pattern is general: a number varies between sites, and the reason sits in local habits. The same conversations work for:
- shrinkage between shops;
- labour cost between hotels;
- scrap rates between production lines;
- complaint rates between branches.
In each case, ask the people who do the work how a normal week runs, and compare the best with the rest. Writing a clear research question helps you frame it.
Where Lontra fits
With Lontra, you write the question in one sentence and review the topics it proposes. Each chef or manager gets a private link and answers in writing on a phone, between services, with no account. Lontra follows up on what they say, for example asking how the last order was actually placed. You read findings with how many people described each practice, where sites differ, and the quotes behind them. It is free to try with 10 conversations. It works like a conversational AI survey: one open question, then follow-ups.
It does not replace stock counts, invoices or your point-of-sale data, and it cannot observe a kitchen. It records what people say they do. Participants are told before they start that you will read their answers, so explain why you are asking.
Ask the kitchens before you write to them
The next time the report shows a gap between sites, hold the reminder email. Ask each kitchen how its week runs, and let the best site show you what to copy.
Frequently asked questions
- Why does food cost percentage differ between sites with the same menu?
- Usually because each kitchen orders, preps, portions and records waste in its own way. Those habits sit with the head chef and whoever places the orders, and they do not show in the stock report.
- How do you find the cause of food cost variance?
- Use the report to see which site and category are off, then ask the people in each kitchen to walk through a recent order and a normal prep day. Compare the lowest-cost site with the others.
- Who should you ask about food cost in a restaurant group?
- The head chef at every site, including the best ones, whoever places the orders, a senior cook, and the general manager. Make clear the purpose is method, with no blame attached.