What Kitchen Staff Training Costs You When You Skip It
TL;DR: Menu dish performance tells you more than a sales count alone. Before cutting a dish, check its margin, placement, and context. A dish selling poorly may be a menu problem, not a recipe problem.
Before you pull a dish from your menu, stop. The fact that it is not selling does not automatically mean the dish is the problem, and cutting it without understanding why is one of the most common and costly mistakes in restaurant menu optimisation.
I have been in kitchens long enough to remember when we would drop something from the menu because the head chef had a bad week and needed a scapegoat. Nobody interrogated the evidence. The dish would disappear, something else would go in, and three months later that one would be struggling too. Same kitchen, same staff, same underlying problem, never identified.
Menu dish performance is not just a sales count. It is the relationship between how often a dish sells, how much profit it generates, and how visible it is in the first place. A dish that sells poorly on a Monday lunch sitting tells you almost nothing useful on its own. You need context before you reach for the red pen.
The proper framework for this is menu engineering, which places every dish into one of four categories based on its popularity and its contribution margin. High popularity, high margin: keep it and protect it. High margin, low popularity: this is the one worth investigating before you cut. Low margin, high popularity: you might be giving money away. Low popularity, low margin: that one can probably go.
The category that causes the most confusion is the second one. A dish with a strong margin that nobody seems to be ordering could be a failure, or it could be a neglected gem sitting in the wrong position on the page, described badly, priced in a way that makes people hesitate, or simply never mentioned by the person taking the order. Those are not problems with the dish. They are operational problems wearing a dish’s face.
When a dish is underperforming, the instinct is to blame the recipe. Sometimes that is right. But before you go there, work through the more mundane possibilities first, because they are far more common than we like to admit.
I once had a lamb shoulder dish that sold about four covers a week. We were close to pulling it. A young sous chef suggested we get the floor team to taste it before service one Friday and actively recommend it over the weekend. We sold nineteen portions in two days. The dish had not changed. The conversation had.
Proper restaurant menu analysis requires more than a glance at your EPOS report. Sales volume tells you what people ordered. It does not tell you what they nearly ordered, what they ignored, or what they asked about and then talked themselves out of. You need to combine your data with observation.
Look at time-of-day patterns. A dish might sell well at dinner but sit untouched at lunch, which tells you something about the audience at each service rather than anything fundamental about the food. Look at day-of-week patterns. Look at whether the dish was featured in any verbal recommendation or table interaction, if your team tracks that at all.
Compare the dish’s performance against its position in the menu mix. If it accounts for three percent of covers in a section where everything else sits between eight and fifteen percent, that gap is telling you something. If it accounts for three percent in a section with twelve options, the maths are at least plausible. Context matters enormously here.
Chefs often fixate on food cost percentage, which is a useful metric but an incomplete one. Contribution margin, the actual cash profit per dish after food cost, is what pays the rent. A dish with a thirty percent food cost but a selling price of £12 generates less cash than a dish with a forty percent food cost at £22. Low volume on a high-margin dish is a different problem from low volume on a low-margin one, and the response should differ accordingly.
Work out which dishes are actually carrying the room financially, and which ones you have been defending out of sentiment or stubbornness. (I have been guilty of both. The wild mushroom risotto incident of 2011 need not be revisited in public.)
Before cutting anything from the menu, run through this sequence. It takes an hour at most and it will save you considerable grief.
If you complete all six steps and the dish still does not move, then you have a legitimate case for removing or reworking it. But you will have done it on evidence rather than frustration, which makes a significant difference to what you replace it with.
Sometimes the dish genuinely is not working, and the right answer is to change it or lose it. The signs are fairly clear when you look honestly. Staff who have tasted it and still cannot sell it with conviction. Guests who order it and do not return it but also do not reorder it on future visits. A concept that made sense in the development kitchen but does not translate to a plate being carried across a dining room at pace.
If the feedback from tables, over time, is consistently muted, that is worth listening to. Not one grumpy Tuesday, but a pattern of indifference. Indifference is worse than a complaint. A complaint tells you what is wrong. Indifference tells you nobody cared enough to bother.
At minimum, six to eight weeks of consistent service, assuming the dish has been tasted by staff and described properly on the menu. New dishes often take time to build familiarity with both the team and returning guests. Pulling something after a fortnight is rarely anything other than impatience.
Change the description and the staff recommendation behaviour first. Those cost nothing and take effect immediately. If there is no movement after two to three weeks of genuine effort, then look at whether the price is misaligned with its neighbours. The recipe should be the last thing you touch unless customer feedback is specifically and repeatedly pointing at the food itself.
Menu engineering is the practice of categorising dishes by their sales volume and profit margin, then using that to make decisions about positioning, pricing, and promotion. It was formalised by Kasavana and Smith in the early 1980s and the core logic holds up well. It is useful precisely because it forces you to be quantitative about decisions that chefs often make on gut feeling. Gut feeling matters, but it benefits considerably from a spreadsheet beside it.
Yes, meaningfully so. Studies on menu layout, including research from Cornell’s hospitality school, show that placement within a section, the use of boxes or visual anchors, and the sequencing of price points all influence ordering behaviour. A dish moved from the bottom of a section to a highlighted position within that same section can see sales increase substantially without any other change.
The question worth sitting with is this: if you cut the dish tomorrow and replaced it with something new, what would be different about how you introduced it, described it, priced it, and sold it? If the answer is ‘nothing much’, you already know where the real problem lives.
Chef Ian McAndrew’s specialist eBooks and guides are available directly on ChefYesChef, including his technical titles and autobiography. If you want more practical, chef-led reading beyond this article, you’ll find the full collection here.
Chef Ian McAndrew works with chefs, businesses, and individuals on a wide range of culinary projects, from concept development to practical problem-solving.
If you’d like to talk through an idea or need informed guidance, you’re welcome to contact him.
Essential cookies required for the site to function. Cannot be disabled.
Cookies that help us understand how visitors use the site.
Cookies used to deliver relevant advertisements.
Privacy Policy Terms of Service
