Skip to content
allanninal.dev/build
Collaborate

Part 4 of 7 · Menu cost calculator series ~5 min read

Why portion size moves more than price does

Everybody watches ingredient prices, which move a few per cent a year and are outside your control. Almost nobody watches portion size, which moves more, is entirely within your control, and does not appear on any invoice.

Key takeaways

  • Portion drift is gradual, unintentional, and larger than annual price movement.
  • The only way to see it is theoretical usage against actual usage.
  • Compute theoretical from dishes sold times recipe quantities.
  • The gap is portion drift, waste, and theft, in roughly that order of size.
  • Scales on the line are the fix, and they need a reason to be used.

The relative sizes

Four influences on annual food cost comparedA bar chart with four bars showing effect on food cost over a year as a percentage. Ingredient prices: four per cent. Yield change: three per cent. Portion drift: twelve per cent. Waste: five per cent. A note says the one everybody watches is the smallest and the least controllable.05101520~4Ingredient prices~3Yield change~12Portion drift~5WasteEffect on food cost over a year, %The one everybody watches is the smallest and the least controllable.
Fig 1. Four influences on food cost over a year. Portion drift is the largest, is invisible on any invoice, and is the only one fully within the kitchen’s control.

Portion drift happens for entirely reasonable reasons. A busy service, a chef being generous, a new person who has never seen the intended portion, a plate that looks empty with the correct amount on it. None of them is a decision and together they are the largest single line on that chart.

Theoretical against actual

How theoretical ingredient usage is compared with actual usageA vertical chain of five steps entered by a box labelled A month of sales by dish. Step one multiplies by recipe quantity per ingredient. Step two produces theoretical usage, what should have been used. Step three computes actual usage as opening stock plus purchases minus closing stock. Step four takes the gap per ingredient. Step five ranks by value rather than by percentage. A note says this calculation is the only way portion drift becomes visible and it needs a stock count.AWS ACCOUNTA month of salesby dishTimes recipe quantityper ingredientTheoretical usagewhat should have been usedActual usageopening + purchases - closingThe gapper ingredientRank by valuenot by percentageThis calculation is the only way portion drift becomes visible. It needs a stock count.
Fig 2. How portion drift is measured. It requires a stock count, which is the reason most kitchens never see the number.
  • Compute
  • Analytics
  • Front-end & mobile

It needs a stock count

There is no way round this. Actual usage can only be computed from opening stock, purchases and closing stock, which means counting the kitchen. Monthly is enough for the high-value ingredients, and counting only the top twenty by spend captures most of the value at a fraction of the effort.

That partial count is worth stating as an option, because a full kitchen count is a half-day and a top-twenty count is forty minutes. The forty-minute version finds nearly all the money.

Rank by value

A twenty per cent variance on herbs is a few pounds; a six per cent variance on beef is several hundred. Ranking the gap by percentage puts the cheap ingredients at the top and sends people to investigate the wrong things.

What the gap is made of

In roughly descending order

  • Portion drift. Usually the largest, and entirely fixable with scales and a reason to use them.
  • Waste. Spoilage, over-production, dropped plates, returns. Some of it is unavoidable.
  • Staff food. Real consumption that belongs in its own line, not in variance.
  • Recipe inaccuracy. The written recipe does not match what is actually made.
  • Count errors. Particularly on part-used containers and on anything measured by eye.
  • Theft. Real, and usually smaller than the first two, and worth investigating only after they have been ruled out.

The order matters because the instinct on seeing a variance is to reach for the last item. In most kitchens the first two account for the great majority of it, and starting there is both more productive and considerably less damaging.

The fourth item is worth checking early and is easy to miss: a recipe that says a hundred and eighty grams when the kitchen has always served two hundred produces a permanent eleven per cent variance that is a documentation problem rather than a control one.

Fixing it

How portion drift is reduced and then re-measuredA horizontal row of five boxes. Scales on the line for the top five dishes. A reason to use them: the number, shared. Recount in a month, same method. Gap falls from twelve per cent to four: measured, not claimed. It drifts back, so measure quarterly. A note says the last box is the honest one, and this is maintenance rather than a project.PORTION CONTROL, MEASUREDScales on the linefor the top 5 dishesA reason to use themthe number, sharedRecount in a monthsame methodGap 12% to 4%measured, not claimedIt drifts backso measure quarterlyThe last box is the honest one. This is maintenance, not a project.
Fig 3. How portion drift is corrected and verified. The last box is the part that gets left out of most improvement stories.
  • Compute
  • Machine learning
  • Management
  • Analytics

Sharing the number with the kitchen is the part that works. A variance figure presented as a problem to be solved by the people portioning, with the value attached, gets a better response than an instruction to be careful, and the recount a month later is what makes it real.

Next: reading the profitability report.

All posts