Two cooks can portion the same dish from the same recipe card and send out plates that look alike but weigh differently. Across a service the gap reaches guests as uneven plates and reaches the kitchen as product that runs out sooner or later than expected. Weighing the portions that actually leave the pass shows what is happening rather than what the card assumes.
This guide covers one narrow check: the net weight of individual portions of one component, measured during real service. It is separate from measuring a batch's total yield or scaling a recipe. The target, review band and weights below are an original fictional example, not a recommended standard.
Fix the check before anyone weighs a plate
Write down the recipe and its revision, the finished product, the component being portioned and the checkpoint where it is weighed, such as in the serving container before it leaves the pass. Record the target as net grams per portion. Adding a sauce, garnish or plate weight into the figure turns the check into a different question.
Set the review band before sampling and have an authorised person approve it for this item. A band is a local trigger for looking more closely. It is not a legal minimum, a nutrition or health threshold, a promise to guests or a tolerance that suits every kitchen. A band chosen after seeing the weights tends to be drawn around whatever happened.
Record how each portion is made and weighed
For each check, note the outlet or station, the service period, the batch reference, the utensil or method used to portion and the role of the person portioning. A ladle, a scoop and a spoon judged by eye are different methods and belong in the record. Use role labels rather than names where the sheet is shared.
Identify the scale, how your site checks it and how tare is handled: either zero the scale with the empty container in place, or weigh the empty container and subtract it. Use one approach throughout a check. The worksheet accepts the resulting net weight; it cannot confirm that the scale was accurate or the tare correct.
Plan which portions to weigh
Decide the selection plan before service, for example a stated number of portions spread across the opening, busiest and closing parts of the period, or every portion of the item at set intervals. Record the plan beside the results. Weighing only the plates that look right, or only those from the most experienced person, produces a reassuring figure that says little about service.
A check weighs a sample, not every portion served. Report it as what was observed in that sample under those conditions. This guide does not set a minimum number of observations or estimate statistical confidence.
Record first, correct later
Write each weight as it is read, before anyone adjusts the portion. If a portion is topped up and weighed again, retain that recheck in the separate manual history, linked to the original observation. Do not add corrected rechecks to the original sample count or replace its light readings: either change would hide the first-pass variation. Start a separately defined sample when assessing a changed method.
A missing reading is not zero. A genuine no-product observation can be recorded as 0 g, explained and investigated. Start a separate sample if the recipe revision, intended target or measurement method changes. Record planned differences in batch, time and portioning role so you can interpret the spread rather than treating unlike observations as interchangeable.
Worked example: five portions, one target
In this fictional check, the target is 125 g net and the locally approved review band is 120 g to 130 g, with both limits counted as within. Five portions were weighed during one defined check: 110 g, 120 g, 125 g, 130 g and 140 g.
Together they weigh 625 g, so the mean is 125 g, exactly the target. Yet two of the five are outside the band: 110 g is below and 140 g is above. Three are within. That is 40 per cent of this five-observation sample outside the band. The lightest portion is 110 g, the heaviest 140 g, and the range is 30 g.
| Fictional observation | Net weight (g) | Difference from 125 g | Against 120–130 g band |
|---|---|---|---|
| P01 | 110 | -15 | Below |
| P02 | 120 | -5 | Within |
| P03 | 125 | 0 | Within |
| P04 | 130 | +5 | Within |
| P05 | 140 | +15 | Above |
| Summary | Total 625, mean 125, min 110, max 140, range 30 | Sum of differences 0 | Below 1 / Within 3 / Above 1; 2 of 5 outside |
Why an on-target average is not enough
The differences from 125 g are minus 15, minus 5, 0, plus 5 and plus 15. They sum to zero, which is why the mean matches the target. A zero sum shows only that the deviations offset each other in arithmetic. It does not show that portions were consistent: one guest received noticeably less and another noticeably more.
The total sample mass can match five target portions while individual servings differ. That does not establish the cost effect or identify waste. Keep any stock, waste or cost investigation separate and supported by its own records.
Decide the next step in a manual record: recheck the utensil or method, review the recipe or portioning guidance, retrain, or retest under stated conditions. Record who decided, on what evidence, and when the next check is due.
Where the principles come from
The Institute of Child Nutrition recipe standardization guide, prepared with USDA Food and Nutrition Service, advises stating a practical serving size and the serving utensil, giving portioning directions, and, when verifying yield, comparing the weights of actual servings with the intended serving weight and reviewing the method or recipe where they differ. This article uses only those general principles. School meal programme rules, crediting, nutrient and safety requirements, batch sizes and source examples are not applied.
Put it into practice
Your action checklist
Your progress saves in this browser. No sign-up needed.
0 of 9 complete
From reading to doing
Portion consistency observation record
The Portion Check worksheet records up to 12 original observations in net grams against one target and an approved local review band. Valid rows show their difference from target and Below, Within or Above. Counts, mean, minimum, maximum, range and the sample fraction outside the band appear only after all entered observations are valid and the sample is confirmed complete. A confirmed empty sample shows zero counts but no averages or rates. Manual sections retain recipe revision, scale and tare method, sampling plan, conditions, separate linked rechecks and the decision. It does not verify weighing accuracy, approve service or estimate costs, waste or statistical confidence.
Open the Portion Check tab in the launch workbook. Keep a copy for your own business.
Download the free workbook See what’s inside the toolkit →Sources & further reading
- USDA Recipe Standardization Guide for School Nutrition Programs (Institute of Child Nutrition with USDA FNS) ↗
2022 edition, retrieved 8 November 2026. Serving Size, Recipe Yield and yield verification passages used only for stating a practical serving size and utensil, portioning directions, and comparing actual serving weights with the intended weight before reviewing method or recipe. School meal programme rules, crediting, nutrient, safety and temperature requirements, batch sizes and source examples are not applied. All figures here are original.
Published by Keystonne Grow. Product links connect to the Keystonne ecosystem; check the linked product or service for its current scope.