Skip to main content
Log InSign Up

Processing yield multiplies, it does not add

Losses at each step compound, so a series of small losses is much larger than their sum. How to measure the whole chain with two numbers you already have, and where giveaway hides.

Updated

Raw material passes through several steps and loses something at each one — trimming, cooking, draining, portioning, packing. The instinct is to add those losses up. That is wrong, and it understates the total every time.

Each step keeps a **fraction of what the previous step handed it**, so the losses compound. A run of individually small losses ends up substantially larger than their sum, and the gap grows with the number of steps.

Why it multiplies

Step two does not lose a share of the original raw material. It loses a share of whatever survived step one. Chain enough steps together and the compounding does real damage — which is also the good news: **an improvement anywhere in the chain multiplies through everything downstream of it.**

This is the structural reason food plants get large results from small yield changes at early steps. An improvement at the first operation carries through every subsequent multiplication; the same improvement at the last one does not.

Measure the whole chain with two numbers

Quantity of raw material issued, and quantity of finished product delivered. Divide. That single ratio is the whole chain, and it needs no new measurement.

Then, if you want to find where the loss sits, weigh between steps for a few batches. Do that second — the overall figure tells you whether it is worth the trouble, and often it settles the question by itself.

Giveaway: the loss that is not a loss

Packs filled above their declared weight to stay safely above it are giving product away. It never looks like waste because nothing is thrown out — the customer simply receives more than they paid for.

It belongs in the material line, because that is what it is: raw material consumed and not sold. Measure it as average actual fill against declared fill, on data your checkweigher already produces.

This guide covers how to measure giveaway, not how much overfill is appropriate. That is a legal and specification question that depends on your product, your jurisdiction and your equipment, and you know it better than any general document does.

Where yield actually improves

  1. Attack the earliest steps first — the multiplication works in your favor there.
  2. Find uses for the byproduct: trim that becomes a second product stops being a loss and becomes revenue.
  3. Reduce giveaway by tightening fill control, which usually needs no capital at all.
  4. Separate variation from level: a step that averages well but scatters widely is losing product on the bad runs, and the average hides it.

Point four is the one most often missed, and it is visible the moment you record yield per batch rather than per month. Two plants with the same average yield and different scatter are not equally good, and only the per-batch figure shows which is which.

Related guides

Processing yield multiplies, it does not add | costdown.org