Food-factory yield: connect material loss to margin

Water is the main source of false yield movement in food processing. A process that adds water and then removes some of it will show different apparent yields depending on whether inputs and outputs are recorded as-received or on a dry-matter basis.

If input is recorded as-received and output on a dry basis, part of the water disappears from the calculation and appears as a yield gain. If the reverse, it appears as a loss. Neither reflects anything the operation did.

The fix is to fix the basis explicitly and apply it consistently across periods. Where a process adds or removes significant water, it is usually clearer to reconcile on a dry-matter basis and track water separately as its own flow, rather than trying to express both in one percentage. That also makes changes in moisture content visible as what they are, rather than as unexplained yield movement.

Once the reconciliation is stable, the useful step is to test which component matters financially. The answer is frequently not the largest tonnage.

Consider the illustrative figures above. Waste to disposal at 22 tonnes is the smallest substantive output, but it carries a disposal cost as well as the lost input cost, so its financial weight per tonne is higher than its volume suggests. The coproduct at 48 tonnes carries revenue, so it offsets rather than costs. Rework at 35 tonnes carries the cost of the process steps it has already been through, and may carry a quality risk if it is reintroduced repeatedly.

Running the sensitivity — what happens to margin if waste falls by a quarter, if coproduct value rises, if rework is halved — identifies which of these is worth attacking. In many cases the answer is that a modest reduction in waste outweighs a larger improvement in coproduct yield, because the former removes a double cost while the latter adds only revenue.

Material lost late in a process route has absorbed more cost than material lost early. Valuing all loss at raw material purchase cost therefore understates losses that occur downstream.

The more accurate approach is to value loss at the cumulative cost of the furthest process step the material reached, then net off any recovery value. Material lost at filling has absorbed preparation, processing and handling; its loss destroys that accumulated value, not merely the ingredient cost.

This reframing often changes priorities. A small tonnage lost at a late stage can outweigh a larger tonnage lost at receiving, and a business chasing the larger tonnage may be addressing the smaller financial problem. Where the loss also has an environmental dimension, the same boundary question arises: the Commission has recommended the use of Environmental Footprint methods to make such assessments comparable, [D] and those methods share the principle that where a loss occurs in the chain changes what it costs to lose it.

Yield comparisons across periods are affected by product mix. A period with a higher share of a naturally lower-yield product will show a lower overall yield without any operational deterioration. Where mix is changing, comparisons should be made at product level, or normalised for mix, before conclusions are drawn.

Process bottlenecks matter for a related reason. A yield improvement at a step that is not constraining output may produce little financial benefit, because the constraint lies elsewhere. Conversely, reducing loss at the bottleneck step can release capacity as well as improve yield, which makes it worth more than the material saving alone.

Connecting the two — a defensible reconciliation, a value-per-tonne that reflects where in the route material was lost, and a view of where the constraint sits — is what makes a yield programme a margin programme rather than an exercise in reporting.