Reducing feed formulation cost without losing performance

On a farm, feed weighs more than everything else. In France, pig producers spend close to 70% of their production cost on it (Inaporc, 2022). In the international InterPIG comparisons, the share runs from 56% in Finland to 80% in southern Brazil (2022 results). For laying hens it is around 60% in the United States. In catfish farming in Thailand it runs from 72 to 87% of cost per hectare depending on intensity (FAO, 2007).

A good order of magnitude to keep: feed represents 60 to 80% of the cost of producing an animal, and more in intensive aquaculture. One point gained on feed cost therefore weighs more than any other item. That is also why you cannot reduce feed formulation cost by looking at the tonne alone: what is saved on the tonne is lost if the animal converts feed less well.

An additive or a reformulation that saves energy or protein is not always worth the same amount: its value follows the price of the nutrient it replaces. Fifty kilocalories per kilo of feed are worth more when oil is expensive, so more in 2026 than in 2025. Conversely, a methionine saving counts for little when methionine is cheap.

Finished feed follows raw materials with a delay: in October 2026, Ifip notes that the rises have not yet been passed on and expects margins to fall in the coming months. For methionine, the world average for DL-methionine rises by 4% between the first and second quarters of 2026, but published series diverge: a firm quote is worth more than an average.

In South-East Asia, the rise weighs on countries that import a large share of their raw materials. In Vietnam, which imports more than 65% of them, the Ministry of Agriculture expects a further 3 to 5% rise in feed prices by the end of 2026.

A matrix value is the nutritional contribution assigned to an additive in the formulation software, as if it were a raw material: energy, methionine, choline, phosphorus. Least-cost feed formulation software then reduces by the same amount the raw materials that supplied those nutrients, and the additive pays for itself through the saving it allows. The software also gives, for each constraint, its marginal price: that is the right measure of what a saving is worth.

An additive is used in two ways. On top of an unchanged formula: a performance gain is expected. In reformulation: the formula is lowered by the matrix value and the same performance is expected at a lower cost. The two are not counted together: a saving already taken in the formula is not added to the performance gain.

Published values give an order of magnitude, not a promise. A meta-analysis published in 2020 in British Poultry Science reports that, in broilers, a lysolecithin compensates in reformulation for 57.88 kcal/kg at 125 g/t and 73.11 kcal/kg at 250 g/t. In chicks, a study published in the Journal of Nutrition in 2007 concludes that half of the choline requirement can be covered by betaine, provided the ration contains a minimum of choline. These values describe a class of molecules and vary with species, age, the fat in the ration and the level of stress. For a Linker product, the value to use is set with you, on your formula, after a trial.

• Start from a published value, then reduce it: begin at half or two thirds of it until it is validated on the mill’s own formula.
• Check that the substrate exists: an emulsifier brings no energy in a ration with almost no fat, and betaine saves nothing if the need for methyl groups is already covered.
• Do not stack: phytase, enzymes, an emulsifier and betaine sometimes claim the same energy. A study published in 2012 in the Journal of Applied Poultry Research showed that a phytogenic additive and an enzyme blend neither added up nor reinforced each other. Count the first matrix in full, reduce the following ones, and validate the combination in a trial.
• Validate in a three-arm trial: complete formula, formula lowered by the matrix, lowered formula with the additive. If the third arm catches up with the first on feed conversion and weight, the matrix is validated for that formula.
• Follow it over time: feed conversion, weight and uniformity, batch by batch and season by season, with a control batch where possible.

Measured results are on each platform’s product page, where published, with their conditions. The references of the publications cited are in the sources; the detail is available on request.