Feed Quality and Formulation: The Connection That Defines Cost and Performance

Every feed is formulated to meet a specific set of nutritional requirements. But the feed coming out of the mixer is rarely identical to what was calculated in the software. This difference exists and is expected to some degree, but when it exceeds acceptable limits it translates into unnecessary costs, below-projected animal performance, and inconsistency between lots. The gap between the formulated feed and the produced feed is largely a function of the quality of ingredients used. And that is why quality and formulation should not be two separate areas: they are two sides of the same process.

This article discusses how quality management directly influences formulation precision, why raw material variability is one of the biggest operational challenges of a feed mill, and how the integration between laboratory and formulation can reduce costs, increase product consistency, and improve customer animal performance.

Formulation begins with data: the risk of working with theoretical values

Feed formulation is, in essence, a mathematical optimization exercise. The formulator defines the animal's nutritional requirements, inputs available ingredients with their composition values, and the system finds the lowest-cost combination meeting the requirements. The quality of this result depends directly on input data precision: if ingredient composition values are incorrect or outdated, the formulation will be wrong, even if the optimization algorithm is perfect.

The problem is that ingredient nutritional composition values vary. They vary between crops, production regions, suppliers, processing methods, and seasons. Soybean meal produced in Mato Grosso during a good-moisture crop may have different composition from the same soybean meal produced in Paraná during an adverse crop. Using a fixed table value to represent all lots of this ingredient throughout the year is a simplification with real cost.

The solution is not to eliminate the use of reference tables, which continue to be useful as starting points. The solution is to complement these values with laboratory analyses of the lots actually available in inventory, updating nutritional matrices continuously and systematically. This transforms formulation from an exercise based on historical averages to a process based on the reality of each ingredient lot available for use at that moment.

What makes up an ingredient's nutritional matrix

An ingredient's nutritional matrix is the set of values representing its chemical composition and digestibility or availability coefficients for the target animal. For energy ingredients like corn, metabolizable energy, starch, ether extract, and moisture values are most critical. For protein ingredients like soybean meal and animal-origin meals, crude protein, total amino acids, and digestible amino acid contents take center stage. For minerals and premixes, each micronutrient's guarantee level is the fundamental parameter.

When the laboratory performs systematic analyses of received ingredients and this data is consolidated and passed to the formulation team with regular frequency, the nutritional matrices used in formulation software begin to reflect the actual profile of ingredients in use. This directly impacts the nutritional precision of the final product and, consequently, the performance of animals consuming it.

Raw material variability: the invisible cost that most mills don't measure

Raw material variability is one of the most studied and at the same time most underestimated topics in feed mill management. The fact that an ingredient has natural variation between lots is widely recognized. But the financial cost of this variability, and how it accumulates over hundreds of production lots per year, is rarely rigorously quantified.

A study published in All About Feed magazine (Volume 29, Issue 2, 2021) by Wisium analyzed crude protein variability of Brazilian soybean meal in different lots throughout the year. Data showed that while customers used a fixed formulation value of 46.8% crude protein, the actual statistical mean of received lots was 46.4%, with significant deviations above and below throughout the months.

The practical effect of this variability manifests in two opposite and equally problematic directions. In lots where soybean meal arrived with protein above the formulation value, the produced feed had excess protein above what was necessary. This excess represents cost without proportional performance return. In lots where meal arrived with protein below formulation value, feed was nutritionally deficient, with direct impact on animal performance. The study estimated that a plant producing 100,000 tons per year would have unnecessary costs exceeding R$260,000 just by working with a soybean meal protein value above ideal for that specific lot. The same reasoning applies to metabolizable energy, amino acids, ether extract, and other critical parameters.

This is the cost variability imposes when there is no integrated system for updating nutritional matrices. The plant pays for excess in some lots and delivers deficiency in others, without visibility into either scenario.

Variability in different ingredients and what to monitor

Soybean meal is the most studied ingredient in this context, but variability is present in virtually every raw material used in animal nutrition. Corn varies in metabolizable energy and can show aflatoxin contamination depending on storage conditions and crop. Animal-origin meals, such as meat and bone meal and fish meal, show relevant variation in crude protein, calcium, available phosphorus, and essential amino acids depending on processed raw material, cooking method, and supplier quality control. Fats and oils vary in fatty acids and quality indices such as acidity and peroxides. Distillery coproducts vary widely in degradable protein and energy.

Knowing which parameters to monitor in each ingredient is a technical decision the quality team must make based on each variable's impact on formulation and each analysis' cost. This prioritization is part of the laboratory's analytical program planning and must be done in collaboration with the formulation team, which knows which nutrients are most critical for each product line.

The role of laboratory analyses in formulation precision

Within the plant, the laboratory is the first point of contact with each received raw material lot. Sampling performed at the time of receiving, followed by bromatological and quality analyses, produces a dataset that, if well utilized, can feed formulation with information far more precise than any reference table can offer.

In practice, the most analyzed parameters in energy and protein ingredients include crude protein, moisture, ether extract, mineral matter, crude fiber, and in some cases amino acids by chromatography or NIRS prediction. From these values, it is possible to calculate or adjust the ingredient's metabolizable energy value for formulation use. Additionally, mycotoxin analyses, such as aflatoxins, zearalenone, deoxynivalenol, and fumonisins, are essential for ingredients like corn and grain by-products, given the direct impact of these contaminations on animal performance and health.

The ideal flow works as follows: the laboratory performs analyses of the received lot, consolidates results, calculates accumulated means and standard deviations for that ingredient from that supplier, and passes these values to the formulation team at a defined frequency. The formulation team, in turn, evaluates whether the found values justify updating the nutritional matrix used in the software and re-optimizes the formula if necessary.

This cycle, when well structured, functions as a continuous calibration mechanism between the reality of available ingredients and the formulas the plant is producing.

NIRS as a bridge between laboratory and formulation in real time

Near infrared spectroscopy, known by the acronym NIRS, occupies an increasingly relevant role in this context of integration between quality and formulation. The equipment allows estimating an ingredient's nutritional composition in seconds, from an optical sample scan, without the need for reagents and time involved in conventional wet chemistry analyses.

When combined with robust calibration curves developed for the operation's specific ingredients and conditions, NIRS allows raw material receiving analysis to be completed quickly enough to inform the use decision within the same production cycle. In short-cycle species like broilers, where animals gain weight by the hour and any nutritional deviation quickly accumulates impact, this agility has direct operational value.

Integration between NIRS and formulation software completes the circuit: the analysis result is transferred to the formulation platform, which recalculates optimization with the actual values of the available lot. This reduces the gap between formulated and produced feed and eliminates the cost of variability that goes unnoticed when using fixed table values.

Supplier qualification as a variability control strategy

One of the most valuable consequences of a well-structured laboratory is the ability to qualify suppliers based on accumulated analytical data over time. When analysis history by supplier is organized and accessible, it becomes possible to evaluate not only whether a specific lot is conforming, but whether a supplier consistently delivers the parameters commercially agreed upon.

A concrete example: a plant purchasing meat and bone meal with a 50% crude protein guarantee that, upon analyzing 30 samples over six months, finds 35% of them below the 45% tolerance minimum has a technical and commercial problem that needs addressing. With organized laboratory data, this problem can be identified, quantified, and presented to the supplier with concrete evidence, opening the possibility of renegotiating conditions, establishing formal improvement plans, or replacing the source.

Without the data, this situation usually generates only the vague perception that "that supplier is inconsistent." With the data, it becomes an evidence-based management decision.

Why quality and formulation still operate separately in most mills

If the integration between quality and formulation brings such clear gains, why is it still not the reality for most operations? The reasons are known and frequently recur in the sector.

The first is structural: the two areas frequently report to different managers, with objectives measured by different indicators. The laboratory is evaluated by response time and regulatory compliance. Formulation is evaluated by feed cost and nutritional precision. When there is no shared objective connecting the two metrics, information flow between areas happens by individual initiative, not by process.

The second reason is technological: data generated in the laboratory frequently stays in spreadsheets or systems that do not communicate with formulation software. The analyst records results in one place, the formulator accesses values in another. Matrix updating depends on someone manually bridging this gap, with the frequency and discipline the routine allows, which is rarely sufficient.

The third reason is cultural: in companies where the laboratory is treated as a compliance department rather than an intelligence department, formulation professionals rarely consult analytical data before reformulating. The perception that "table values are enough" persists as long as the cost of variability isn't measured and made visible.

Building quality-formulation integration in practice

Integrating the two areas does not require a radical operational transformation. It starts with practical, incremental decisions that create the habit of using real analytical data in the formulation process.

The first step is defining which ingredients will have their nutritional matrices updated based on in-house analyses and at what frequency. Not all ingredients need continuous updates at the same cadence. High-volume, high-variability ingredients should have priority. Marginally-used or low-variability ingredients can be maintained with table values updated semi-annually or annually.

The second step is standardizing the format in which laboratory data is consolidated and transferred to the formulation team. Means, standard deviations, and number of samples analyzed per period are the minimum data needed for the formulator to make an informed decision about matrix update. This periodic report, monthly or bi-weekly depending on analysis volume, is the practical link between the two areas.

The third step is using tools that allow this flow to happen with minimum friction. Laboratory platforms like Labinfy allow consolidating analytical data by ingredient and supplier, calculating descriptive statistics, and exporting values in format compatible with formulation software. On the other side, platforms like Formulamix allow receiving these values and recalculating optimization with updated matrices, closing the cycle between analytical data and formulation decision.

When this cycle functions systematically, the plant begins operating with a nutritional precision level that translates into more consistent products, more controlled formulation costs, and better customer animal performance. The gap between formulated and produced feed decreases. And the quality department, instead of being seen as a cost and bureaucracy sector, begins to be recognized as one of the main contributors to operational profitability.