R&D tax incentives for food and agriculture

Food, drink and agriculture businesses run trials all the time: a new recipe, a new variety in the field, a new feed regime, a new line speed. Much of that work is about what customers prefer, and no scheme treats preference as R&D. The dividing line is simple to state and hard to hold: a technical target measured by instrument, such as water activity, microbial count, shear force, yield or disease incidence, can be the subject of a scientific or technological uncertainty; a preference measured by a panel or by sales cannot. The same product can carry both kinds of project in the same year, and only one of them is a claim.

This guide follows the shape of the manufacturers guide, because most food businesses are manufacturers and most farms are process businesses, and it ends with each country's wording on the taste and cosmetic exclusion.

What this guide covers

• Where technical uncertainty sits in formulation, process, packaging and field work.
• What the taste, style and cosmetic exclusions remove, and how to separate a sensory measurement from a preference test.
• Evidence from the development kitchen, the plant and the trial plot.
• Costs often missed, including ingredients consumed in trials and the treatment of trial output that is sold.
• Each country's exclusion wording and where it is explained.

Where to look

Formulation to a technical target

Extending shelf life, removing a preservative, allergen or ingredient while holding texture and stability, reaching a nutritional profile, or keeping an emulsion or foam stable, where the available data did not predict whether it could be done. The failed formulations are the evidence.

Process development

Thermal, drying, fermentation, extrusion, high pressure or cold chain processes where the parameters for a new product or ingredient had to be found by structured trial, or where scale-up from the development kitchen to the line failed in ways that had to be investigated.

Packaging and product interaction

Barrier, modified atmosphere and migration behaviour of a new pack with a particular product, where the combined effect on shelf life was unknown. Pack graphics and shape for appearance are not.

Crop and livestock trials

Varieties, inputs, irrigation, soil treatments, husbandry and feed regimes tested with controls and measured outcomes under conditions where the result was not known. Normal growing and rearing around the trial is not part of it.

Agritech

Sensing, automation, modelling and control systems for farms and processors. The software rules of each country apply alongside the agricultural ones; see the industry examples module for how the software boundary is drawn.

Food safety interventions

Reducing pathogen load or spoilage with a new treatment or process step whose effectiveness on this product could not be predicted and had to be measured.

What to exclude

• Changes to taste, appearance, style, colour and seasonal variants.
• Consumer preference panels, market research and market testing.
• Routine quality control, release testing and shelf life confirmation of an unchanged product.
• Normal farming, growing and rearing operations.
• Labelling, regulatory and certification work.
• Cost reduction by substituting a known equivalent ingredient.

Sensory work needs a careful split. A trained panel scoring a defined texture attribute on a numbered scale as part of resolving a technical uncertainty is a measurement and belongs in the evidence. A consumer panel asked which sample it prefers is a preference test and belongs in marketing. The table shows how typical projects fall.

Evidence in food and agriculture

• A formulation log with numbered versions, the change made in each, the target and the result.
• Storage trial schedules with the pull dates, conditions and the measurements at each pull.
• Laboratory results for microbial, moisture, texture and nutritional tests, dated and linked to the version tested.
• Process trial sheets with the parameters set, the output measured and the operator's name.
• Trial plot plans and treatment group records: layout, randomisation, controls, inputs applied and dates.
• Yield, grading and animal performance data by treatment, with photographs where they help.
• Disposal and sale records for trial output, because several schemes adjust for output that is sold.
• A note at the start of each project stating what was unknown and why the available data did not answer it.

Costs often missed

Ingredients, raw materials, seed, feed and packaging consumed in trials are eligible costs in most schemes and are usually lost in cost of sales; separate trial consumption from production stock. Trial output that is sold is treated differently by country: the United Kingdom excludes the cost of consumable items incorporated into products that are sold or transferred, Australia applies a feedstock adjustment where R&D activities produce marketable products, New Zealand allows expenditure on goods or materials that become part of products sold or used only to the extent it exceeds the amount derived from the sale or use, and Canada applies recapture rules where materials transformed in the work are incorporated into property that is later sold. Trial runs on the production line are limited in several schemes: Canada counts only the experimental work and its incremental cost where production is commercial, New Zealand excludes preproduction trial runs, and Ireland excludes pilot plants that are not wholly and exclusively for R&D. Field and growing trials span seasons, and in Australia activities must be registered for each income year in which they are conducted, while in New Zealand activities not covered by an existing approval must be included in an application by the general approval due date for the year. External laboratories, universities and research providers enter at each scheme's rules for contracted work.

Getting started

Run a project identification workshop with the technical, operations and agronomy leads, and sort every project in the year into two columns: measured by instrument, or measured by preference. Drop the second column. For the first, ask whether the outcome could have been predicted from the available data; drop anything that could. What remains is usually a short list of shelf life, stability, process and trial projects, and the formulation logs and trial records already exist. Then set up the version numbering and trial sheets so that next year's evidence is captured as the work happens.

Illustrative example

Illustrative example. Brindle Mill Ltd, a hypothetical bakery supplier, ran two projects on the same loaf in one year. The first developed a new seasonal flavour, tested with consumer panels, at a cost of 40,000. It is not claimed: the question was what people prefer. The second set out to produce a gluten free loaf with a crumb firmness within a measured range and a mould free life of 10 days without the current preservative, where the published data on the alternative structurants and water activity did not predict the result. Over five months the technologists ran 80 trial bakes across 14 formulation versions, with texture analysis and microbial testing at intervals; 11 versions failed on texture or mould, and version 14 met both targets, at which point the work ended. The project cost 180,000: 120,000 of technologist time, 30,000 of ingredients consumed in trial bakes, 20,000 paid to an external laboratory for microbial testing, and 10,000 for a trial run on the production line, 4,000 of whose output was sold. The candidate base is 180,000 before each scheme's rules on the laboratory fees, the production trial and the 4,000 of sold output are applied; the benefit at that country's rate applies to what survives those rules. The three further months of line commissioning for the new loaf are outside.

Country differences

One rule per country, in the scheme's own wording, on taste, style and cosmetic work. The country pages hold the detail.

Related pages

• Guide: R&D tax incentives for manufacturers
• Technical Teams Program, Module 5: Industry examples
• Guide: qualifying versus non-qualifying activities
• Guide: how to run a project identification workshop
• In-House Finance Team Program, Module 2: Cost capture
• Resource hub: eligibility

Educational material, not advice. Eligibility and treatment depend on the scheme and the facts.