Why your input matters to an R&D claim

After this module you will be able to explain what the tax authority actually reads when it assesses an R&D claim, why a finance team cannot produce that material on its own, and where a small amount of your time changes what the claim is worth and whether it survives review.

What this module covers

What the tax authority reads, and the questions behind every form.
Why the people who did the work are the only reliable source for the technical content.
The three points in a claim where technical input changes the result.
What a few hours of technical time looks like, and what happens without it.

What the tax authority reads

Every one of the seven schemes asks the claimant to describe the work in writing, at a registration or approval step, with the claim or on review, and keeps the right to ask for the records behind that description. The forms differ in headings and limits, but the reader behind every form is trying to answer the same questions:

Was there a real technical unknown at the start?
What did the team do to resolve it, and was that work systematic?
Did the people whose costs are claimed do the work, and when?
Where did the work to resolve the uncertainty end?
Do the costs tie to the work?

Reviewers typically read the narrative first and then the earliest dated document for the project, to see whether the narrative was written from the evidence or the evidence was arranged to fit the narrative. In a deeper review several authorities ask to speak to the technical staff, and some bring in an expert in your field. The narrative is judged against what a competent professional in that field would have known at the time.

Why finance cannot write it alone

The finance team owns the numbers, the deadlines and the forms. It does not know which of the forty tickets in a sprint were a genuine unknown and which were routine, and it cannot find out from the ledger. Four things make the technical content yours to supply.

The test is judged in your field. Whether something was uncertain is judged against what a competent professional in the field could have known or worked out at the start. Only such a professional can make that judgement.
The boundary is technical. The qualifying work ends when the unknown is resolved or abandoned, not when the product ships. Only the people doing the work can say when that happened.
The names must match. Reviewers expect the people in the cost schedule to appear in the technical record, and they interview them. A technical lead who does not recognise the narrative tells a reviewer the claim was assembled rather than evidenced.
Advisors edit, they do not originate. An external advisor can structure the claim, test it against the definition and make it consistent with the forms.

A good claim is a joint product: technical content from the people who did the work, structure and numbers from finance.

Question the reader is asking	Who can answer it	What goes wrong when finance answers alone
What was unknown at the start?	The technical lead	The narrative describes the product; the unknown is implied, never stated
Why could it not be looked up or worked out?	An engineer or scientist in the field	"Nobody had done this before", with no account of what was searched, consulted or tried
What was tried, and what failed?	The people who ran the tests	Only the final design is described; the failures are left out
Who did the work, and when?	The team lead, from tickets, commits and notebooks	Round percentages applied to whole departments
Where did the uncertainty end?	The technical lead	The claim runs to release or year end and sweeps in routine build and polish
Do the costs tie to the work?	Finance, once the project boundary is set	Costs are allocated to a project name nobody in engineering recognises

Three points where your input shapes the claim

1. Identifying the projects

Ask a finance team which projects contain R&D and it will name the big, visible ones. Ask the engineers and they will name the problems that kept them up at night, which are usually smaller and far more likely to qualify. The candidate list should start with you, and it should record the candidates you rejected and why.

2. Drawing the boundary

For each candidate, state when the unknown was recognised, when work to resolve it started, and when it was resolved or abandoned. Work inside that window that contributed to resolving it is a candidate for the claim; work outside it, however hard, is not. The boundary decides the size of the claim more than any other judgement, and it is a technical one.

3. Writing and confirming the narrative

The person who led the work drafts or dictates the narrative, in their own words, referring to dated documents by name, and confirms every sentence before filing. Under review that person will be asked to explain the uncertainty; a narrative they wrote is one they can defend.

What a few hours looks like

A dated half-page note at the start of each project stating what is unknown and why.
A project key on the tickets, commits, notebooks and test logs you already create, so finance can find them.
A short candidate review each quarter with finance, while the work is fresh.
One drafting session per project for the narrative, and one read-through before filing.

Without it, the narrative is written from the marketing deck, the whole cost of a commercial project is claimed, and the first reviewer to ask for the earliest dated document finds nothing older than the claim. Over-claimed amounts are repaid, usually with interest and penalties can follow. A smaller claim built on records made at the time is worth more than a larger one that cannot be defended.

Illustrative example

Illustrative example. Harbour Robotics Ltd, a hypothetical company, employs 10 engineers at a total cost of 1,000,000 a year. Finance's first draft claims the whole "Warehouse Robot v3" product at a cost of 600,000, with a narrative taken from the product sheet. Two technical leads spend one afternoon with the finance lead. The only genuine unknown, they find, was whether a grasp planner could reach a 99% pick success rate at a four second cycle time on partially occluded items with the company's two-finger gripper; the two published approaches evaluated reached 91% and 94%. That work ran from March to August and ended when the test rig met the target over a 2,000 pick run; it cost 300,000. Path planning used a published algorithm and the dashboard, user interface and integrations were routine, so none of that is claimed. The leads also mention a motor thermal model in the drives team, unknown to finance, which cost 100,000 and had its own uncertainty. The claim is 400,000 rather than 600,000, every line of it can be explained by someone who was there, and the benefit at that country's rate rests on evidence rather than a product sheet.

Across the seven countries

The headline in each country's own words; the country pages hold the detail.

Country	What the tax authority reads	Country page
United States	Keep records that link each business component to the four-part test and to the wages, supplies and contract research costs you claim. For tax years beginning after 2025, Section G of Form 6765 asks for information about each business component, with some exceptions.	Documentation: United States
United Kingdom	Every claim needs an additional information form that describes the R&D projects and breaks down the qualifying costs.	Documentation: United Kingdom
Ireland	Keep records made at the time that show the scientific or technological uncertainty, the work carried out to resolve it and the expenditure incurred.	Documentation: Ireland
Canada	Form T661 asks you to describe the work and calculate your expenditure. Keep records made as the work happens, such as hypotheses, test results and time records, that show the systematic investigation.	Documentation: Canada
Australia	The department asks registrants to describe each core R&D activity by its hypothesis, experiment, observation and evaluation, and logical conclusions.	Documentation: Australia
New Zealand	Approval comes before the claim. The general approval application asks for each core activity: the scientific or technological uncertainty, the new knowledge sought, the systematic approach and the start and end dates.	Documentation: New Zealand
Singapore	Keep documents that show each project's objective, its novelty or technical risk, the work done and the qualifying expenditure.	Documentation: Singapore

Module 1: Why your input matters