What a Weak Write-Up Looks Like to HMRC: An R&D Technical Narrative Example

An R&D claim now carries a written description of each project, and a first-time claimant has no earlier write-up to copy. An R&D narrative that describes the product instead of the problem invites questions, and the window to make or amend the claim closes two years after the end of the period of account (HMRC manual CIRD81800, checked 19 October 2026). The R&D technical narrative example below shows one project written badly and then well, as a guide to how to write an R&D technical narrative that answers HMRC's questions before it asks them.

What Is an R&D Technical Narrative?

The short answer to what is a technical narrative: it is the written account of each R&D project that explains why the work qualifies. It sets out the field, what was already known, what the company set out to achieve, what it could not work out and how it tried to.

The narrative is not filed with the company tax return. Since 8 August 2023, the project descriptions have to be entered on the additional information form for R&D claims before the claim is made (HMRC manual CIRD81800; GOV.UK, "Submit detailed information before you claim Research and Development (R&D) tax relief", last updated 26 September 2025). How the form is filed is a separate subject; this is about what goes into it.

The Same Write-Up, Different Names: Narrative, Technical Report and AIF Project Description

Anyone asking what is an R&D technical report is asking about the same document. Advisers call it the technical narrative, the technical report or the project description, and on the additional information form (AIF) it is the set of project details HMRC asks for. GOV.UK also accepts a separate R&D report covering the methodology, sampling and competent professionals, but it supplements the form's project details and does not replace them.

What the Write-Up Must Prove: Advance, Uncertainty and the Competent Professional

The question of what is R&D in technology has a statutory answer, and the write-up exists to evidence it. Under the 2023 DSIT Guidelines, which apply to accounting periods beginning after 31 March 2023 (HMRC manual CIRD81910), a project must show:

• An advance: it sought an advance in overall knowledge or capability in a field of science or technology, "not a company's own state of knowledge or capability alone" (CIRD81910, para 6).
• An uncertainty: whether the advance was possible, or how to achieve it, was "not readily available or deducible by a competent professional working in the field" (CIRD81910, para 13).

The competent professional is the yardstick for both tests. A problem an experienced engineer in the field could solve from published knowledge is not an uncertainty, however hard it was for the company, and "improvements, optimisations and fine-tuning" that leave the underlying technology unchanged do not qualify (CIRD81910, para 14).

Failed projects still count, because R&D takes place even if the advance is not achieved (CIRD81910, paras 10 and 38). A write-up that records the dead ends is evidence of the uncertainty. Whether borderline work qualifies at all belongs to the guide to everyday problem-solving versus qualifying R&D.

What HMRC's Guidance Says a Good Write-Up Contains

The HMRC guidance R&D claimants must follow sets out the structure. The GOV.UK additional information form page (last updated 26 September 2025) asks for five details for each project:

1. the main field of science or technology;
2. the baseline: the level of knowledge or capability when the project started;
3. the advance the company aimed to achieve, described against that baseline;
4. the scientific or technological uncertainties it faced, and why they apply to the industry and not just the company;
5. how the project sought to overcome them, and whether it did.

The same page warns that "a problem is not an uncertainty if it can be resolved in a discussion with peers".

Not every project needs a write-up. A claim for one to three projects describes all of them. For four to ten, the company selects three or more that together account for at least half of the qualifying expenditure. Above ten the same rule applies, but if reaching half would take more than ten projects, it describes the ten with the highest qualifying expenditure (GOV.UK, last updated 26 September 2025).

What a Weak Write-Up Looks Like to HMRC

A weak R&D narrative leaves HMRC unable to see the test being met, so it asks. Several weak signals together can lead to a compliance check, which the guide to what happens in an HMRC R&D enquiry and how to respond covers. Mistakes outside the write-up, from missed forms to the wrong scheme, are in the guide to why R&D tax relief claims get rejected.

Weak signal	The question it invites	The fix
Describes what was built	What was uncertain about building it?	Describe the problem the team could not solve
Commercial risk presented as uncertainty	Was the doubt technical, or about cost and market?	Keep only scientific or technological unknowns
Boilerplate that fits any company	What did this company actually do?	Name the system, the constraint and the methods
No baseline	Advance compared with what?	State what the field could do when work began
No systematic work	How did the team try to resolve it?	Record the approaches, tests and results
Padding and jargon	Which part of this is the R&D?	Cut to the five details, in plain words
Written long after the work	Is this how it was understood at the time?	Draw on records made during the project

On commercial risk, the guidelines place a project's "financial, marketing, and legal aspects" outside R&D (CIRD81910, para 37). A weak write-up also costs time, because the two-year window to make, amend or withdraw the claim runs from the last day of the period of account where that period is not longer than 18 months (FA 1998 Sch 18 para 83E, as set out in CIRD81800).

Worked Example: A Weak Passage and the Rewrite

This R&D technical narrative example uses a fictional project. No R&D tax credit claim template for SMEs can supply the detail in the rewrite, because that detail belongs to the project.

Weak version: "We developed a new monitoring platform for our clients' production lines. We needed to process data quickly and it was challenging. Our talented team used cutting-edge technology to overcome these challenges and deliver an innovative solution."

That passage has no baseline, no uncertainty and no account of the work, and "challenging" says nothing about what a competent professional could deduce.

Rewrite, same project:

• Field: software engineering for embedded signal processing.
• Baseline: published stream-processing methods could filter machine vibration data at the required rate on server hardware, but not within the memory of the controllers fitted to the client's machines.
• Advance sought: filtering on the controller that kept the vibration patterns used to predict bearing failure.
• Uncertainty: the senior engineer could not determine from published methods whether reducing the data to fit the controller's memory would remove those patterns.
• Work done: the team tested three reduction methods against recorded failure data. Two removed the patterns and were abandoned; the third kept them once its sampling stage was redesigned.

How to Write a Write-Up HMRC Will Accept

To write an R&D technical narrative that holds up, take each project in turn:

• Get the technical lead to write or review it. The uncertainty test is judged against what a competent professional could readily deduce (CIRD81910), which a finance lead cannot answer for alone.
• State the project's dates and scope. The guide to where an R&D project starts and stops sets the boundary; the write-up records it.
• Write the baseline before the advance, so the advance has a starting point to be measured against.
• Name each uncertainty precisely, with why a competent professional in the field could not resolve it.
• Record the work, including the failures: the approaches tried, what each showed and what was abandoned.
• Write from records made at the time. An R&D narrative reconstructed a year later is weaker than one drawn from contemporaneous notes and test logs; the guide to R&D record keeping covers how to capture them.
• Tie each project to its costs. The costs claimed should match the work described, and the guide to missing costs in an R&D claim covers what qualifies.

A write-up that fills all five details evidences the test but does not prove the work qualifies. The project itself still has to meet the advance and uncertainty test in CIRD81910.

R&D Technical Narrative FAQs

What is an R&D technical narrative?

It is the written account of an R&D project that explains why the work qualifies for relief. It covers the field, the baseline, the advance sought, the uncertainties and how the company tried to resolve them, and since 8 August 2023 those project details go on the additional information form (CIRD81800).

What is an R&D technical report?

It is the same document under a different name. Advisers use technical narrative, technical report and project description interchangeably, and GOV.UK also allows a separate R&D report with extra detail such as methodology and sampling alongside the form.

What does HMRC look for in an R&D technical narrative?

HMRC looks for evidence of an advance in overall knowledge or capability in a field of science or technology, achieved by resolving uncertainty a competent professional could not readily resolve (CIRD81910, paras 6 and 13). The GOV.UK guidance asks for that evidence in five details: field, baseline, advance, uncertainties and how they were overcome.

What makes an R&D technical narrative weak?

A weak narrative describes what was built instead of what was uncertain. Other signals are commercial risk presented as technical uncertainty, boilerplate, a missing baseline, no account of the work done and a write-up reconstructed long after the project.

What happens if my R&D technical narrative is weak?

HMRC is more likely to ask questions, and a write-up that cannot answer them can lead to a compliance check. The claim can still be amended within two years of the end of the period of account where the period is not longer than 18 months (FA 1998 Sch 18 para 83E, CIRD81800).

What is technological uncertainty?

It exists when whether something is technologically feasible, or how to achieve it in practice, is not readily available or deducible by a competent professional working in the field (CIRD81910, para 13). Difficulty for the company alone, or doubt about cost and market, does not count.

How long should an R&D technical narrative be?

HMRC's additional information form guidance sets no page length; it asks for enough detail to cover the five project details. Kene Partners, an R&D adviser, recommends one to two pages per major project (updated 19 February 2026), which is a guide rather than a rule.

How many projects should I write up in my R&D claim?

All of them when the claim covers one to three projects; four to ten need three or more that together account for at least half of qualifying expenditure; above ten, the same applies, capped at the ten with the highest qualifying expenditure (GOV.UK, last updated 26 September 2025).

Who should write the R&D technical narrative?

The technical lead who did or ran the work should write it or review it. The uncertainty test is judged against what a competent professional in the field could readily deduce (CIRD81910), which an accountant or finance lead cannot judge alone.

How long do I have to make an R&D claim?

Two years from the last day of the period of account, where that period is not longer than 18 months; otherwise 42 months from its first day (FA 1998 Sch 18 para 83E, CIRD81800). The claim can be made, amended or withdrawn within that window.

Conclusion: Get the Write-Up Right Before You Claim

Rework a weak write-up before the company makes its claim, while the technical lead still remembers the dead ends and the window to claim r&d tax credits is open. When the narrative is ready, run the claim through the R&D claim estimator.