R&D tax relief •12 min read

Flame Tree Publishing Ltd v HMRC

MJ

Matthew Jones ACA CTA

Last reviewed September 2026 · Editorial standards

Tribunal
First-tier Tribunal (Tax Chamber)
Decided
25 April 2024
Judge
Judge Anne Redston and Mr Michael Bell

A publisher claimed £266,644 of enhanced R&D expenditure for digitising its back catalogue. The appeal was dismissed on two independent grounds: nobody put forward as a competent professional was qualified in software or computing, and the costs claimed were never evidenced. A third problem sat underneath both — the programming that might have qualified was done by an outside supplier, under a statutory route the claim had not been made on. Three later tribunals have applied it on the competent professional: Tanglewood Care Services, Beer Express and Tills Plus.

What was at issue

Flame Tree Publishing is a music and art publisher, set up by its owner in 1992. Its accountants filed the corporation tax return for the year ended 30 June 2018 on 8 March 2019, showing profits of £344,994 and tax of £65,548.96, with no reference to R&D. On 30 June 2020 a specialist adviser filed an amended return claiming R&D expenditure of £205,111 — £147,292 of employee costs, £5,280 of subcontracted work, £49,429 of other costs, and the balance in utilities and software. That produced an enhanced deduction of £266,644 and reduced the tax payable by £50,662.

The claim described seven projects, among them a hybrid server archive, digital tools, an e-commerce payment system and a foiled notebooks project. What the company had actually done was convert its historic archive into machine-readable form, organise the data, and build a search process so customers could use it as a curated resource (paragraph 57). One fact ran underneath everything: “The entirety of the programming, computing, and the building of the customised search engine was carried out by Image File” — an outside website developer — “and no part of those costs are included in the claim” (paragraph 57(4)).

Three questions reached the tribunal. Was this one project or seven? Did the people relied on count as competent professionals in the relevant field? And was there evidence that the money claimed had been spent on the work?

How the dispute got to a hearing

HMRC opened an enquiry into the amendment on 18 May 2021 and asked why each project met the Guidelines, for project documentation and third-party contracts, and for the basis of the costs. The adviser answered on 17 June 2021, and three sentences from that reply, recorded at paragraph 25, decided the case. On documents: “there are no specific documents for reference … The R&D was generated internally and carried out as an ongoing activity throughout the period”. On costs: the time had been estimated “because no time recording system was in place”, covering 12 of the company’s 18 employees and amounting to 24.15% of total staff costs. On contracts: “no contracts were entered into in relation to R&D activities”.

Job titles and percentages followed on 21 August 2021. The case officer changed twice, in January and March 2022, and technical points went to HMRC’s in-house Chief Digital and Information Officer group (paragraph 29). At a meeting on 10 March 2022 HMRC asked for each claimed employee’s expertise in software; the answer, on 5 April 2022, described editorial and design roles instead.

A closure notice increasing the tax payable by £50,662 was issued on 1 July 2022. The company appealed on 19 July, asked for a statutory review, and the review officer upheld the decision on 10 November 2022. The appeal was notified to the tribunal on 29 November 2022, heard on 19 and 20 March 2024, and released on 25 April 2024.

What each side argued

HMRC’s case was put by Mr Dave Lewis. The Guidelines require evidence from a competent professional. The term is not defined, but “its natural meaning is self-explanatory, and that it goes beyond having an intelligent interest in the field”. To be accepted as one, “an individual would need to be able to demonstrate appropriate qualifications, experience and up-to-date knowledge of the relevant scientific and technological principles involved” (paragraph 66). Because the claimed advance lay in digitising and searching an archive, a competent professional here “would have up-to-date software, programming and computing skills and knowledge”, and neither witness had them (paragraph 67). The company had shown only that the technology was new to itself, which paragraphs 22 and 24 of the Guidelines exclude. Separately and independently, there was no evidential basis for the costs (paragraphs 70 and 71).

The company opened on construction: guidelines “are not rigid instructions, and a measure of interpretation is implied” (paragraph 43), a point conceded in reply. At the hearing its case shifted. The owner gave oral evidence that the company had “developed a single solution”, and expressed “regret” that it had “boxed ourselves into discussing it” as seven projects (paragraph 49). Its representative then argued that the combined result was a “Software as a Service” amounting to “a novel advance in technology” for the publishing sector (paragraph 50). The owner was said to be a competent professional with computer expertise. The production manager was said to be “a competent professional in digital production and publishing – with a focus on digital formats, extraction and data structure” (paragraph 64). The staff costs were said to rest on a reasonable apportionment of time.

HMRC did not object to the change of position, and expressly did not ask the tribunal to find that the company had been careless or deliberate in filing the claim (paragraph 52).

The provision the case turns on

The claim was made under section 1052 CTA 2009, as it stood for a period ending on 30 June 2018. That is the in-house route: “A company’s ‘qualifying expenditure on in-house direct research and development’ means expenditure incurred by it in relation to which each of conditions A, B, D and E is met.” Condition B is what shuts the supplier’s work out: “Condition B is that the expenditure is attributable to relevant research and development undertaken by the company itself.” Work done by an outside supplier belongs, if anywhere, in section 1053 — expenditure “incurred by it in making the qualifying element of a sub-contractor payment” — which carries its own conditions. No claim was made under it. (The tribunal reached the same destination through Condition D at paragraph 81(2).)

The phrase that decided the case appears in neither section. “Competent professional” is not a statutory term at all: it is used in the Guidelines, which take effect as tertiary legislation through section 1006 ITA 2007 and regulation 2 of SI 2004/712, and which the tribunal recorded therefore “have the force of law” (paragraph 41). Paragraph 13 defines scientific or technological uncertainty as existing where the knowledge in question “is not readily available or deducible by a competent professional working in the field”. Paragraph 23 measures an appreciable improvement by what “would generally be acknowledged by a competent professional working in the field as a genuine and non-trivial improvement”. HMRC restates the point at CIRD81300, amended 19 January 2026.

What the tribunal decided

HMRC won, on every point.

On construction, the tribunal followed Gripple at [12] — “a detailed and prescriptive code of this nature leaves little room for a purposive construction” — as Hadee Engineering had done (paragraphs 44 to 46).

On the boundaries, it found a single project (paragraph 55). No start or end date had ever been given for any of the seven; no costs had ever been allocated between them; and the company had told HMRC there were no specific documents about any of them (paragraph 54). The explanation offered for the change of position was rejected outright: “We reject these explanations, and find that none of the responsibility rests with HMRC” (paragraph 56). Seven projects would not have saved the claim either (paragraph 79).

The competent professional point did the work. The tribunal had “no hesitation in agreeing with HMRC, for the reasons given by Mr Lewis”, and added four of its own (paragraph 68). The owner was “a competent professional in the field of publishing who had some knowledge of IT and computing, but he was not a competent professional in software, programming or computing”. The production manager was “a competent professional in digital archiving” but had “no IT or software qualifications and no software or programming skills”. Ten other employees whose time was claimed brought “publishing judgements”, and none gave evidence. The finding that generalises beyond the facts is at paragraph 68(2). The owner had said one technology had “probably” been used in a similar way before but did not know, and that a search company “must have already had” a capability. The tribunal’s answer: “A competent professional in the relevant field could reasonably have been expected to be able to give definitive evidence on both points.” Paragraph 69 draws the consequence: the company “has failed to show that the Project had either (a) resolved uncertainties which could not have been resolved by a competent professional or (b) made an improvement which a competent professional would have acknowledged as being ‘non-trivial’”.

The costs failed separately. There was “no evidential basis for the quantum of any of the employee costs claimed” (paragraph 73). There was no time recording system, and the percentages had been agreed with an adviser who “was not involved in the Project”. A 50% allocation for the owner was “not credible” given his wider responsibilities. The production manager had “no recollection” of the time he had spent. A 30% allocation for the art director came with no explanation, and there was “no information at all” about the publishing assistant or the designer. The subcontractor figure had “no explanation as to how the 70% figure had been arrived at” (paragraph 74), and a 1.5% slice of cost of sales carried “no explanation … for that percentage” (paragraph 75). The appeal was refused “for the additional reason that it has failed to prove that the sums claimed were spent on the Project” (paragraph 76).

Finally the route. Because the claim was made under section 1052, it could not reach the work the supplier had done. The tribunal could have considered that work only “had FTP had made an entirely different claim under CTA s 1053, supported by evidence from Image File” (paragraph 82).

What it changes for a claim being prepared now

  • Name the field first, then find the person. If the advance is in software, the competent professional must be qualified and current in software. A director who has run the business for decades is a competent professional in the business, which is not the same thing.
  • Test the person on the question they will actually be asked. The tribunal’s sharpest point was not about qualifications: a competent professional could have given definitive evidence about what was already available in the field, and this witness could only say “probably” and “must have already had”. Ask your witness what the published state of the art was in the claim period. If the answer is a guess, they are not the right witness.
  • Record time while the work happens. Percentages reconstructed two years later, with an adviser who was not there, are not evidence. Timesheets, sprint records, project codes in the payroll — anything contemporaneous beats anything remembered.
  • Explain every apportionment in writing, at the time. One line per figure is enough. On invented facts: the lead developer logged 142 hours against project code DT-17 in the year to 31 March, of a 1,830-hour year — 7.8%, applied to salary, employer’s NIC and pension.
  • Check which statutory route the claim is on. Work done by an outside supplier is not in-house R&D, and claiming it under the wrong section is not corrected at the hearing.

Who counts as a competent professional takes the first two points further; the records a claim needs covers the third and fourth.

Where it sits against the other decisions

Flame Tree is the decision later tribunals reach for. Tanglewood Care Services applied it at paragraph 112, refusing a care operator’s claim for want of anyone who could speak to the field. Beer Express Ltd [2026] UKFTT 672 (TC) set out paragraphs 65 to 69 in full and adopted them at paragraph 91. Tills Plus Ltd [2024] UKFTT 614 (TC) adopted the same description at its paragraph 121, taking it from Flame Tree at [66] and [68]. Behind it stands AHK Recruitment Ltd [2020] UKFTT 232 (TC), where the tribunal found it “remarkable that the Appellant did not provide evidence from someone that was contemporaneously involved in the project … and/or from someone with relevant expertise” (paragraph 73). Hadee Engineering supplies the strict-construction step through Gripple.

The one decision that pulls the other way is Get Onbord, decided three months later on the same HMRC submission by the same litigator. It accepted a witness without formal qualifications, distinguished Flame Tree on the facts at paragraph 82, and expressly reserved the question Flame Tree answered — whether evidence from a competent professional is required at all (paragraph 84).

One further decision shows the practical risk. In M-Sport Wheels Ltd [2026] UKFTT 977 (TC), an R&D appeal was withdrawn in February 2026 because the company could not produce a signed witness statement (paragraphs 2, 4, 15 and 17). The witness it first wanted had changed jobs, and his current employer would not agree to him giving one (paragraphs 8 and 9). The replacement, based in China, supplied an unsigned statement and never a signed one (paragraph 11). The tribunal accepted that the difficulties “arose from external constraints, including employer restrictions and corporate approval processes” (paragraph 30). HMRC’s costs application of £6,202.07 was refused — “Repeated procedural failures, without more, do not automatically amount to unreasonable conduct for the purposes of Rule 10” (paragraphs 24, 38 and 47) — but the claim was gone. The person who did the work is the claim’s single point of failure, and people move on.

Has it been appealed?

No onward appeal has been reported. There is no Upper Tribunal decision in the case on the published record, and later tribunals continue to cite it as it stands. Applications for permission to appeal are not consistently published, so absence of a reported appeal is not proof that none was sought. HMRC has not changed its published guidance as a result: nothing in the CIRD manual’s dated change log is attributed to this decision, and no Revenue and Customs Brief or Agent Update names it.

The decision is First-tier Tribunal, so it binds only the parties and sets no precedent. Its weight is practical: it is the decision HMRC officers and later tribunals reach for on what a competent professional has to be qualified in. It sits with the rest of the R&D tax case-law register, one page per decision.

Sources

Every CIRD paragraph cited above (CIRD81300 and CIRD83800) is indexed on our CIRD reference index, with HMRC’s own title for each, a line on what it says and the date HMRC last revised it.

First-tier Tribunal decisions bind only the parties to them and set no precedent. This entry describes the decision as it stood at the review date above, as general information rather than advice on your circumstances — see our terms. For an answer on your own facts, talk to us.