[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback] [DONATE]
	First-tier Tribunal (Tax)
You are here: BAILII >> Databases >> First-tier Tribunal (Tax) >> Realbuzz Group Ltd v Revenue and Customs (CORPORATION TAX - DISCOVERY ASSESSMENT - research and development relief - whether officer should reasonably be expected to have been aware of excessive claim for relief when enquiry window closed - what information was made available to the officer?) [2025] UKFTT 493 (TC) (01 May 2025) URL: https://www.bailii.org/uk/cases/UKFTT/TC/2025/TC09502.html Cite as: [2025] UKFTT 493 (TC)

[New search] [Contents list] [Printable PDF version] [Help]

FIRST-TIER TRIBUNAL
TAX CHAMBER

		Taylor House, 88 Rosebery Avenue, London EC1R 4QU
		Heard On: 28 March 2025 Judgment Date: 1 May 2025

CORPORATION TAX – DISCOVERY ASSESSMENT – research and development relief – whether officer should reasonably be expected to have been aware of excessive claim for relief when enquiry window closed – what information was made available to the officer?

DECISION

Introduction

1. This is the Appellant's appeal against a discovery assessment issued on 1 June 2023 disallowing a Research and Development (R&D) Relief claim in the sum of £335,452.57 for the accounting period ended 30 April 2020 (APE 2020). The claim was made in an amended corporation tax return received on 31 March 2021.
2. It is common ground that the officer who made the assessment (who gave oral evidence at the hearing) made a relevant discovery and that the R&D relief which had been given was or had become excessive.
3. The sole ground of appeal is that HMRC were not entitled to raise the discovery assessment as an officer "could have been reasonably expected, on the basis of the information made available to him, to have been aware [of the excessive relief] before he ceased to be entitled to give notice of enquiry" into the return for APE 2020.
4. A secondary issue is what constituted the "information made available" to the officer and, in particular, whether it included the R&D Report provided on 15 November 2021 after HMRC opened an enquiry into the Realbuzz Group Ltd's (Realbuzz/the Company) corporation tax return for the accounting period ended 30 April 2021 (APE 2021).
5. Statutory references are to schedule 18 to the Finance Act 1998 unless otherwise specified.

The procedural history

6. On 22 December 2020, Realbuzz filed its corporation tax return for APE 2020. It filed an amended return on 23 March 2021 and a further amendment on 31 March 2021. This last amendment included a claim for R&D relief of £335,452.57. The claim was supported by an R&D report which accompanied the amended return. The report (the 2020 Report) was prepared by the Company's accountants, Cowgill Holloway LLP (Cowgills) and specifically by Cowgills' R&D team following extensive discussions with the Company's technical team and an assessment of the individual projects undertaken in the period to ascertain whether they qualified for R&D relief. In making this assessment, Cowgills considered the tax legislation, the BIS Guidelines (which we refer to below) and HMRC's published guidance in their Manuals.
7. Realbuzz filed its corporation tax return for APE 2021 on 13 July 2021. HMRC opened an enquiry into that return, which also included a claim for R&D relief, on 17 September 2021. On 15 November 2021 Cowgills responded to HMRC's questions and sent an R&D report for APE 2021 (the 2021 Report) to HMRC. The 2021 Report was similar in format to the 2020 Report and included a number of projects which had begun in 2019 and which had also been included in the 2020 Report.
8. On 13 April 2022, the officer then dealing with the enquiry, a Mr Patel, wrote to Cowgills stating that most of the projects did not fall within the definition of R&D for tax purposes. He considered that there were a few sub-projects which might qualify. Mr Patel invited Cowgills to explain why they thought that the projects which Mr Patel had rejected in fact qualified. He sent a questionnaire to be completed in relation to specific projects and indicated that he would then obtain the opinion of HMRC's software specialists.
9. 30 April 2022 was the last date for HMRC to open an enquiry into the APE 2020 return (LDE).
10. Ms Martin, who gave oral evidence at the hearing took over the case from Mr Patel in September 2022.
11. On 8 March 2023 HMRC closed the enquiry into the 2021 return denying the R&D claim for APE 2021 and reducing it to nil. In the same letter, Ms Martin said she was considering raising a discovery assessment for APE 2020 on the basis that some of the 2021 projects began in 2019 and so she believed that the inaccuracies identified in 2021 would also have occurred in 2020 as R&D relief had been claimed in the earlier period on the same or similar projects and on the same basis as in 2021.
12. The discovery assessment was issued on 1 June 2023.
13. The Company appealed the assessment on 30 June 2023. Following a review which upheld the original decision, the Company made an in-time appeal to the Tribunal on 30 November 2023.

What is R&D?

14. The Corporation Tax Act 2009 (CTA 2009) makes provision for R&D relief.
15. The definition of "research and development" is contained in section 1138 Corporation Tax Act 2010, which, less than helpfully, provides broadly that "research and development" means "activities that fall to be treated as research and development in accordance with generally accepted accounting practice". The Income Tax Act 2007 section 1006 provides for the Secretary of State to issue guidelines on the meaning of R&D. Pursuant to section 1006, the Department for Business, Innovation and Skills has issued "Guidelines on the Meaning of Research and Development for Tax Purposes. The version in force at the relevant time is that updated on 6 December 2010 (the Guidelines). The most important of the Guidelines for present purposes are set out below. The words in bold are defined elsewhere in the Guidelines.

"3. R&D for tax purposes takes place when a project seeks to achieve an advance in science or technology.

4. The activities which directly contribute to achieving this advance in science or technology through the resolution of scientific or technological uncertainty are R&D….

6. An advance in science or technology means 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). This includes the adaptation of knowledge or capability from another field of science or technology in order to make such an advance where this adaptation was not readily deducible.

7. An advance in science or technology may have tangible consequences (such as a new or more efficient cleaning product, or a process which generates less waste) or more intangible outcomes (new knowledge or cost improvements, for example).

8. A process, material, device, product, service or source of knowledge does not become an advance in science or technology simply because science or technology is used in its creation. Work which uses science or technology but which does not advance scientific or technological capability as a whole is not an advance in science or technology.

9. A project which seeks to, for example,

(a) extend overall knowledge or capability in a field of science or technology; or

(b) create a process, material, device, product or service which incorporates or represents an increase in overall knowledge or capability in a field of science or technology; or

(c) make an appreciable improvement to an existing process, material, device, product or service through scientific or technological changes; or

(d) use science or technology to duplicate the effect of an existing process, material, device, product or service in a new or appreciably improved way (e.g. a product which has exactly the same performance characteristics as existing models, but is built in a fundamentally different manner)

will therefore be R&D. …

12. However, the routine analysis, copying or adaptation of an existing product, process, service or material, will not be an advance in science or technology.

13. Scientific or technological uncertainty exists when knowledge of whether something is scientifically possible or technologically feasible, or how to achieve it in practice, is not readily available or deducible by a competent professional working in the field. This includes system uncertainty. Scientific or technological uncertainty will often arise from turning something that has already been established as scientifically feasible into a cost-effective, reliable and reproducible process, material, device, product or service.

14. Uncertainties that can readily be resolved by a competent professional working in the field are not scientific or technological uncertainties. Similarly, improvements, optimisations and fine-tuning which do not materially affect the underlying science or technology do not constitute work to resolve scientific or technological uncertainty."

16. Software development is clearly within the realm of technology. It is also clear that R&D requires an overall advance in knowledge or capability in a technological field which includes a significant improvement to existing technology. The advance is to be achieved by resolving "technological uncertainty" which exists where it is not known whether a scientific or technological goal is possible or if it can be achieved in practice. Where a competent professional in the field could readily deduce the answer, there is no "uncertainty"
17. It is also important to note that research and development in the general sense which only increases the particular company's knowledge and capability is not R&D for tax purposes.

The facts

18. Realbuzz was established in 2000. It focusses on the fitness, sports and running industries, providing online content, social networking and, importantly, online entry systems for events such as the London Marathon and related charity fundraising programmes. It also enables participation in "virtual events" where the participants take part in, for example, a marathon or half marathon, running on their own in their own area rather than joining an organised mass participation event.
19. It is an SME for the purposes of the R&D relief provided for in schedule 20 Finance Act 2000. The applicable rate of relief in APE 2020 was 230%.
20. The 2020 Report states that in APE 2020, the directors of Realbuzz, in conjunction with Cowgills, identified ten projects which were considered to qualify for R&D tax relief in accordance with the Guidelines and legislation. The 2020 Report sets out, in some detail, what was involved in each of eight projects. We understand that HMRC practice is to request information on a sample of projects rather than all of them. The form of the Report was the same for each project, setting out:

(1) The title of the project

(2) In a prominent position under the title, the commencement date and the end date. In each case, the "end date" was described as "Ongoing".

(3) The aim of the project

(4) The technical uncertainties faced and overcome. This included an account of the activities undertaken and, in some cases, was divided into "sub-projects" under separate headings.

(5) The outcome of the project. This included what had been achieved to date and further activities to be carried out in the future.

21. The projects are summarized below.

Project 1-Amazon Web Services (AWS) Security Posture Enhancements

22. AWS is the platform which Realbuzz uses to provide its online services. In order to comply with the UK GDPR rules and international data privacy laws it wished to improve its understanding of its current security position. Having identified the risk areas, it researched the software products on the market which would allow the company to address the identified risks. Ultimately, the Company decided to use products offered by AWS.
23. The next stage of the project was to enhance further the Company's AWS defences and threat intelligence and aligning its security with industry standards.

Project 2-Product Development and Integration

24. The Company wished to integrate its four main products which were:

(1) Its main website, realbuzz.com

(2) Its virtual events website

(3) Its registration software which enables participants to sign up for events such as marathons

(4) Its charity consoles which allow event organisers to manage their charity programmes.

25. In order to achieve integration, the company identified a number of efficiencies:

(1) The need to introduce automatic testing systems to ensure financial accuracy. The Company used existing software products but needed to research and experiment with software with which it was unfamiliar in order to implement it.

(2) The Company developed a series of tools to manage its database. In particular, the Company needed to produce anonymised data so that its developers could work on it without breaching GDPR. There was no off-the-shelf product which was suitable for the Company's needs and so it developed its own bespoke anonymisation system.

(3) Each client has a separate entry management console which is independent of other consoles. This means updates, upgrades and client requested features have to be dealt with on a case-by-case basis which is inefficient. The aim was to integrate the consoles so global updates and upgrades could be applied. This project was put on hold

(4) The Company needs to maintain separate records for each user which allows it to provide accurate data which feeds into the results scoring/ranking system. This involves hundreds of thousands of database objects. The Company was using an AWS product but researched how the third party software worked to see if they could produce something better or whether they could improve functionality to do the same thing faster and more efficiently.

(5) The Company needed a detailed breakdown of the costs of running the AWS systems in order to bill clients correctly and analyse the costs to increase financial efficiency. AWS does not provide this breakdown. The Company undertook a web project which enabled it to extract the AWS information and import it into its own systems for analysis. It believed there was nothing in the market which would provide this level of detail on billing for AWS services.

26. The project enabled the Company to expand its commercial advertising inventory and laid the foundations for a subscription-based training model which are two vitally important revenue streams for the Company

Project 3-AWS User Access Key Rotation

27. The Company identified a security vulnerability as user access keys were not being periodically rotated, which is industry best practice. The research entailed reviewing guides and documentation provided by AWS. The Company ultimately decided to use an AWS product which met its needs. The project enhanced the Company's security by reducing the risk of unauthorised access to its systems and better aligned the Company's security with industry standards.

Project 4-Database Backups and Anonymisation

28. The Company needs to back up the databases from its live websites and sought to supplement the existing solution provided by AWS to reduce cost and improve efficiency. It also sought to improve the ability of the Company's developers to work effectively on the up-to-date databases whilst complying with its GDPR obligations by anonymising the databases. The project involved the development of a bespoke solution and those involved researched methods online or in technical documentation. The Database Tools project developed the anonymisation process.
29. The project has improved developer capability by providing access to real but anonymised data and has achieved one of its GDPR compliance goals.

Project 5-Realbuzz Registration Software

30. The registrations platform enables clients to set up and manage events, primarily running, walking, cycling and swimming mass participation events. The project sought to enhance various parts of the system to improve the product. Unlike other registration software in the market, the Company's platform is tailored to the needs of the event organisers and charities. The aspects which were the subject of the work were as follows:

(1) Improved event capacities. The software needs to be flexible to accommodate different sizes of events, different groups of entrants and ensuring that fixed capacity of events is not exceeded.

(2) API development. Realbuzz provided the entry system for the London Marathon but another company MIKA handled race day management and timing. This required Realbuzz to integrate its systems with those of MIKA. This required extensive technical liaison between Realbuzz and MIKA.

(3) Improve customer authentication strength of the Company's payment gateways. This involved improving the integration of a large number of payment gateways such as PayPal and Stripe. These well-known payment methods have good quality developer documents and integration guides. Many gateways are poorly documented or not documented at all which made it difficult to obtain technical support.

(4) Fundraising Platform Integration. Users have the opportunity to create a charity fund raising page as part of the registration process for some events. This project focussed on improving the integration of JustGiving and Virgin Money Giving into the registration software.

(5) Worldpay Corporate payment gateway development. This project was aimed at integrating Apple Pay and Google Pay mobile payments into the high-capacity payment gateway previously developed. This required compliance with Apple's and Google's strict requirements on details including button colour, icons, fonts, positioning and text size to provide consistent brand awareness and look-and-feel across all online merchants. Getting sign-off from the companies was an arduous process.

(6) Ballot-improvements to notification emails. This related to sending emails following a ballot for places in an event so that successful entrants were notified in a timely manner, non-delivery was minimised and multiple notifications of rejection were avoided. The challenges arose from the volume of emails required; 450,000 for the London Marathon. The AWS platform was used.

31. In carrying out the project, the developers had to use new (to them) technologies which they had not previously used and had no experience of.
32. This project is part of the Company's aim to be the market leader in registration software.

Project 6-Virtual Events

33. This project was to provide software that enabled users to enter virtual events and facilitate charity fundraising. Initially, this was provided through the existing website.
34. A particular issue was the provision of proof that an entrant had completed the event. They decided to allow all forms of evidence such as GPS applications and photos of smart watches or treadmills. Participants were to receive targeted emails to encourage fundraising and would be provided with automated training support. The project involved the integration of different Company products.
35. Initially, the project enabled the Company to launch an alternative revenue stream during the Covid pandemic. It now proposes to develop the product to support proof of completion, maximise fundraising potential and further engaging participants as part of a community.

Project 7-Realbuzz.com

36. This project involved the extensive updating of the Company's website so that it could run in a newer hosting environment with better integration of the various systems. This increased cost efficiencies and performance and ensured the site remained competitive. Aspects of this were:

(1) Find-my-Nearest: This was the provision of a directory of businesses connected with fitness within a defined area of the user's home. Advertisers were able to buy enhanced listings. This was more focussed than common listings websites such as Google, Bing and Yell but broader than one service niche websites.

(2) Race Finder: The Company wished to develop an event calendar which users could search according to a number of criteria. Challenges included determining how events were to be displayed and whether to use map-based listings and desktop v mobile versions of the site.

(3) Run the World: this aimed to create a global collective of marathon and half marathon events to enable runners to participate in events around the world and to encourage charity fundraising. Challenges included how the events were to be displayed in terms of listings and maps. The solutions included the use of Google Maps.

(4) Now's the Time: This aimed to encourage Realbuzz members to take up a new sport.

(5) Enhanced Community Functionality: The Company sought to develop and improve the social networking aspects of its website. It is likely to do this incorporating third party product solutions. This is a key pillar of the Company's brand strategy.

Project 8-AWS Technology Research

37. To ensure that the technology the Company uses remains robust, secure and up-to-date, the Company spends considerable time researching and understanding the new AWS services and products which are regularly released onto the market. This enables the Company to understand how these new products might benefit the Company and its offering to its members by reading the AWS documentation. If of interest, it would be examined in a test environment for functionality and capabilities. The Company has also researched changing from AWS to an alternative cloud provider.
38. The 2020 Report also contained a qualifying cost analysis, setting out the data showing it qualified as an SME and listing the cost categories for which it was claiming.
39. These were:

(1) Staff costs

(2) Consumable costs

(3) Software costs

(4) Subcontractor costs

(5) Reimbursed travel expenses.

40. The 2020 Report provided totals for the direct and indirect qualifying costs in each category. They were not broken down by project or sub-project. In the case of staff costs, the Report included an Appendix which broke down the employment costs of each employee on a month-by-month basis and estimated for each employee for each month, how much of the total was attributable to R&D time and therefore cost.
41. The 2021 Report followed a similar pattern. There were eight projects of which two were new and four were a continuation from the previous accounting period. The overlaps are shown in the table below.

Project number: 2021 Report	Project number: 2020 Report	Name of Project	Commencement date shown in 2021 Report
1	6	Virtual Events	June 2019
2	2	Product Development and Integration	May 2019
3	-	Master Console	May 2020
4	5	Realbuzz Registrations Software	June 2019
5	-	Runclusive	May 2020
6	7	Realbuzz.com	May 2019

42. No claims were made in APE 2021 in respect of 2020 projects 1 (AWS Security Posture Enhancement), 3 (AWS User Access Key Rotation, 4 (Database Backups and Anonymisation) and 8 (AWS Technology Research).
43. The 2021 Report was submitted on 15 November 2021, after the APE 2021 corporation tax return had been submitted (on 13 July 2021).
44. On 13 April 2022, before the enquiry window for the APE 2020 return expired on 30 April 2022, Officer Patel responded to the 2021 R&D claim. His response on the overlapping projects was as follows:

(1) Project 1-Virtual Events: Not R&D for tax purposes.

(2) Project 2-Product Development and Integration: Not R&D for tax purposes.

(3) Project 2-Series event functionality: Not R&D for tax purposes.

(4) Project 4-Ballot-Improvements to notification emails: Not R&D for tax purposes.

(5) Project 6-Find-my Nearest-Race Finder-Run the World: Not R&D for tax purposes.

45. Officer Patel asked the Company to explain why it thought the rejected projects were within the scope of R&D. He provided a "TU (Technical Uncertainty) questionnaire to provide further analysis of some of the projects in the period. This would be submitted to HMRC's software experts for further analysis. He commented that there might be some elements of R&D in some of the projects and, again, this would be discussed with the software experts.
46. Cowgills responded on 12 July 2022 providing further information.
47. Officer Martin took over the case in September 2022. As Mr Patel was off work at that time she was not able to have any handover discussions with him. She only had the information with was on the system which included correspondence and notes of telephone calls. She responded to Cowgills' letter on 27 September 2022 and concluded that the projects included in the claim for relief in APE 2021 did not meet the definition of R&D for tax purposes. She requested further information and sent a Technological Uncertainty Table Template (TU Table) which set out further questions designed to identify whether certain projects constituted R&D. They were asked to complete the Table in relation to projects 1, 2 and 4 (Virtual Events, Product Development and Integration and Realbuzz Registration Software), all of which were continuations of projects started in APE 2020. Cowgills returned the completed TU Table on 14 October 2022.
48. In a letter dated 18 November 2022 HMRC said they were going to disallow the whole of the APE 2021 R&D claim.
49. On 20 January 2023, HMRC (Ms Martin and software specialists), Realbuzz and Cowgills attended a telephone conference to discuss the APE 2021 claim. The call focussed on Project 2 although some aspects of Project 1 was also discussed.
50. On 8 March 2023 HMRC wrote to the Company disallowing the whole of the APE 2021 claim. The Company did not appeal this decision. In addition, Ms Martin noted that a claim had been made for relief in APE 2020 in relation to the same or similar projects as were included in the R&D claim for 2021. She concluded that the inaccuracies identified in relation to the APE 2021 return had also occurred in the earlier period and she would be considering raising a discovery assessment in relation to the 2020 return.
51. The Discovery Assessment was issued on 1 June 2023 disallowing the whole of the claim for R&D relief so that the company was required to repay £335,452.57 plus interest.

The law

52. The provisions concerning discovery assessments in relation to corporation tax returns are contained in schedule 18 to the Finance Act 1998. Paragraph 41, so far as material, provides:

"41 (1) If [an officer of Revenue and Customs][discovers] as regards an accounting period of a company that—

(a) an amount which ought to have been assessed to tax has not been assessed, or

(b) an assessment to tax is or has become insufficient, or

(c) relief has been given which is or has become excessive,

[he] may make an assessment (a "discovery assessment") in the amount or further amount which ought in [his] opinion to be charged in order to make good to the Crown the loss of tax."

53. Paragraph 42 imposes restrictions on the circumstances in which a discovery assessment may be raised. The relevant restrictions are contained in paragraph 44, which provides:

"44 (1) A discovery assessment for an accounting period for which the company has delivered a company tax return, or a discovery determination, may be made if at the time when [an officer of Revenue and Customs]—

(a) ceased to be entitled to give a notice of enquiry into the return, …

[he] could not have been reasonably expected, on the basis of the information made available to [him] before that time, to be aware of the situation mentioned in paragraph 41(1) or (2).

(2) For this purpose information is regarded as made available to [an officer of Revenue and Customs] if—

(a) it is contained in a relevant return by the company or in documents accompanying any such return, or

(b) it is contained in a relevant claim made by the company or in any accounts, statements or documents accompanying any such claim, or

(c) it is contained in any documents, accounts or information produced or provided by the company to [an officer of Revenue and Customs] for the purposes of an enquiry into any such return or claim, or

(d) it is information the existence of which, and the relevance of which as regards the situation mentioned in paragraph 41(1) or (2)—

(i) could reasonably be expected to be inferred by [an officer of Revenue and Customs] from information falling within paragraphs (a) to (c) above, or

(ii) are notified in writing to [an officer of Revenue and Customs] by the company or a person acting on its behalf.

(3) In sub-paragraph (2)— "relevant return" means the company's company tax return for the period in question or either of the two immediately preceding accounting periods, and "relevant claim" means a claim made by or on behalf of the company as regards the period in question [or an application under section 751A of the Taxes Act 1988 made by or on behalf of the company which affects the company's tax return for the period in question]."

54. The Appellant's case is that the officer could "have been reasonably expected, on the basis of the information made available to [him] before [the LDE], to be aware of the situation mentioned in paragraph 41(1) or (2)." Accordingly, HMRC were not permitted to raise the discovery assessment.

Discussion

55. It is common ground that no R&D relief was due for APE 2020 and that additional tax should be due. It is accepted that HMRC made a "discovery", and that the assessment was properly notified to the Appellant. It is agreed that paragraph 41(1) is satisfied.
56. HMRC accept that the Appellant's corporation tax return was not completed carelessly, so that a discovery assessment cannot be raised by virtue of paragraph 43.
57. The main issue between the parties is whether paragraph 44(1) is satisfied. That is to say, whether an officer of HMRC, at the time when the enquiry window closed, could not have been reasonably expected, on the basis of the information made available to him before that time, to be aware of the situation mentioned in paragraph 41(1). If the section 44 test is satisfied, that is, if the officer would not have been aware of the insufficiency, HMRC may raise the discovery assessment. If the test is not satisfied, that is if the officer should have been aware, HMRC is barred from making the discovery assessment.
58. It is common ground that the 2020 corporation tax return and the 2020 Report was "information made available". The second issue is whether the 2021 Report was also information made available in relation to APE 2020.
59. Mr Hellier submits that it should have been obvious, from the 2020 Report alone that no R&D relief was due and that HMRC are therefore barred from making the discovery assessment.. HMRC accept that it was apparent that some of the projects did not qualify but that other projects were not obviously non-qualifying and Mr Marks submits that that enables HMRC to assess the whole of the underpaid tax.
60. There are two main points of disagreement between the parties.
61. Mr Marks submits that even if it is clear that insufficient tax has been charged, the officer is not barred from raising the assessment if the amount of the assessment cannot be quantified. Mr Hellier argues that there is no requirement for quantification.
62. The parties also disagree on the nature of the officer's awareness and whether the officer needs to be aware that none of the projects qualified or if it is sufficient that he is aware that some projects do not qualify but others might.
63. The burden of proof is on HMRC to prove, on the balance of probabilities that the discovery assessment is valid.
64. There is much case law analysing the principles to be applied in determining the necessary level of the officer's awareness. Most of the cases relate to income tax and capital gains tax discovery assessments under section 29A of the Taxes Management Act 1970 (TMA), but it is agreed that they apply equally to the almost identically worded discovery provisions for corporation tax purposes.
65. The principles are summarised in the Court of Appeal case of HMRC v Fisher and Others [2021] EWCA Civ 1438 (Fisher) which draws on the earlier cases of Langham (Inspector of Taxes) v Veltema [2004] ECWA Civ 193 (Veltema), HMRC v Lansdowne Partners LLP [2011] EWCA Civ 1578 (Lansdowne), Sanderson v HMRC [2016] EWCA Civ 19 (Sanderson) and Clive Beagles v HMRC [2018] UKUT 380 (TCC) (Beagles). We have also considered the other authorities cited by the parties.
66. In Fisher at [126] Newey LJ set out the principles as enumerated in Sanderson as follows:

"126. Patten LJ, with whom Briggs and Simon LJJ agreed, summarised principles relating to section 29 of the TMA in these terms in Sanderson v Revenue and Customs Commissioners [2016] EWCA Civ 19, [2016] 4 WLR 67, at paragraph 17:

"The power of HMRC to make an assessment under section 29(1) following the discovery of what, for convenience, I shall refer to as an insufficiency in the self-assessment depends upon whether an officer 'could not have been reasonably expected, on the basis of the information made available to him before that time, to be aware of the insufficiency'. It is clear as a matter of authority:

(1) that the officer is not the actual officer who made the assessment … but a hypothetical officer;

(2) that the officer has the characteristics of an officer of general competence, knowledge or skill which a reasonable knowledge and understanding of the law: see HMRC v Lansdowne Partners LLP [2012] STC 544;

(3) that where the law is complex even adequate disclosure by the taxpayer may not make it reasonable for the officer to have discovered the insufficiency on the basis of the information disclosed at the time: see Lansdowne at [69];

(4) that what the hypothetical officer must have been reasonably expected to be aware of is an actual insufficiency: see Langham v Veltema [2004] STC 544 per Auld LJ at [33]–[34]:

'33. More particularly, it is plain from the wording of the statutory test in section 29(5) that it is concerned, not with what an Inspector could reasonably have been expected to do, but with what he could have been reasonably expected to be aware of. It speaks of an Inspector's objective awareness, from the information made available to him by the taxpayer, of 'the situation' mentioned in section 29(1), namely an actual insufficiency in the assessment, not an objective awareness that he should do something to check whether there is such an insufficiency …;

(5) that the assessment of whether the officer could reasonably have been expected to be aware of the insufficiency falls to be determined on the basis of the types of available information specified in section 29(6). These are the only sources of information to be taken into account for that purpose: see Langham v Veltema, at [36]:

'The answer to the second issue - as to the source of the information for the purpose of section 29(5) though distinct from, may throw some light on, the answer to the first issue. It seems to me that the key to the scheme is that the Inspector is to be shut out from making a discovery assessment under the section only when the taxpayer or his representatives, in making an honest and accurate return or in responding to a section 9A enquiry, have clearly alerted him to the insufficiency of the assessment, not where the Inspector may have some other information, not normally part of his checks, that may put the sufficiency of the assessment in question. If that other information when seen by the Inspector does cause him to question the assessment, he has the option of making a section 9A enquiry before the discovery provisions of section 29(5) come into play.'"

67. From these passages, it is clear that:

(1) We must consider the awareness, not of the actual officers involved, but of a hypothetical officer. The hypothetical officer has the characteristics of an officer of general competence, knowledge (in this case of the R&D legislation and guidelines) or skill, which includes a reasonable knowledge and understanding of the law.

(2) The hypothetical officer must be clearly alerted to an insufficiency of tax. While the information provided need not be sufficient to enable HMRC to prove its case it must be more than would prompt the hypothetical officer to open an enquiry i.e. there must be more than a suspicion of insufficiency (Beagles).

(3) Where the law is complex the hypothetical officer will not be expected to resolve all legal issues.

(4) The hypothetical officer is required only to take account of the information provided as defined in section 44(2) and is not required to take any additional steps to verify the information or look at other information which that provided indicates may be relevant.

68. The points set out at [67] above are not contested.

The quantification issue

69. The first point which is disputed is whether the hypothetical officer's level of awareness must enable him to quantify the insufficiency of tax based on the information provided.
70. Mr Hellier submits that no quantification is required and referred to an extract from Simon's Taxes at division A6.710 which states

"…the first point to consider is what is meant by the officer being reasonable expected to be aware of the situation.

This is an objective test of awareness of a situation; it does not require the hypothetical officer to be able to quantify the under assessment. The quantification is done by the actual officer who raises the assessment…"

71. Mr Marks observes that this is simply commentary, and no cases are cited in support. Indeed, the cases are silent on the issue. Sanderson at [22] and Veltema at [36] require only that the hypothetical officer to be made aware of an actual insufficiency in the assessment.
72. The question remains whether HMRC is to be "shut out" from making a discovery assessment where he is aware of an actual insufficiency of tax but is unable to quantify it. Mr Marks submits that he is not. As there is no case law on this point, he starts from the legislation.
73. Paragraph 44 (1) provides that a discovery assessment may be made if, when the enquiry window closes, the hypothetical officer could not have been reasonably expected to be aware of "the situation" mentioned in paragraph 41(1). Paragraph 41 (1) provides that:

"41 (1) If [an officer of Revenue and Customs][discovers] as regards an accounting period of a company that—

(a) an amount which ought to have been assessed to tax has not been assessed, or

(b) an assessment to tax is or has become insufficient, or

(c) relief has been given which is or has become excessive,

[he] may make an assessment (a "discovery assessment") in the amount or further amount which ought in [his] opinion to be charged in order to make good to the Crown the loss of tax."

74. Mr Marks submits that "the situation" referred to in paragraph 44(1) is the whole of paragraph 41(1) and includes the provision that the officer may make an assessment in the amount which in his opinion ought to be charged in order to make good the loss of tax. The claim is for a specific amount of relief. Where it is clear that some of the claim is wrong, but possibly not all of it, the hypothetical officer cannot be in a position to issue an assessment because he does not know how much he needs to assess. He has not therefore been made aware of the situation.
75. He contends that we must look at the assessment actually made (which determined that none of the claim qualified for relief) and consider whether that conclusion could have been reached by the hypothetical officer at the LDE. Mr Marks took us to Sanderson at [25] where the Court of Appeal stated that sub-sections (1) and (5) of section 29 TMA (the equivalent of paragraphs 41(1) and 44(1)) set out different tests.

"[25] I do not accept that sub-ss 29(1) and (5) import the same test and that the Revenue's power to raise an assessment is therefore directly dependent on the level of awareness which the notional officer would have based on the s 29(6) information. The exercise of the s 29(1) power is made by a real officer whois required to come to a conclusion about a possible insufficiency based on all the available information at the time when the discovery assessment is made. Section 29(5) operates to place a restriction on the exercise of that power by reference to a hypothetical officer who is required to carry out an evaluation of the adequacy of the return at a fixed and different point in time on the basis of a fixed and limited class of information. The purpose of the condition is to test the adequacy of the taxpayer's disclosure, not to prescribe the circumstances which would justify the real officer in exercising the s 29(1) power."

76. Mr Marks submits that the reference to "adequacy" must involve quantum as the amount of the assessment is part of "the situation" referred to in paragraph 44(1). He contends that the test in this case is whether the hypothetical officer would have been aware, at the LDE, that no relief at all was due.
77. He further submits that the hypothetical officer could not have been aware of the amount of the overclaim at that point.
78. The 2020 Report analysed the R&D costs claimed in terms of the nature of the costs; staff costs, consumables, software, subcontractor costs and reimbursed travel costs. It also provided a detailed breakdown of staff costs and the percentage of each person's time spent on R&D activities in each month. Mr Marks submits that that is not sufficient. There was no attribution of costs to the different projects and no indication of what items made up the claim. It was not therefore possible to quantify the amount of the overclaim as some items might have qualified, but others did not and the hypothetical officer could not determine, from the Report which was which.
79. He also suggested that it was unclear whether items which were not thought to be R&D had been included in the claim. We do not agree. Section 8 of the 2020 Report states that the company's directors and Cowgill "have identified 10 projects with are considered to qualify for R&D tax relief in consultation with HMRC's BIS guidelines and relevant legislation…A sample of projects considered to qualify for R&D tax relief has been summarised below". We infer from that statement that the Appellant considered that all the costs set out in the Report related to R&D and were included in the claim.
80. Mr Hellier submits that it is not necessary for the hypothetical officer to be able to quantify the amount of the assessment. As noted, he referred to extracts from Simon's Taxes on this point. He also took us to statements in the authorities which he submits support this contention although they do not in terms refer to quantification.
81. In Beagles, at [100(6)] the Court of Appeal said "the information made available must "justify" raising the additional assessment or be sufficient to enable HMRC to make a decision whether to raise an additional assessment."
82. In Veltema at [33] the Court referred to section 29(5) TMA (the equivalent of paragraph 44(1)) as speaking of "an Inspector's objective awareness …of "the situation" mentioned in section 29(1) (the equivalent of paragraph 41(1)), namely an actual insufficiency in the assessment"
83. Mr Hellier also took us to paragraph 25 in Sanderson, set out at [75] above but drew a different conclusion from that of Mr Marks. He submitted that the extract drew a distinction between the section 29(1)/paragraph 41(1) test, which requires the actual officer to quantify the insufficiency based on all the information available at the time the discovery assessment is raised, and the section 29(5)/paragraph 44(1) test which requires the hypothetical officer to evaluate the adequacy of the return at a different point in time on the basis of specified and limited information. As a result, of the evaluation the hypothetical officer must reach the conclusion that not enough tax has been paid; that there is an actual insufficiency of tax but does not require him to decide the amount of the insufficiency.
84. We prefer Mr Hellier's arguments on this point. In our view, the natural meaning of "the situation" in paragraph 41(1) is the officer's discovery (in this case) that an assessment to tax is or has become insufficient or that relief has been given which is or has become excessive. The final part of paragraph 44(1) ("the officer may make an assessment" etc) sets out the consequence of "the situation" and is not itself part of the situation. This is the interpretation in Veltema. "The situation" means that the Inspector must be aware of an actual insufficiency". In Beagles, the requirement is stated to be that the information available must enable HMRC to make a decision whether to raise an additional assessment or "justify" raising an assessment. It does not say that the hypothetical officer must be able at that point to raise an assessment in a specific amount; that is the job of the real officer who has all the relevant information available, as confirmed in Sanderson.
85. There is further support for this view in the wording of paragraph 44(1): "the officer may make… a "discovery" assessment in the amount or further amount which ought in their opinion to be charged in order to make good to the Crown the loss of tax". [emphasis added] Although the real officer must raise the assessment for a specific amount of money, they may not know what the actual amount is and can effectively make a "best judgment" assessment based on their knowledge at the time. The burden is then on the taxpayer to show that they have been overcharged by the assessment.
86. We accordingly conclude that the hypothetical officer does not have to quantify the amount of the tax underpaid/relief overclaimed in order to be "aware of the actual insufficiency" in the assessment. He has only to be aware that there is an actual insufficiency.

The complexity issue

87. The case law indicates that the complexity of the relevant law is a factor in determining what the hypothetical officer is aware of. See for example point (3) in the passage from Fisher at [66] above. Similarly, Beagles at [100(4)] states "In some cases it may be that the law is so complex that the inspector could not reasonably be expected to be aware of the insufficiency".
88. Mr Marks submits that the complexity in this case is such that the hypothetical officer would not be aware of the insufficiency based on the 2020 Report. He submits that the test is not limited to complex law but relates to all the circumstances of the claim. Although the hypothetical officer will be familiar with and understand the R&D legislation, Guidelines and case law, he will not necessarily be an expert on all types of business or scientific endeavour. Software is a complex field and the hypothetical officer is not a software specialist. If the officer would need to ask questions of a specialist colleague or research terms used in the Report to understand it and determine whether there is an insufficiency, the paragraph 44(1) test is not satisfied. He is only permitted to use the information available as specified in that paragraph and cannot take any further action even if it would be a small step to do so (see Freeman v HMRC [2013] UKFTT 496 (TC) at [73] (Freeman)).
89. Mr Marks contends that although some of the projects or sub-projects in the 2020 Report were clearly non-qualifying, other projects were not obviously non-qualifying but would need further enquiries to be made to determine whether they did or did not qualify.
90. Officer Martin's actual conclusion as set out in her letter on the APE 2021 return, was that none of the projects (including those which had begun in APE 2020) qualified for R&D relief as the Company had not sought to achieve any overall advance in science or technology; they advanced the Company's knowledge but not the field of technology.
91. Officer Patel, in his letter of 13 April 2022 reached the conclusion (on the 2021 Report) that most of the projects (including those started in 2020) were non-qualifying. He also said there were some aspects which might qualify and he required further information which he would refer it to his specialist colleagues.
92. Mr Marks contends that the actions and conclusions of the real officers are irrelevant to the test. Irrespective of their conclusions and decisions, he submits that the hypothetical officer could not have reached those conclusions at the LDE just by reading the return and the 2020 Report. As software is such a complicated field, the hypothetical officer could not be expected to be aware of whether something was an innovation or not and would need further information to understand what was claimed and whether the items which did not immediately appear to be non-qualifying did or did not qualify.
93. He further submits that the hypothetical officer could not, from the 2020 Report alone, be aware whether the claim was not allowable in its entirety or only in part because of the lack of detailed costings.
94. Mr Hellier submits that the cases refer to "complexity" in the context of complex law rather than facts. In Lansdowne at [69] the Court of Appeal said "the legal points were not complex or difficult…. there may be circumstances in which an officer could not reasonably be expected to be aware of an insufficiency by reason of the complexity of the relevant law". And at [56] the Court stated that the hypothetical inspector does not need to resolve points of law. Any disputes of fact or law can be resolved by the normal appeal processes. This comment was endorsed in Sanderson at [23].
95. Mr Hellier contends that the law in this case is straightforward.
96. We agree that the law relating to R&D applicable in this case is relatively straightforward. The principles as set out in the Guidelines are clear in concept. To qualify for R&D relief, the research and development must advance knowledge in a scientific or technical field generally or significantly improve an existing process.
97. The application of those principles to the facts of a case may be more difficult.
98. Whether it is the law or the facts of a case which are complex, paragraph 44(1) focusses primarily on the adequacy of the disclosure by the taxpayer, as explained in the Upper Tribunal case of HMRC v John Hicks [2020] UKUT 0012 (TCC) (Hicks), in relation to section 29(5) TMA at [196]. The Tribunal went on to say at [199]-[200]:

"199. Plainly, the greater the level of disclosure, the greater the officer's awareness can reasonably be expected to be. If a disclosure on a tax return includes all material facts and, in complex cases, an adequate explanation of the technical issues raised by those facts and the position taken in relation to those issues, it would be reasonable to expect an officer to be aware of an insufficiency. What constitutes reasonable awareness is linked to the fullness and adequacy of the disclosure – the expertise of the hypothetical officer remains that of general competence, knowledge or skill which includes a reasonable knowledge and understanding of the law.

200. In argument before us Mr Nawbatt came close to suggesting, as we understood it, that a hypothetical officer could not be expected to understand complex or specialist areas of tax law. We disagree. If the disclosure (factual and technical) is adequate in the circumstances of the case, a hypothetical officer can reasonably be expected to be aware of an insufficiency even in a complex case or one involving specialist technical knowledge. …"

99. We conclude that the hypothetical officer would have had an adequate understanding of the tax law relating to R&D. He is not, however, assumed to be an expert in software technology, and it is here that the adequacy of disclosure is important.

What should the hypothetical officer have been reasonably aware of?

100. Although we must address the awareness of the hypothetical officer, it will be convenient to start by considering the awareness of the actual officers involved in the case.
101. In his letter of 13 April 2022, Mr Patel considered the projects and sub-projects which were subject to the R&D claim in APE 2021. Four of those projects were continuations of projects which began in APE 2020.
102. Mr Patel conclusions were as follows, in relation to the continued projects:

(1) Project 1: Virtual Events: Not R&D as it was an advancement of the Company's commercial aims not an advance in science or technology.

(2) Project 2: Product Development and Integration: The Consoles sub-project had been put on hold. The Automated Testing System and Series Event Functionality sub-projects did not qualify for relief as they were advancing the commercial aims of the business and adapting an existing process to make the functionality faster and more efficient respectively.

(3) Project 4: Realbuzz Registrations Software: overall not R&D as no overall advance. The software used in creating the Company's platform would be readily deducible. Several of the 2021 sub-projects were not included in the 2020 Report. The Ballot sub-project was a commercial advance but not an advance in overall knowledge.

(4) Project 6: Realbuzz.com: Not eligible. The Company enhanced its website. While it extended its own knowledge, Mr Patel could not identify any advance in overall knowledge or capability. In relation to the sub-projects:

(a) Find-my-nearest: not R&D. It was the creation of a database.

(b) Race Finder: not R&D.

(c) Run the World: Not R&D but the adaptation of an existing product.

(d) Personalised Training Support: Not R&D. Machine learning tools are readily available, and a customisation of individual needs is not R&D for tax purposes.

103. Under the heading "Next Steps", Mr Patel stated:

"Therefore I have attached/provided a questionnaire to be filled, to provide further analysis/evidence of the projects which have taken place during the APE in question. I will then submit this information to our specialists in software who will further review the information you provide on the questionnaire as well as the project document you have already provided.

Within this letter I have regularly informed you of why certain projects/mini projects and enhancements do not fall into the scope for tax purposes. For sections of the letter in which I have mentioned this please go onto explain using HMRC guidance and guidelines why you think these projects would be within the scope of R&D for tax purposes.

There are elements within this letter which I have not questioned this is because I am pleased that there may in fact be elements of R&D taking place within your projects however as mentioned specialists within HMRC who look at software projects will review the answers on the questionnaire and the R&D project document who may well in fact differ in my opinion and therefore more questions/decisions will be made once the questionnaire and your response has been received." [our emphasis]

104. We have two comments on this. First, Mr Patel did not mention all of the projects/sub-projects because he considered it possible that there may have been some qualifying R&D in the projects he did not mention. The following sub-projects which were in both the 2020 and the 2021 Reports were not mentioned:

(1) Project 2: Database Tools and AWS Costing Tool.

(2) Project 4: Improved Event Capacities, API development, Improve customer authentication of payment gateways, Worldpay Corporate payment gateway development and Fundraising platform integration.

(3) Project 6: Now's the Time, Enhanced Community Functionality.

105. The implication is that Mr Patel considered that these sub-projects or aspects of them might qualify as R&D.
106. Secondly, Mr Patel proposed to submit the additional information to be provided in the questionnaire together with the Report to HMRC's software specialists for their views which, he anticipated, might differ from his view.
107. In summary, Mr Patel was confident that many of the projects did not qualify for relief, but he considered there were some projects which might possibly qualify and in any event, he would require the software specialists sign off before making a final decision.
108. Miss Martin then took over the 2021 enquiry.
109. The TU questionnaire related only to Projects 1, 2 and 4, all of which were projects begun in APE 2020. There was a conference call between Ms Martin, two software specialists, Realbuzz and Cowgills which mainly discussed Project 2 and some aspects of Project 1. Following the telephone call further information was provided on Projects 1, 2 and 4. No further information was provided about any other project in APE 2021nor was any information asked for or received in relation to the projects from APE 2020.
110. Ms Martin confirmed that no further information emerged, in the course of the 2021 enquiry, which changed HMRC's view. HMRC's view all along was that the claim was excessive. The additional enquiries were to ensure HMRC was comfortable with their initial impression. Ms Martin explained that the field of technology was notoriously complex and that ordinary officers were not experts in any particular field of technology. Software is a particularly difficult area, and it was felt to be important to understand the Company's approach to R&D, the methodology it used and the process it adopted in determining whether work was R&D. Understanding the methodology assisted in judging the quality of the company's R&D decisions and whether its approach aligned with the legislation. There was, however, nothing specific in the information or process which changed or confirmed HMRC's view.
111. On 8 March 2023, Ms Martin wrote to Realbuzz to confirm that the whole of the APE 2021 claim would be disallowed. The same letter she stated that she was considering raising a discovery assessment for APE 2020.
112. Ms Martin issued the discovery assessment on 1 June 2023. The basis for the assessment was that "the same projects are being claimed for in the accounting period ending 31 April 2020, as in the accounting period ending 31 April 2021, and that these projects have been found to not qualify for Research and Development tax relief." She disallowed the entire amount of the 2020 claim, even though additional information had been provided in relation to only three of the projects which had continued into 2021 and no information beyond the 2020 Report had been provided in relation to Projects 1, 3, 4 and 8 in APE 2020.
113. In summary, the actual officers must have made their decision on the 2020 claim solely based on the 2020 Report. Although further information was provided and discussed in relation to three of the continuing Projects in relation to APE 2021, this did not produce anything new which changed or confirmed HMRC's view.
114. We now turn to the hypothetical officer.
115. Mr Marks sought to show that parts of the 2020 Report indicated activity which might constitute qualifying R&D. He highlighted some examples which could potentially fall within the Guidelines and also highlighted the technical complexity and jargon which would make it difficult for a non-specialist to reach a definite conclusion. He submitted that this indicated that there were projects/sub-projects which were not obviously non-qualifying. For example, in Project 1:

"Solutions were not readily available by default due to the nature of SaaS and PaaS cloud services. As stated earlier in this report, the company identified 4 core areas, or technical uncertainties they needed to overcome. The company could not find a readily available single solution 'off-the-shelf' which would enable them to address all 4 of these core areas."

116. There was much technical jargon in the Report and this extract indicates that there was no readily available solution so the Company had to fill the gap-which could be qualifying R&D. Similarly in relation to Project 2 under Database Tools:

"The company looked at great length for off-the-shelf tools that would do this level of database anonymisation, despite GDPR, no solution was readily available meaning the company's solution was entirely bespoke.

The company spent a lot of time looking for off-the-shelf products to do the trimming and anonymisation that they required, there was nothing that came close to what the company needed. A fully bespoke system was created. Within the implementation, the company had not seen the following elsewhere:

➢ An abstraction layer over different database backends to allow the same functionality to work over mySQL and postgre

➢ The ability to run commands at any level in the hierarchy of database objects, host, database, table, and column."

117. Further:

"Having the anonymiser consistently anonymise the same piece of information to the same randomised replacement, via backing onto a redis persistence layer, is also something the company had not seen in other anonymisation systems. This was important to the company as some projects are multi database and require PII like email addresses to be consistent across multiple databases."

118. In relation to Series Event functionality:

"The company wished to analyse and improve their import process to efficiently match new results to the database of users via per event configurable matching criteria. Where third party software was being used, the company looked at how it worked to see if they could produce something better or whether they could implement functionality where they had researched alternative methods to do the same thing, but faster and more efficiently."

119. Under AWS Costing Tool:

"At the time of writing, the company does not believe there was anything on the market that would allow visualisation and fine grain introspection of billing for AWS services that was provided by the company's offering. The software created fulfilled the need of the company, the company implemented this as it allowed the company to understand and reduce their bill substantially, in addition to improving client billing."

120. Mr Patel did not immediately reject the Database Tools or the AWS Costing Tool sub-projects meaning he considered that these items could possibly contain elements of qualifying R&D. He did reject Series Event Functionality having interpreted it as the "routine analysis, copying or adaptation of an existing process…" within Guideline 12.
121. However, a more efficient product or cost improvements can be R&D within paragraph 7 of the Guidelines and a project which seeks to make an "appreciable improvement" to an existing process, product or service through technological change can be R&D within Guideline 9. Making something faster and more efficient can be an appreciable improvement.
122. It is not necessary to consider all the projects, but we note that there were other passages which referred to bespoke solutions or improvements to what was on the market, and many passages incorporating highly technical language.
123. Mr Hellier's approach was to consider what the projects were aiming to do at a high level. He also reminded us that the burden was on HMRC to show that the disclosure was insufficient as a whole.
124. The projects are summarised from [22] above and again, we consider a selection of them.

Project 1

125. This involved researching software products already on the market to improve the security of the Company's online services. The extract Mr Marks highlighted (see [115]) might suggest R&D but when the aim and outcome of the project is considered, it is clearly not R&D for tax purposes.

Project 2

126. The aim of the project was to integrate the Company's core products. It allowed the Company to expand its commercial advertising inventory and laid the foundation for a subscription-based training model which were important commercially. The project was not aiming for an overall technical advance, but improving the Company's knowledge and commercial outcomes, which is not R&D. However, as we have noted, there were some elements of the work carried out to achieve these aims which might have qualified.

Project 3

127. The aim was to improve security in relation to access to the Company's services by periodically changing the access keys to its AWS platform. The Company wanted to implement best practice and had to research what that was. The Company's research involved reviewing guides and documentation produced by AWS and they ultimately decided to use a new AWS service to achieve their aims. While this advanced the Company's knowledge, it simply used existing products to improve its security posture. This is clearly not R&D.

Project 8

128. This simply involved keeping up to date with the products issued by its existing platform provider, AWS. This is clearly not R&D.
129. In summary, there were some projects, notably projects 1, 3 and 8 which obviously did not qualify for R&D and no special technical knowledge was required to come to that conclusion. The other projects included long, highly technical, explanations of activities undertaken, some of which were aimed at developing new technical solutions not available in the market and others looking to increase speed, efficiency or cost effectiveness which are capable of constituting R&D.
130. Returning to the actual officers, Mr Patel's initial conclusion was that most projects did not qualify, but some might do so, and he would want to consult his software specialist colleagues before making a final decision. It is notable that the TU questionnaire related only to three projects (out of six in 2021) and HMRC's software specialists only considered one project in depth and asked some questions about a second in the telephone meeting. The whole of the claims for both 2021 and 2020 were thereafter rejected. Ms Martin indicated that the purpose of the meeting was to probe the Company's approach to identifying R&D, it's methodology and process. Based on those discussions, HMRC did not feel it needed any further information on any of the other projects. We infer that the Company's approach and methodology in general indicated that it was treating work as R&D when it did not in fact qualify. This confirmed HMRC's view that the various projects advanced the Company's knowledge and commercial aims but did not advance science or technology overall. This meeting was therefore an important part of the decision-making process by which the claims were rejected.
131. We also note that the Reports did not cover all of the projects which were the subject of the claims. Only eight out of ten were included in the 2020 report and six out of eight in the 2021 Report.
132. The hypothetical officer is an "officer of general competence, knowledge or skill with a reasonable knowledge and understanding of the law". He is not a software specialist. We consider that, based on the 2020 Report, it would be reasonable to expect the hypothetical officer to be aware that several of the projects could not qualify for R&D relief. It would not be reasonable to expect him to be aware that none of the projects qualified for the relief. He had no information at all about the two projects which were not contained in the Report and so could not have formed any view about those projects. As noted, there were elements of other projects which, on the face of it, might have qualified. The hypothetical officer would have needed more information and/or to collaborate with specialist colleagues, to reach a definite conclusion on those aspects of the claim.
133. The hypothetical officer would not have been expected at the LDE to be aware that the entire claim was non-qualifying, but he would have been expected to be aware that some of the projects and sub-projects did not qualify for relief. It follows that he would be expected to be aware that the claim for R&D relief was excessive. Accordingly, we conclude that, at the LDE, the hypothetical officer could, on the information then available to him, have been reasonably expected to be aware of the situation in paragraph 41(1), namely that the R&D relief given in APE 2020 was excessive so that an assessment to tax had become insufficient.

Must the hypothetical officer have been aware that none of the projects qualified for relief?

134. Mr Marks acknowledged that some of the projects were clearly non-qualifying. In particular, he singled out Project 8 which consisted of reviewing new products issued by AWS to see whether they might benefit the business. Quite clearly this did not involve any general advance in science or technology.
135. Mr Marks submitted that if it was clear that some of the claim was wrong, but not clear that all of it could not qualify, the hypothetical officer would not have been made aware of "the situation" as he would not know the amount to assess and would not be in a position to issue an assessment. Mr Marks went on to say that the hypothetical officer must not only be qualitatively aware of an insufficiency but must be able to quantify the insufficiency in order that he can assess only the amount which is needed to make good to the crown the loss of tax. We have dealt with the quantification issue above and concluded that it was not necessary that the hypothetical officer be able to quantify the amount of the insufficiency at the LDE.
136. Mr Marks then submitted that one must look at the assessment actually made and ask whether those conclusions could be reached at the LDE. The actual conclusion was that none of the R&D claim was good, and he argues that HMRC are only precluded from raising the discovery assessment if, at the LDE the hypothetical officer would be aware that there could be no claim for relief at all.
137. Mr Hellier argued that Mr Marks was confusing the actual assessment with the test for adequate disclosure.
138. The Court of Appeal set out the distinction between the paragraph 41(1) test and the paragraph 44(1) in Sanderson at [25], set out at [75] above. Paragraph 41(1) looks at the assessment which the actual officer makes, based on all the information he has gathered during the enquiry. Paragraph 44(1) is a limitation on the actual officer's ability to raise an assessment based on what the hypothetical officer should have been aware of in the light of specified information at a different time.
139. We agree with Mr Hellier on this point. That is, we must consider what the hypothetical officer was aware of at the LDE and whether he should have been aware of an insufficiency of tax. It is not necessary that he was aware, at that point, that none of the claim qualified for relief.
140. HMRC's primary case is that paragraph 44(1) only prevents them from raising the discovery assessment if, at the LDE, the hypothetical officer should reasonably have been aware that no R&D relief was available. Mr Marks submits that if it was clear that some parts of the claim were non-qualifying but other parts might have qualified, HMRC remains entitled to assess the whole of the insufficiency, including those parts where the disclosure was adequate such that the hypothetical officer was aware of the insufficiency.
141. If we are not with HMRC on this point, Mr Marks submits that paragraph 44(1) only protects that part of the claim which was obviously non-qualifying and HMRC can assess the remainder of the insufficiency arising from the doubtful parts of the claim. As authority for this, he took us to the case of Brian Lynch v HMRC [2025] UKFTT 300 (TC) at [79] which considered how section 29 TMA operates where there are "distinct parts of an insufficiency of tax". Mr Lynch had participated in a tax avoidance scheme which produced two separate "insufficiencies"; a "dry" tax charge to income tax which, it was claimed, was not taxable and a claim to interest relief. The FTT first referred to the case of Hargreaves v HMRC [2022] UKUT 34 (TCC) in which Mr Hargreaves had claimed to be non-resident in the UK and had completed his tax return accordingly. He was subsequently found to be resident and as a result he had underdeclared his income. He had not declared his capital gains at all. Mr Hargreaves argued that HMRC should have been aware of an insufficiency in income tax and therefore could not assess either income tax or capital gains tax (CGT) even though HMRC could not have been aware of an insufficiency in the latter. HMRC argued that, as section 29(1) TMA referred to "any" income tax or CGT not being assessed, then they could make a discovery assessment to both income tax and CGT in this situation. Although the Upper Tribunal did not regard this point as straightforward and did not need to decide the point to reach its conclusion it commented that it considered the better view to be that HMRC could make assessments in relation to both income tax and CGT. The Tribunal in Lynch arrived at that conclusion without considering whether the Upper Tier's reasoning was obiter. The Tribunal also decided that the same principle applied where there were two different insufficiencies relating to income tax.
142. Mr Hellier submits that if Mr Marks is right in his contention that adequate disclosure in relation to some parts of the claim can be overridden by inadequate disclosure of other parts (i.e. so it is not clear whether or not they qualify for relief) and HMRC can then assess the whole of the claim, it would denude the legislation of the protection intended to be provided.
143. Mr Hellier's primary submission is that if the hypothetical officer should have been aware of an excess in the relief, HMRC is precluded from raising the discovery assessment. If the disclosure is adequate to show that the claim was excessive, the whole of the claim is protected.
144. His secondary contention is that, if the whole of the claim is not protected, those parts of the claim where adequate disclosure was made are protected.
145. Mr Marks' submissions on Lynch and Hargreaves were made in reply and Mr Hellier did not have an opportunity of commenting.
146. Leaving aside the issue as to whether the Upper Tribunal's comments in Hargreaves were obiter, both Lynch and Hargreaves can be distinguished from the present case. They dealt with a situation where there were two separate and distinct tax charges or insufficiencies relating to difference inaccuracies, whether they were in relation to different taxes (Hargreaves) or the same tax (Lynch).
147. In Realbuzz's case there is only one inaccuracy and one potential tax charge. There is a single claim to R&D relief for a specified amount. If some projects qualify for relief and some do not, that means that "relief has been given which is or has become excessive" and it follows that "an assessment to tax is or has become insufficient" as described in paragraph 41 (1) (b) and (c). Where there is a single claim to relief which is excessive because some parts of it qualify and other parts do not, and at the LDE, the hypothetical officer should reasonably have been aware that the claim for relief was excessive, we consider that paragraph 44(1) prevents HMRC from raising a discovery assessment in relation to that claim.
148. That, in our view, is in accordance with the purpose of paragraph 41 which is to strike a balance between HMRC's right to collect the right amount of tax and the taxpayer's right to finality and certainty.

Overall conclusion

149. We have concluded that the hypothetical officer would have been aware that there was an insufficiency of tax because the claim for R&D relief was excessive. It would have been obvious that some projects/sub-projects did not qualify even if others might have. As there was a single inaccuracy being a single excessive claim for relief, the hypothetical officer only had to conclude that the claim was excessive, he did not have to conclude that the entire claim was non-qualifying. Nor did he have to quantify the amount of the insufficiency of tax. That was for the actual officers to decide based on all the information available to them.
150. Accordingly, the protection provided by paragraph 44(1) is available to the Appellant and HMRC are not able to raise a discovery assessment by reference to the APE 2020 claim for R&D relief.
151. That disposes of the appeal. However, the parties made submissions as to whether the 2021 Report was "information available" in relation to the 2020 claim and we briefly consider this below.

Was the 2021 Report "information made available?

152. The Appellant contended that the 2021 Report was "information made available" to the hypothetical officer and was an additional document which made the officer "aware" of the insufficiency in the APE 2020 tax return. The 2021 Report contained nothing new in relation to the 2020 claim. The difference was that HMRC opened an enquiry into the 2021 claim and Mr Patel wrote, challenging the claim, on 13 April 2022 which was before the LDE. The Appellant argued that this should have alerted HMRC to the insufficiency in APE 2020, as there was an overlap in projects.
153. Paragraphs 44(2) and (3) define information made available. They provide:

"(2) For this purpose information is regarded as made available to [an officer of Revenue and Customs] if—

(a) it is contained in a relevant return by the company or in documents accompanying any such return, or

(b) it is contained in a relevant claim made by the company or in any accounts, statements or documents accompanying any such claim, or

(c) it is contained in any documents, accounts or information produced or provided by the company to [an officer of Revenue and Customs] for the purposes of an enquiry into any such return or claim, or

(d) it is information the existence of which, and the relevance of which as regards the situation mentioned in paragraph 41(1) or (2)—

(i) could reasonably be expected to be inferred by [an officer of Revenue and Customs] from information falling within paragraphs (a) to (c) above, or

(ii) are notified in writing to [an officer of Revenue and Customs] by the company or a person acting on its behalf.

(3) In sub-paragraph (2)—

"relevant return" means the company's company tax return for the period in question or either of the two immediately preceding accounting periods, and

"relevant claim" means a claim made by or on behalf of the company as regards the period in question…"

154. Sub-paragraphs d(i) and d(ii) are potentially relevant.
155. In order for paragraph d(i) to apply, the hypothetical officer would have to be able to infer, from the 2020 Report and APE 2020 tax return and the 2020 claim itself that there was another document which was relevant. The 2020 Report indicated that a number of projects were "ongoing", but it would not be reasonable to expect the officer to infer from that that the 2021 Report existed and was relevant to the insufficiency in the 2020 tax return.
156. Turning to d(ii), the 2021 Report was certainly information which was notified in writing to HMRC by the Company, but the Appellant must also show that it would have been clear from the 2021 Report that it was relevant to the insufficiency of tax in 2020.
157. In a prominent position, underneath the heading for each project, was the start date. The start dates for the four ongoing projects were stated to be in 2019. There were also references to what was done in the previous year in the body of the Report. This would not, of itself, alert the hypothetical officer to an insufficiency in APE 2020.
158. If the hypothetical officer was aware of an insufficiency in APE 2021, he should realise from this that there might be an insufficiency in 2020, but he would not know whether an R&D claim was made in 2020 or what it related to. It might raise a suspicion that he ought to look into the 2020 return (which is what the actual officers did) but that would require taking additional steps to look at other documents which the hypothetical officer is not permitted to do.
159. We conclude that the 2021 Report is not information which would clearly be relevant to an insufficiency in tax for APE 2020.
160. Accordingly, the 2021 Report was not information made available to the hypothetical officer.

Decision

161. For the reasons set out above, we have decided that the hypothetical officer should reasonably have been aware at the LDE of the excessive claim for R&D Relief and the consequent insufficiency in tax. Accordingly HMRC is not entitled to raise a discovery assessment under paragraph 44(1) for APE 2020.
162. We allow the appeal.
RIGHT TO APPLY FOR PERMISSION TO APPEAL
163. This document contains full findings of fact and reasons for the decision. Any party dissatisfied with this decision has a right to apply for permission to appeal against it pursuant to Rule 39 of the Tribunal Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009. The application must be received by this Tribunal not later than 56 days after this decision is sent to that party. The parties are referred to "Guidance to accompany a Decision from the First-tier Tribunal (Tax Chamber)" which accompanies and forms part of this decision notice.
Release date: 01^st MAY 2025