CIRD81995 - R&D tax relief: conditions to be satisfied: DTI guidelines (2000): HM Revenue & Customs commentary

(This commentary was issued by the Inland Revenue in April 2003 to assist in interpretation of the 2000 DTI guidelines.)

Commentary on the guidelines on the Meaning of R & D for Tax Purposes

Introduction

1. R&D tax credits are a positive measure, intended to focus support on innovative activities that benefit UK industry and indeed, deliver wider benefits for the UK. For this to be successful it is necessary that a boundary be drawn between activity that qualifies and activity that does not - so a definition of "research and development" is required.
   
   
2. For tax purposes, what constitutes R&D is defined by reference to accounting standards, which are set out in SSAP13, and to the ‘guidelines on the Meaning of R&D for Tax Purposes’ issued by the Secretary of State for the DTI (‘the guidelines’). This commentary is intended to help in the interpretation of the guidelines; it does not supersede them. Paragraph references in this document are to paragraphs of the guidelines unless otherwise indicated. Not all the paragraphs in the guidelines require further comment and as such several do not feature in this commentary.
   
   
3. The guidelines and this commentary must both be read in their entirety in order to understand their meaning and intent. This is because many concepts in the definition of R&D are interrelated and cannot be applied in isolation. Taking extracts out of context can produce misleading conclusions.
   
   
4. Paragraphs 1 to 10 of this commentary are introductory. Paragraphs 11 to 78 of this commentary are extracted from the Corporate Intangibles R&D (CIRD) Manual, which has been prepared for the staff of HMRC. They are being published for the information of taxpayers and their advisors in accordance with the Code of Practice on Access to Government Information.
   
   
5. It should not be assumed that the CIRD manual in general or this commentary in particular are comprehensive, nor that they will provide a definitive answer in every case. The staff of HMRC are expected to use their own judgement, based on their training and experience, in applying the guidance to the facts of particular cases. In particular difficult or complex cases they are able to obtain further guidance from specialists in Head Office.
   
   
6. The CIRD manual is based on the law as it stood at date of publication. HMRC will publish amended or supplementary guidance if there is a change in the law or in the Department's interpretation of it.
   
   
7. Subject to these qualifications, readers may assume that the guidance in this commentary given will be applied in the normal case; but where HMRC considers that there is, or may have been, avoidance of tax the guidance in this commentary will not necessarily apply.

Tax relief for R&D

8. Prior to the introduction of the R&D tax credits in April 2000, tax relief was given for expenditure on ‘scientific research’. Claims were concentrated in the area of pure science. The introduction of the guidelines made clear that R&D for tax purposes is undertaken in other sectors such as engineering and software.
   
   
9. Most current R&D expenditure qualifies for a deduction - that is, it receives 100% tax relief. In addition, there are three special types of relief for expenditure on R&D.

• R&D allowances give 100% relief for virtually all capital expenditure on R&D.
• R&D tax credits for SMEs, introduced in April 2000, give an extra 50% relief (making 150% in all) for current R&D expenditure on staff costs, consumable stores and some sub-contract costs.
• R&D tax credits for large companies, introduced in April 2002, give an extra 25% relief (making 125% in all) for current R&D expenditure on staff costs and consumable stores.

10. ‘Staff costs’ are defined in the R&D tax credit legislation to include only those staff working actively and directly on R&D. The indirect activities listed at paragraph 29 of the guidelines do not qualify for the R&D tax credits.

Commentary on the guidelines

11. Paragraph 3 sets out the basic principles of what constitutes R&D activity for tax purposes, which are then expanded upon in the rest of the guidelines. To be R&D, an activity must be:

• creative work i.e. non-routine work, work containing novel elements or outcomes,
• undertaken on a systematic basis i.e. excluding one off or ‘lucky’ discoveries. Novelty, innovation or uniqueness in the product or process is not sufficient to demonstrate technological or scientific advancement. It is how such attributes arise that is important. An effort to achieve scientific or technological advancement will be accompanied by experimentation or analysis in a situation where there is scientific or technological uncertainty about whether or how the advance can be achieved.

Of course many important advances have resulted from fortuitous discoveries (for example, penicillin and saccharine), and a subsequent effort to exploit such a chance discovery could be R&D. For example, a laboratory doing routine analyses of chemical samples to assure the quality of a production process discovers a new and potentially useful substance in one of its samples, arising from a defect in the production process. The discovery of this substance does not mean that the analysis work can be reclassified as R&D. However, if a separate programme of work is then carried out to determine how the new substance was formed and what its properties are, then this would be R&D.

• to increase the stock of knowledge not just the knowledge within the company or to use the stock of knowledge to devise new applications etc. R&D activity can lead to the creation of a new product, but the creation of a new product is not necessarily the result of R&D. In a business context, this means that when a new or improved product or process is created, it must embody a scientific or technological advancement to be qualifying R&D. For the avoidance of doubt, this means that development work can qualify on an equal basis with research work.
• within the fields of science or technology, excluding the humanities and social sciences. See paragraph 16 for the treatment of activities outside the scope of science and technology but which are integral to an R&D project.

12. Paragraph 4 sets out the relationship between the accountancy definition of R&D in SSAP13 and the guidelines. To be R&D for tax purposes an activity must meet the requirements of both, but in this regard, only that part of SSAP13 to do with defining qualifying activities should be considered. The remainder relating to accountancy treatment (for example the treatment of subcontracted R&D) has no relevance in this context.
    
    
13. Paragraph 5 outlines the three basic categories of R&D. Of these, ‘development’ requires the most clarification of its nature and limits for tax purposes, including the need to establish what constitutes substantial improvement.
    
    
14. The key question in a number of sectors, such as engineering, where technology generally advances by small amounts at a time, is ‘what constitutes substantial improvement?’
    
    
15. Substantial improvement means to change or adapt something to the point where it is obvious to a competent professional user that the ‘improved’ version is different to the original. As with all R&D, work aimed at substantial improvement must be creative (contain novel elements, involve innovation) and involve scientific or technological advance. If something falls well within the current capability of the industry sector as a whole, then even if it is a major improvement on a previous version, work aimed at developing it by a company is not R&D (but see also paragraphs 19 and 20 of this commentary).
    
    
16. It is important to stress that substantial improvement is not limited to companies at the very leading edge of R&D activity in a field or sector. As long as the work represents innovation relative to the sector as a whole, it may be R&D. For example, the existence of high-fidelity audio equipment does not prevent activity to create lower-performance equipment from being R&D (for example, through technological advances leading to lower cost through innovative circuit design or speaker construction). However, if a company is simply catching up with the general state of the industry the work will not be R&D.
    
    
17. Paragraph 6 identifies the presence of an appreciable element of innovation as a requirement for an activity to be R&D for tax purposes. We interpret ‘appreciable’ as meaning ’significant or perceptible’ i.e. a non-trivial advance. In this context innovation is the creation or the attempted creation of a new device, product or process resulting from study and experimentation or the introduction of something novel.
    
    
18. It is not sufficient that something is novel or innovative to the company. The work must be more than merely duplicating what has been done before.
    
    
19. It can be difficult on occasions to establish and verify whether something is novel or innovative. But we accept that there can be innovation where several companies are working at the cutting edge in the same field, and are doing similar work independently.
    
    
20. There can also be circumstances in which companies are carrying out genuine innovation unaware that someone else has already covered the same ground. Where work has already been done but this is not known in general because it is a trade secret, and another company repeats the work, this would not be barred on the grounds of duplication. Nor would it be barred if it were known that something has been achieved, but the details of how were not publicly available. In judging such circumstances we would have regard to whether or not a competent scientist or engineer, or R&D manager, in the relevant field ought to have been aware that the ground had already been covered (for example, if the information could have been obtained by a search of the Patent Office databases).
    
    
21. Paragraph 7 comments on qualifying activities listed in SSAP13.
    
    
22. Testing will fall within the guidelines up to the point of resolution of scientific or technological uncertainty, discussed below in the context of paragraph 10. Testing which does not form part of the resolution of scientific or technological uncertainty is almost always a routine activity and hence is not R&D.
    
    
23. A prototype is an original model on which something new is patterned, and of which all things of the same type are representations or copies. It is a basic experimental model possessing the essential characteristics of the intended product. The design, construction, and testing of prototypes generally fall within the scope of R&D for tax purposes. This applies whether only one or several prototypes are made, and whether they are made at the same time, or one following the other. Constructing several copies of a prototype after successfully testing the original is not part of R&D for tax purposes. The role of prototypes in development is further discussed in the commentary on paragraph 19 of the guidelines. The use of prototypes in the software field is covered in the commentary on paragraphs 21 - 28 of the guidelines.

The boundary of R&D and other related activities

24. In paragraphs 9 - 20 the guidelines explore the dividing line between R&D and other related activities.
    
    
25. Paragraph 10 outlines the importance of the concepts of ‘novelty’ and scientific or technological uncertainty in determining whether an activity is R&D. For work to be R&D it must both be novel and involve the resolution of scientific or technological uncertainty. (This should be understood in the context of the basic principles set out at paragraph 5).
    
    
26. ‘Novelty’ means that work must be intended to lead to a scientific or technical advance or substantial improvement of a product, process or service. In other words, for activity to be R&D, it must be aimed at creating something - which need not be tangible; new knowledge is a legitimate aim of R&D - or at improving substantially something that already exists. Periodic changes to existing products, services or processes - even though they may represent some improvement - would normally be excluded from R&D, unless they involve an appreciable element of innovation and either break new ground or represent technical advances (paragraph 14). Note that R&D requires only that the aim of work should be substantial improvement; if the result of such work is incremental improvement then it is still R&D (paragraph 18). Work which aims to make only incremental improvements is not R&D, but there are likely to be activities which, taken as a whole, result in incremental improvements, but which include some elements of much more significant development. Such activities would count as R&D.
    
    
27. It follows also that a rolling programme of R&D, intended to lead, over the years, to substantial improvements but which is exploited during that time to produce successive (incremental) improvements to products, should qualify.
    
    
28. R&D also includes ‘novel work which draws on or creates a new source of knowledge which might lead to the breaking of new ground or a technical advance and which subsequently might entail the creation or development of a new or substantially improved product, process or service’ (paragraph 13).
    
    
29. ‘Scientific or technological uncertainty’ means whether there is reasonable doubt whether something is scientifically possible or technically feasible, or how to achieve a desired technological advance. It is presumed that those carrying out the work will have the appropriate skill, experience and knowledge. Where such a person is able to specify, with reasonable confidence, how to achieve the advance, the activity would be routine and not R&D.
    
    
30. The uncertainty may derive from a number of sources, such as the need to engineer to a particular set of constraints or specification, or the need to make a complex system of components function effectively. (Note however that the converse is not true: the need to meet a particular specification does not of itself mean that there is any technological uncertainty.) It is assumed that a company would be able to list the technological uncertainties being resolved (or that have been resolved) in the course of the R&D project.
    
    
31. In engineering and software engineering, technological uncertainties will typically arise from turning something that has already been established as scientifically feasible into a cost effective, reliable and reproducible product or process. In the case of software projects, for example, uncertainties may arise due to system complexity, feature interaction and software conflicts.
    
    
32. Scientific and technological uncertainty applies at the level of the project as a whole: individual tasks contributing to the overall project need not themselves be subject to technological uncertainty, as long as the whole project is aimed at resolving technological uncertainty (paragraphs 13 - 15, 25 - 27). In other words, routine tasks that are required to help resolve the overall scientific or technical uncertainty are R&D for tax purposes if the project viewed as a whole is R&D. This includes the writing of software to assist in carrying out the work (paragraphs 25 - 27) - and the use of research in the humanities and social sciences where it has bearing on resolving the technological uncertainty (paragraph 16). For example, the testing of a prototype device or system (including a software prototype) may itself be straightforward, but it will count as R&D if done as part of a project to develop a novel device. Because of the structure of the R&D tax regime a subcontractor carrying out a routine task as part of a larger project e.g. testing, may not qualify for the tax credits even though the activity would qualify if the client company undertook it directly.
    
    
33. The scope of the term ‘project’ in the previous paragraph of this commentary may not be obvious. Consider a company that designs and manufactures petrol engines for cars. At one extreme, it may wish to develop improved spark plugs for an existing engine. The technological uncertainty associated with this work is resolved once prototype plugs have been fully tested in the engine: the ‘project’ therefore comprises the work up to the end of that testing (including work which would, considered alone, be ‘routine’ - such as the testing). At the other extreme, the company may require an engine incorporating new spark plugs, a new combustion chamber design, lighter materials and other improvements such that the overall engine is substantially improved (for example, using less petrol to achieve slightly greater performance, generating less pollution). In this case, not only the work on the spark plugs, but also development of the engine through to testing the prototype is a ’project’.
    
    
34. Work done before tackling a technological uncertainty (e.g. pre-production planning or bringing a firm’s level of expertise up to the norm or beyond for the sector) does not count as R&D (paragraph 12). Work done after the technological uncertainty is resolved is not R&D (paragraphs 13 - 15). Work not relevant to resolving the technological uncertainty but still relevant to the project (for example, cosmetic or marketing-related changes to a product’s design, or work on intellectual property rights connected with the product) is not R&D (paragraph 13).
    
    
35. Technical and scientific planning activities directly supporting an R&D project can qualify as part of that project. These may include defining scientific or technological objectives, assessing scientific or technological feasibility, identifying scientific or technological uncertainties, estimating development time, schedule, and resources, and high-level outlining of the scientific or technical work, as well as the detailed planning and management of the work. Other elements of a company’s planning activity relating to a project but not directly contributing to the resolution of scientific or technological uncertainty, such as examination of the project’s financial, marketing, and legal aspects, are not R&D within the meaning of the guidelines.
    
    
36. The set of activities directly contributing to the resolution of the overall scientific and technological uncertainty at the project level are R&D, within the meaning of the guidelines (as long as the project as a whole is novel, in the sense described above).

Scientific and Technological uncertainty arising from complex systems

37. Scientific and technological uncertainty may result from the complexity of a system rather than uncertainty about how its individual components behave. For example, in electronic devices, the characteristics of individual components or chips are fixed, but there can still be uncertainty about the best way to combine those components to achieve an overall effect. However, simply assembling a number of components (or software sub-programs) to an established pattern or following routine, known methods for doing so involves little or no uncertainty and is therefore not R&D.
    
    
38. Similarly, work on combining standard technologies, devices, and/or processes can be R&D if non-trivial combinations of established, known technologies and principles for their integration carry a major element of technological uncertainty; this may be called a ‘system uncertainty’. If the technological specifications or objectives to resolve the ‘system uncertainty’ are such that the basic design of the underlying technologies must be changed to achieve the integration, the overall project may be R&D. (In both cases, the intended outcome of the work must be novel if it is to be R&D).
    
    
39. In contrast, if R&D is done to develop a new component to slot into an existing system, or a system, which is substantially the same as an existing system, there is unlikely to be significant system uncertainty. Assuming that use of the new component in the existing system constitutes a substantial improvement to the system, then necessary development work on closely related components are also R&D. If substantial change to the system is required to incorporate a new component or components, and the work is intended to create a new or substantially improved product, process or service, this is also R&D.
    
    
40. Paragraph 12 refers to the requirement that the activity must be intended to lead to a substantial improvement in the business’s products, processes or services.
    
    
41. ‘Substantial improvement’ - mentioned in the notes on paragraph 5 of the guidelines above - means to change or adapt something to the point where it is obvious to a competent professional user that the ‘improved’ version is in some way ‘better’ than the original. How big a change is necessary to achieve this depends on the state of knowledge and technology in a particular industrial sector; expectations in some newer or faster-moving sectors can differ from those in more established or constrained ones. For example, historically, it was long accepted that the power of new computers roughly doubled every 18 months, so a chip which was 1% faster might not be considered much of an improvement (all other things being equal), but an increase of 1% in the efficiency of electricity generation from, say, a wind turbine may well be a ‘substantial improvement’.
    
    
42. Scientific and technological improvement is not limited to physical properties: reducing the production cost (and hence the price) for a product could be a substantial improvement, but again how big a reduction counts depends on conditions prevailing in the particular sector.
    
    
43. ‘Substantial improvement’ in the context of engineering and other development is contrasted by the guidelines with ‘incremental’ development (paragraph 18). Simply changing or updating a product, process or service is not of itself sufficient to qualify as substantial improvement (paragraph 14). If a product, process or approach is substantially set then any further changes to it are not R&D (paragraph 13).
    
    
44. In judging whether there was intent to make ‘substantial improvement’ of a product, process or service, we would have regard to the nature of the improvement or improvements sought by the R&D project, and arguments as to why they would constitute a substantial improvement. (This could for instance include examples of new applications made possible by the improved version, such as devices made practical by markedly lighter batteries.)
    
    
45. Paragraph 13 introduces the phrase ‘experimental development’ to emphasise that not all work that a company might view as development is necessarily R&D in the meaning of the guidelines. ‘Experimental development’ should therefore be read as meaning ‘activity which is development and hence R&D within the meaning of the guidelines’, i.e. work which is subject to technological uncertainty and aimed at creating new or substantially improved materials, devices, products, processes and services.
    
    
46. This contrasts with ‘commercial development’ involved in turning a functioning prototype into a final commercial product is not R&D; this exclusion for example covers some items such as market research which are specifically excluded from R&D by SSAP13 (paragraph 8) and the activities outlined in paragraph 31. However, it may be necessary to undertake further R&D before a product can be manufactured and sold; paragraph 58 of this commentary gives some examples of the kind of work that may be required.
    
    
47. Paragraph 13 highlights that R&D is still taking place while the (scientific and technical) viability of earlier R&D is being established or further technical improvements are being made, as long as the overall requirement for novelty is also fulfilled. Once work moves on from eliminating technological uncertainty (i.e. the product, process or approach is substantially set) then further work on ‘finishing’ the product, process or approach is not R&D.
    
    
48. Paragraph 14 distinguishes R&D from periodic updating.
    
    
49. Work on periodic updating is not R&D for tax purposes unless it involves an appreciable element of innovation (see paragraph 6). As discussed above, the outcome of a programme of R&D may be an incremental or series of incremental improvements.
    
    
50. Periodic updating includes activity that is directed at changing the physical appearance or superficial characteristics of a product without altering its utility, efficiency or function.
    
    
51. Paragraph 15 covers the border between R&D and subsequent commercial development. It is recognised that after the initially identified uncertainty has been resolved there may be situations where further R&D is required.
    
    
52. A project is complete when the activities associated with resolution of the scientific and/or technical uncertainty are complete.
    
    
53. When the technological character of the product or process is substantially set, and the primary objective is to develop markets, to do pre-production activity, or to get a production or control system working smoothly, then the work is no longer R&D for tax purposes. However, if further scientific and/or technological uncertainty exists under these circumstances, then work on studies to resolve these scientific and/or technological problems may still be R&D for tax purposes.
    
    
54. Paragraph 16 excludes humanities except as part of a wider qualifying R&D project. For example in the development of night vision on-screen instrumentation for pilots it may be necessary to conduct ergonomic and psychological testing as part of the wider project.
    
    
55. Paragraph 17 draws out the idea that it is the context or aim of activities rather than the activities alone that determines whether they amount to qualifying R&D, as discussed in the commentary on paragraph 10 of the guidelines.
    
    
56. Paragraphs 19 & 20 refer to the use of prototypes and pilot plant as a guide to identifying the end of the R&D activity.
    
    
57. In many cases the process or product is developed to the prototype or pilot stage for experimental or technical trial purposes. That is, prototypes are used to test the feasibility of the concept or hypothesis. Possibly, the construction of a whole series of pilots or prototypes may be involved, as problems are met and either overcome or bypassed. It may be that, in this phase of the development, the original objectives have to be modified significantly or perhaps even changed entirely, depending on the technological opportunities that become apparent. Such work would fall within R&D for tax purposes.
    
    
58. Subsequently the product or process is developed to meet the requirements of commercial use or production. This may involve activities such as scaling up from pilot plant size to commercial size, certification, or developing an economical means for commercial production. In some cases, the technologically successful prototype may have been fabricated using methods or materials that are too expensive to be practical for commercial use. Such activities are not necessarily of themselves R&D for tax purposes but if they require scientific or technological uncertainties to be resolved, then further R&D may take place. For example, in scaling up a process, it may become clear that the behaviour of a substance in bulk is markedly different from its behaviour in the pilot plant in a way that can only be understood by carrying out new R&D.

Software (paragraphs 21-28)

59. The guidelines distinguish between two sorts of software activity which may qualify as R&D:

(i) software as the aim of the R&D (i.e. work aimed at creating software or knowledge about software engineering), and

(ii) software used as a means of achieving an R&D goal, within a larger R&D project (paragraphs 25-28). In both cases, software is given equal treatment to other forms of technological activity (paragraph 21).

60. It is recognised that functionality is often implemented in products using software rather than hardware. The technological uncertainties that have to be addressed in creating software are the same in concept as those faced by engineers seeking a physical implementation of a product or process. In the case of software projects technological uncertainties may arise, for example, from system complexity, feature interaction and software conflicts.

Software as the aim of the R&D (paragraphs 21-24)

61. For a project to be software R&D (i.e. R&D in sense (i) above), it must fulfil the same criteria as any non-software activity to qualify as R&D:

• Scientific or technological uncertainty, which will be resolved if the R&D project is successful;

and

• ‘Novelty’, which in the context of development means seeking to create new or substantially improved materials, devices, products, services or systems.

62. The guidelines include some examples of work in the software field that might qualify as R&D (paragraph 22) and of software-related work that is considered to be routine/non-novel and therefore does not qualify as software R&D (paragraphs 23 and 24). This is meant as an illustration of how the fundamental tests might apply to software, rather than a hard and fast rule.
    
    
63. For example, an ‘enhanced query language’ (paragraph 22) would need to constitute a significant improvement on the original query language to be ‘novel’, while extensive ‘adaptation’ of software could in some circumstances require development of new functionality and new technical knowledge, and therefore could fulfil the uncertainty and novelty criteria and hence count as R&D - it would in effect create a substantially improved product, service or system.
    
    
64. Software R&D is usually aimed at resolving technological rather than scientific uncertainty. This can include (paragraph 7):

• design of products, processes, services or systems involving new technology or substantially improving those already produced or installed;
• testing in search for, or evaluation of, product, service or process alternatives;
• design, construction and testing of pre-production prototypes and models and development batches;

and

• construction and operation of prototypes.

65. ‘Prototypes’ in this context can be read as ‘builds’ of the software in question, for example, alpha and beta stages of the normal development and testing process prior to commercial release although the criteria for the end of R&D still apply: once the scientific or technological uncertainty is resolved, or the structure and functionality of the software is substantially set (paragraph 13). As with all R&D, if, subsequently, unforeseen problems (e.g. system and software conflicts) arise, resolving them may constitute an additional phase of eligible R&D.
    
    
66. Projects aimed at developing the initial release of saleable software that is ‘leading edge’ in some technological way are often R&D for tax purposes. To develop software at the leading edge of today's technologies generally requires the developer to come up with new constructs, such as new architectures, algorithms or database management techniques (i.e., make substantial technological advance), and there can be specific uncertainties as to the viability of these. If the software's competitive edge stems merely from advance in an area other than technology, such as business management, or improvements in financial management techniques, the project is unlikely to be eligible unless implementation requires technical innovation.
    
    
67. As with any other product, simply claiming to have developed the first or best software suite for a given purpose does not in itself prove that the taxpayer has made a technological advancement. A new and unique software suite can be built using only well known combinations of constructs, tools and methods without technological advancement. This is analogous to designing and building a unique and complex office building without making any advances in the field of civil engineering.
    
    
68. Note that an advance in technology can rarely be described by listing software functions and features at an ‘end-user’ level. Advances are typically made through innovation in software architectures, designs, algorithms, techniques or constructs.
    
    
69. There is always some uncertainty about anything. In software development, as in other scientific and technological fields, uncertainties that can be resolved through brief discussions with peers, or simply through a few lines of analysis, are routine design uncertainties rather than technological uncertainties. Likewise there is routinely the need to calibrate or optimise or clean up new software. These challenges are not technological uncertainties unless it can be shown there is a fundamental problem with the technologies that must be addressed. Refer to paragraph 29 of this commentary.
    
    
70. In engineering or software, as in other fields, to determine whether or not R&D is taking place the test to be applied to a particular project (see paragraph 33 of this commentary for a definition of ‘project’) is whether persons of appropriate skill and expertise would be able to specify, with reasonable confidence, how to achieve the desired advance. Where such a person is able to specify, with reasonable confidence, how to achieve the advance, the activity would be routine and not R&D.
    
    
71. As with other forms of R&D, the scope of the overall R&D project is defined by the technological uncertainty: once the uncertainty is substantially resolved (for example, once the overall structure/architecture of the software and of its key component routines and engines is resolved and functioning) the R&D is over. Subsequent tasks such as the preparation of user documentation will not be R&D. Refer to paragraph 29 of this commentary.
    
    
72. As with R&D in general, not every component of or activity involved in a software R&D project needs to be novel in order for the whole project to be novel.

Software used as a means of achieving the R&D goal (paragraphs 25-28)

73. Software is frequently used as a tool or mechanism to help resolve technological uncertainty, i.e. to conduct research and development. The guidelines refer to this as ‘Software-based R&D’, but ‘Software-using R&D’ is probably a more accurate general description. Again, software has equal treatment with other forms of technological activity and is treated just like any other tool or equipment used to further an R&D project. Individual tasks contributing to the overall project need not themselves be subject to technological uncertainty, as long as the whole project is aimed at resolving technological uncertainty (paragraphs 13 - 15, 25 - 27).
    
    
74. This includes the use of existing or slightly customised software to assist in carrying out the work (paragraph 25). Paragraphs 26 and 27 give examples of situations where software is used as a tool as part of a larger R&D project. Note that the larger project must itself be R&D. Also note that the cost of acquiring or obtaining access to such software is not qualifying expenditure for the R&D tax credits.
    
    
75. Paragraph 28 describes a situation where a project (analysis of market research data) lies outside the scope of science and technology, and hence would not be R&D. For this reason, developing software for the purpose would not constitute software-based R&D. However, if the software development was expected to result in a significant scientific or technological advance such as a new query language, then the work contributing to that advance would be software R&D, but in the sense (i) described above (relating to paragraphs 21 - 24).

Use of Software in Engineering

76. Functionality in technological products is often implemented through software. As long as there is technological uncertainty to be resolved, both software and hardware implementations of function would be R&D in the context of the overall project. It is conceivable that a situation might arise in which implementing a hardware solution was routine (and as a result technological uncertainty was at an end) while a software solution still had significant uncertainty associated with it, or vice versa. In this case, work on the ‘uncertain’ solution would be R&D while work on the ‘routine’ solution would not, because all of the technological uncertainty associated with the project had been resolved. However, in practice there will usually be continuing technical uncertainty associated with the project regardless of how function is implemented.

Qualifying indirect activities

77. Paragraph 29 lists a number of qualifying indirect activities. For the purposes of the R&D tax credits, only staff costs of those actively and directly engaged in R&D, consumable stores and certain sub-contracting costs are allowable. The list in the guidelines does not overrule the tax legislation prescribing qualifying expenditure.
    
    
78. Paragraphs 30 - 31 list a number of exclusions. R&D would normally exclude work on:

• market research or sales promotion;
• calibration of standards and quality control or routine testing of materials, devices, products, or processes;
• routine data collection;
• research in the social sciences or the humanities;
• the commercial production of a new or improved material, device, or product, or the commercial use of a new or improved process;
• style changes;
• periodic alterations to existing products even though they bring about some improvement;
• operational research not tied to specific R&D activity;
• corrective action in connection with break-downs during commercial production;
• legal and associated work on patent applications, records and litigation;
• acquisition of rights in, or arising from, R&D;
• construction, relocation etc. of facilities or equipment other then those used solely for R&D;
• education and training and provision of scientific or technical information;
• routine computer maintenance or software development;
• specialised, but routine medical care;
• policy studies.