Did customizing prewritten business software qualify for Washington's former high-technology R&D B&O credit when projects followed known plans rather than experimental uncertainty?

Plain-English summary

The Washington Department of Revenue held that an IT consulting company's customization of prewritten business software did not qualify for the former high-technology research-and-development B&O credit. The work was technical, but it was not “nonroutine” under the rule because it did not involve experimental evaluation of alternatives or uncertainty about how to achieve the intended result.

The company supplied consultants for large software implementation projects. They customized off-the-shelf enterprise-management programs according to statements of work agreed with each client. The company did not develop its own hardware or software and was not licensed to alter the underlying programs.

For the former credit, “research and development” included technical and nonroutine activities translating technological information into new or improved products, processes, or software. Rule 24003 required nonroutine activity to pursue improvement through a process of experimentation where capability, method, or design was uncertain at the beginning.

The project documents instead identified the client's problem and a detailed plan before work started. Nothing showed testing alternatives to resolve essential uncertainty. Ordinary-language complexity or extensive programming did not satisfy the rule's specialized definition of nonroutine.

The work also appeared to adapt existing products rather than substantially improve them. The base software did not function fundamentally differently and was reused for later clients. A Board of Tax Appeals case involving creation of new software marketed to other manufacturers was factually different and did not change the result.

The source states that the credit statute expired January 1, 2015. The petition was denied.

What this means for you

Software consultants and implementers

Complex customization is not automatically tax-credit R&D. Preserve evidence of technological uncertainty, alternatives tested, experiments performed, and how the process resolved uncertainty.

Companies configuring prewritten software

Adapting a canned product to a client's known requirements can look like implementation or installation rather than creation of a fundamentally improved product.

Businesses reviewing historical credits

This ruling concerns a former Washington credit that the source says expired in 2015. It remains useful for understanding the Department's R&D definitions, not as evidence of a current credit.

Tax professionals

Apply the regulatory elements rather than ordinary meanings of “technical,” “custom,” or “nonroutine.” The experimentation and uncertainty requirements were decisive.

Common questions

Q: Was the company's work technical?

A: Yes. It applied computer-science methods, but technical activity alone did not satisfy the R&D definition.

Q: What was missing from the projects?

A: Evidence of a process of experimentation evaluating alternatives to resolve uncertainty about the method or design.

Q: Did customizing prewritten software substantially improve it?

A: Not on these facts. The underlying software was not changed to function fundamentally differently.

Q: Why did the cited Cimlinc case not control?

A: Cimlinc involved new or improved programs marketed to other manufacturers and lacked analysis applicable to this company's planned customization work.

Q: Is the high-technology credit still available?

A: The ruling states that former RCW 82.04.4452 expired January 1, 2015.

Citations and references

• Former RCW 82.04.4452(1), (2), and (7) (expired high-technology credit)
• RCW 82.63.010(16), (18) (qualified research and research-and-development definitions)
• WAC 458-20-24003(3)(c) (technical, nonroutine, experimentation, and substantial improvement)
• Cimlinc, Inc. v. Department of Revenue, BTA Docket No. 54862 (June 13, 2000)

Source

• Landing page: Washington Tax Decisions ZIP files
• Original PDF: 41WTD112.pdf
• Source archive: Volume 41 ZIP (member WTDvol41.zip!41 WTD 112.pdf)

Original ruling text

Det. No. 19-0164, 41 WTD 112 (April 12, 2022)

112

Cite as Det. No. 19-0164, 41 WTD 112 (2022)
BEFORE THE ADMINISTRATIVE REVIEW AND HEARINGS DIVISION
DEPARTMENT OF REVENUE
STATE OF WASHINGTON
In the Matter of the Petition for Correction of )
Assessment of
)
)
)
...
)
)

DETERMINATION
No. 19-0164
Registration No. . . .

RCW 82.63.010; FORMER RCW 82.04.4452; WAC 458-20-24003 – BUSINESS AND OCCUPATION TAX – HIGH TECHNOLOGY CREDIT – DEFINITIONS OF “QUALIFIED RESEARCH AND DEVELOPMENT” AND “NONROUTINE.” For the purposes of the high technology credit provided by former RCW 82.04.4452, taxpayer cannot claim to be engaged in “research and development” on the grounds that it performs “technical and nonroutine activities” unless the taxpayer’s work involves a process of experimentation to evaluate alternatives and the outcome of the work is uncertain. Headnotes are provided as a convenience for the reader and are not in any way a part of the decision or in any way to be used in construing or interpreting this Determination. Farquhar, T.R.O. – An information technology company that specializes in project management services and customization of preprogrammed business management software for clients protests the Department’s disallowance of a business and occupation (“B&O”) tax credit offered to companies engaged in “qualified research and development.” The company argues that because its business activities involve “translating technical and nonroutine activities into new or improved processes and software,” it was engaged in the type of research and development for which the credit is available. In fact, the company’s activities do not satisfy the Department’s definition of the term “nonroutine,” thus the activities cannot be considered “research and development” and the company does not qualify for the credit. Petition denied.1 ISSUE Whether customizing preprogrammed software for a client is a “nonroutine” activity that constitutes “qualified research and development” such that the taxpayer qualifies for the B&O high technology tax credit under RCW 82.04.4452.

1

Identifying details regarding the taxpayer and the assessment have been redacted pursuant to RCW 82.32.410.

Det. No. 19-0164, 41 WTD 112 (April 12, 2022)

113

FINDINGS OF FACT
. . . (“Taxpayer”) is an information technology company based in . . . Washington. Taxpayer does not design or develop its own computer hardware or software. Instead, Taxpayer “provides consultants to assist its clients in planning, designing, and executing on critical information technology initiatives.” Memorandum accompanying Taxpayer’s Review Petition (“Memo”), pg.

1. Taxpayer describes its consultants as “subject matter experts with extensive industry experience
   who apply specialized knowledge precisely and efficiently.” Id. Prior to beginning a project, Taxpayer and the client enter into an agreement that includes a description of the work Taxpayer will perform. Taxpayer’s consultants . . . use “off the shelf” or “canned” software, such as preprogrammed enterprise management programs, which the consultants customize to meet the client’s needs. [Taxpayer has no licensing agreements authorizing it to modify the software that they integrate for their clients.] Taxpayer frequently works with “large clients on large scale implementation projects” that “almost always require a significant level of customization.” Id. In October of 2017, the Department’s Audit Division (“Audit”) began a review of Taxpayer’s books and records for the period of January 1, 2014, through September 30, 2017 (“the Audit Period”). During the first year of the Audit Period, Taxpayer claimed a B&O high technology tax credit under RCW 82.04.4452 (“the Credit”) (the statute authorizing the Credit expired on January 1, 2015). Taxpayer argued that its work constituted “qualified research and development” in the field of advanced computing. Audit determined that Taxpayer’s activities did not meet the definition of “advanced computing” because it did not design or develop new computer hardware or software and disallowed the Credit. On April 24, 2018, Audit issued an assessment in the amount of $ . . . (“the Assessment”). The Assessment consists of $ . . . in service and other activities B&O taxes (including $ . . . for the disallowed Credit Taxpayer claimed for 2014 less a credit of $ . . . for over-reported service and other activities B&O taxes), a $ . . . assessment penalty, $ . . . in delinquent penalties, and $ . . . in interest. Taxpayer has not paid the Assessment. On June 18, 2018, Taxpayer submitted a timely petition for review (“the Petition”). In it, Taxpayer argues that it qualifies for the Credit and Audit’s decision to disallow it was improper. Taxpayer argues that its activities during the Audit Period involved “translating technical and nonroutine activities into new or improved processes and software.” Memo, pg. 5. It also argues that WAC 458-20-24003 “expressly notes that post-release software development can qualify as research and development for the purposes of the credit when it involves both technical and nonroutine activities concerned with translating technological information into improved software.” Id. Taxpayer states that its business involves “large scale integration projects requiring computer programming, which itself is nonroutine and results in the creation of new and improved processes and software.” Id. Finally, Taxpayer argues that the Board of Tax Appeals previously ruled that customization of “canned” software may qualify as research and development in Cimlinc, Inc. v. Dep’t of Revenue. Along with the Memo, Taxpayer provided 83 pages of statements of work between itself and several clients that document the type of work Taxpayer would perform on several projects. Taxpayer also provided a table summarizing sales to various clients and a copy of the Assessment notice, audit results, and the auditor’s Detail of Differences.

Det. No. 19-0164, 41 WTD 112 (April 12, 2022)

114

ANALYSIS
Former RCW 82.04.4452, which expired on January 1, 2015, provided the following business and occupation (“B&O”) tax credit for research and development spending: In computing the [B&O] tax imposed under this chapter, a credit is allowed for each person whose research and development spending during the year in which the credit is claimed exceeds 0.92 percent of the person’s taxable amount during the same calendar year. Former RCW 82.04.4452(1) (expired January 1, 2015). The statute then states how to calculate the value of the Credit. To begin, a taxpayer would determine its amount of “qualified research and development” by taking “the greater of the amount of qualified research and development expenditures of a person or eighty percent of amounts received by a person other than a public educational or research institution in compensation for the conduct of qualified research and development.” RCW 82.04.4452(2)(a) [(emphasis added)]. Therefore, to qualify for the Credit, a taxpayer must first show that it has “qualified research and development” costs or receipts. As relevant here, RCW 82.04.4452(7)(c) stated that “qualified research and development” had the same meaning as in RCW 82.63.010, which defines the term as follows: “Qualified research and development” means research and development performed within this state in the fields of advanced computing, advanced materials, biotechnology, electronic device technology, and environmental technology. 2 RCW 82.63.010(16). Thus, to meet the definition of “qualified research and development,” a taxpayer must demonstrate, first, that its activities constitute “research and development” and, second, that the research and development is performed within one of the five fields specified in RCW 82.63.010(16). Our analysis begins with whether Taxpayer’s activities involve “research and development.” RCW 82.04.4452 does not define the term, but RCW 82.63.010, the statute that provides definitions for terms used in tax deferrals for high-technology businesses, defines “research and development” in relevant part as follows: “Research and development” means activities performed to discover technological information, and technical and nonroutine activities concerned with translating technological information into new or improved products, processes, techniques, formulas, inventions, or software. . . . The term does not include adaptation or duplication of existing products where the products are not substantially improved by application of the technology, nor does the term include surveys and studies, social science and humanities research, market research or testing, quality control, sale promotion and service, computer software developed for internal use, and research in areas such as improved style, taste, and seasonal design. 2

The language contained in RCW 82.63.010(16) in 2014 remains in effect as of the date of this determination.

Det. No. 19-0164, 41 WTD 112 (April 12, 2022)

115

RCW 82.63.010(18) (emphasis added).
In other words, “research and development” generally consists of two distinct types of activities: activities “performed to discover technological information” and “technical and nonroutine activities concerned with translating technological information into new or improved products, processes, techniques, formulas, inventions, or software.” RCW 82.63.010(18). WAC 458-2024003(3)(c), which includes the same definition for “research and development,” clarifies that a taxpayer need not perform both types of activities in order to be “engaged” in research and development. Taxpayers “may perform either activity alone and be engaged in research and development.” WAC 458-20-24003(3)(c)(i). Taxpayer does not argue that its business activities are intended to “discover technological information.” Instead, Taxpayer uses consulting experts and known software to develop business management tools for its customers based on a predetermined plan. Therefore, the first type of “research and development” activity does not apply to Taxpayer and our analysis turns to whether Taxpayer’s activities are “technical and nonroutine.” WAC 458-20-24003(3)(c)(iv) provides definitions for both “technical” and “nonroutine”
[activities]. Activities are considered technical if they involve “the application of scientific, engineering, or computer science methods or principles.” WAC 458-20-24003(3)(c)(iv)(A). WAC 458-20-24003(3)(c)(iv)(B) defines “nonroutine” as follows: An activity is nonroutine if it: (I)

Is undertaken to achieve a new or improved function, performance, reliability, or quality; and

(II)

Is performed by engineers, scientists, or other similarly qualified professionals or technicians; and

(III)

Involves a process of experimentation designed to evaluate alternatives where the capability or the method of achieving the new or improved function, performance, reliability, or quality, or the appropriate design of the desired improvement, is uncertain at the beginning of the taxpayer's research activities. A process of experimentation must seek to resolve specific uncertainties that are essential to attaining the desired improvement.

(Emphasis added).
There is no dispute that Taxpayer’s activities are “technical.” Providing assistance to its customers on information technology projects necessarily involves “the application of scientific, engineering, or computer science methods or principles.” However, to meet the second definition of “research and development” in RCW 82.63.010(18), Taxpayer’s activities must also be “nonroutine.” Taxpayer believes its activities are indeed

Det. No. 19-0164, 41 WTD 112 (April 12, 2022)

116

“nonroutine” because computer programming “itself is nonroutine and results in the creation of new and improved processes and software.” Memo, pg. 5. Taxpayer’s argument appears to be based on a more colloquial definition of “nonroutine” and does not address the definition provided in WAC 458-20-24003. As discussed above, “nonroutine” is specifically defined by WAC 45820-24003(3)(c)(iv)(B) as a process of “experimentation” to evaluate alternatives and where the outcome of the work is “uncertain.” Nothing in the record suggests that Taxpayer’s work involves “experimentation” or that the outcome of its projects are “uncertain.” The statements of work provided by Taxpayer show that both parties are aware of the client’s problem, as well as Taxpayer’s detailed plan to resolve the problem, before work begins. Therefore, Taxpayer is not using a process of experimentation to evaluate alternatives. Additionally, it appears from the facts that the parties believe this plan will resolve the problem, suggesting that the outcome is not “uncertain.” For those reasons, we find that Taxpayer’s activities do not meet the definition of “nonroutine”, and Taxpayer’s activities cannot be considered “research and development” as defined by RCW 82.63.010. Though Taxpayer argues that WAC 458-20-24003 “expressly notes that post-release software development can qualify as research and development,” the argument fails because Taxpayer’s work with “post-release software” does not involve both technical and nonroutine activities. Because Taxpayer is not engaged in “research and development,” it cannot satisfy the definition of “qualified research and development” and we need not address whether Taxpayer’s activities constitute “advanced computing.” Because our analysis turns on the definition of “nonroutine,” we need not discuss whether Taxpayer satisfies all other components of the definition of “research and development” under RCW 82.63.010. Nevertheless, we do note that Taxpayer’s activities do not appear to result in a “substantially improved” product. RCW 82.63.010(18) states that the definition of “research and development “does not include adaptation or duplication of existing products where the products are not substantially improved by application of the technology.” [“A product is substantially improved when it functions fundamentally differently because of the application of technological information.” WAC 458-20-24003(3)(c)(v).][3] Taxpayer does not design or develop its own hardware or software and any customization it performs is for the purpose of fitting a “canned” product to a customer’s specific needs. The original product – the “canned” software – is not changed, let alone substantially improved, by Taxpayer’s activities. [Indeed, Taxpayer is not authorized to make any changes to the underlying software. Because Taxpayer does not change the software, the software does not function in a fundamentally different manner as a result of Taxpayer’s services—the software functions exactly how the creator designed the software.] If Taxpayer subsequently sells the same program to another customer, the sale would involve the same base product that Taxpayer would then customize to the new customer’s unique needs. In
[3 WAC 458-20-24003(3)(c)(v) defines “substantially improved” as: A product is substantially improved when it functions fundamentally differently because of the application of technological information. This fundamental difference must be objectively measured. Examples of objective measures include increased value, faster operation, greater reliability, and more efficient performance. It is not necessary for the improvement to be successful for the research to qualify.]

Det. No. 19-0164, 41 WTD 112 (April 12, 2022)

117

that sense, Taxpayer’s activities are more akin to a complex installation process, not the development of new or improved products. Taxpayer cites to Cimlinc, Inc. v. Dep’t of Revenue, a case that was heard by the Washington Board of Tax Appeals (“BTA”), to support its argument that customizing “canned” software can qualify as advanced computing. See Cimlinc, Inc. v. Dep’t of Revenue, BTA Docket No. 54862 (June 13, 2000). In Cimlinc, an out-of-state corporation sold “canned,” or preprogrammed, software to a Washington customer and sought to take the research and development credit. The company also provided “programming services necessary to implement the canned software, and [contracted] to develop new software programs as needed.” Cimlinc, pg. 2. The BTA found that the “necessary research and development done for this customer resulted in creation of new or improved software programs which Cimlinc markets to other unrelated manufacturers.” Id. The decision did not include a discussion as to how the BTA concluded that the company’s activities constituted “research and development.” Instead, after quickly concluding that Cimlinc was engaged in “qualified research and development,” the bulk of the BTA’s analysis is spent on how the credit is calculated and whether the credit can be assigned to another person. While we will distinguish the Cimlinc case on factual grounds, we first note that nothing in the statute authorizing review of Department of Revenue actions by the Board of Tax Appeals states that its decisions are binding on the Department, except with regard to the specific parties before the Board. See RCW 82.03, et seq. Taxpayer is correct that the BTA allowed the company to take the credit in Cimlinc, but the decision lacks a useful analysis of the issue before us here, namely whether the company performed statutorily-defined “research and development.” Additionally, Cimlinc is factually distinct from the case at hand. As noted above, the decision lacks an explanation of how the BTA determined Cimlinc’s activities constituted “research and development,” which is the root issue in this case. Without such an analysis, we cannot apply the BTA’s reasoning here. Furthermore, the cases are factually different because Cimlinc actually did produce “new or improved software programs” which it marketed to other unrelated manufacturers. Id. Here, as discussed above, Taxpayer does not produce any new or improved software to market to other customers. Because the facts are different, and relevant analysis is lacking, Cimlinc does not alter our analysis herein. DECISION AND DISPOSITION Taxpayer’s petition is denied. Dated this 20th day of June 2019.