Strata R&D Tax Group
Case studyJuly 15, 2026

Case Study: How a SaaS Company Identified $1.1M+ in R&D Tax Credits

Graphite Connect identified more than $1.1M in federal R&D credits for work its engineering team was already doing. Here is what a credit study actually looks like from the inside.

Client
Graphite Connect
Vertical
saas
Credit
$1.1M
Case Study: How a SaaS Company Identified $1.1M+ in R&D Tax Credits

TL;DR

  • Graphite Connect, a B2B SaaS company, identified more than $1.1M in federal R&D tax credits working with Strata's team. Individual results vary based on each company's activities, documentation, and facts.
  • The credits came from work the engineering team was already doing: building and improving the product itself.
  • A study typically starts with a $0 initial assessment and often runs four to eight weeks, most of it carried by the credit team rather than yours.

What is a seven-figure credit actually made of?

When SaaS founders hear that a company identified over a million dollars in R&D tax credits, the reaction is usually some mix of interest and skepticism. What did they do that we don't? The honest answer, in most cases: nothing exotic. They built software. The difference is that someone looked carefully at how that work maps to the federal credit and documented it properly.

Graphite Connect, a B2B SaaS company in the supplier management space, is a useful example. Working with Strata's team, Graphite identified more than $1.1M in federal R&D credits. Individual results vary, and no two studies land in the same place, but the anatomy of the claim is worth walking through because the underlying activities look like what most funded SaaS engineering teams do every quarter.

Graphite Connect Case Study Cover

Read the full case study: How Graphite Connect's CFO Turned the R&D Tax Credit Into a Growth Lever (PDF)

What the federal R&D credit covers

The credit for increasing research activities comes from IRC Section 41, with qualification standards in Treasury Regulation 1.41-4. Activities generally need to pass the IRS four-part test:

  • Technological in nature. The work relies on computer science or engineering principles. Software development typically clears this easily.
  • Permitted purpose. The goal is a new or improved product or process in terms of function, performance, reliability, or quality.
  • Technical uncertainty. At the outset, the team was uncertain about capability, method, or appropriate design.
  • Process of experimentation. The team evaluated alternatives systematically: architecture spikes, prototyping, load testing, iterative refactoring toward a performance target.

Claims are filed on IRS Form 6765, supported by documentation tying qualified wages, contractor costs, and certain supply and cloud costs to the qualifying work.

Where the qualifying work usually lives in a SaaS company

For a company like Graphite, and for most SaaS businesses, the bulk of a credit tends to come from engineering payroll. Depending on the specific activities, documentation, and facts, qualifying work may include:

  • Designing and building new product functionality where the right technical approach was uncertain
  • Rearchitecting for scale, reliability, or performance (database redesign, caching strategies, service decomposition)
  • Developing integrations that required real technical problem-solving rather than following a published recipe
  • Building data pipelines, search, or machine learning features through iterative experimentation
  • Substantial improvements to security architecture or multi-tenancy design

What typically does not qualify. Routine bug fixes and maintenance, UI styling changes with no technical uncertainty, configuring off-the-shelf tools, content and documentation work, sales engineering, and customer support generally do not meet the four-part test. A credible study separates these out rather than sweeping the whole engineering payroll in, which is exactly the kind of overreach that creates audit risk.

What the process looked like, step by step

Every engagement differs, but the Graphite study followed the shape most of our SaaS engagements do:

  • $0 initial assessment. A short conversation and document review to gauge whether qualifying activity and credit-eligible spend likely exist. No fee, no commitment.
  • Technical interviews. Structured conversations with engineering leads to identify projects with genuine technical uncertainty and experimentation. Typically a handful of hours of your team's time, not weeks.
  • Cost mapping and documentation. The credit team ties qualified wages and eligible costs to the qualifying projects and builds the substantiation file that supports the number.
  • Delivery. A completed study your tax preparer can file from, with documentation designed to hold up if the IRS ever asks questions.

Start to finish, a study often runs four to eight weeks depending on records and complexity. The recurring theme from clients is that the burden on their own team was smaller than they expected, because most of the evidence already exists in commit histories, tickets, and payroll records.

FAQ

We're not profitable yet. Does a credit even help us?
It may. Qualified small businesses can often elect to apply the credit against payroll taxes rather than income tax, which can put real money back into a pre-profit company. Eligibility rules apply, so this is worth checking early.

Our engineers don't track time against projects. Is that a dealbreaker?
Usually not. Studies commonly reconstruct reasonable allocations from the records a dev team already produces: repos, sprint boards, tickets, and interviews. Better contemporaneous records strengthen a claim, but their absence rarely kills one.

Can we claim credits for prior years?
Often, yes. Amended returns can generally reach back to open tax years, typically the prior three. A first-time study frequently covers multiple years at once, which is part of how totals can grow.

Is a number like $1.1M realistic for us?
It depends on your engineering spend and how much of it qualifies. Graphite's result reflects its specific facts; individual results vary. The point of the $0 assessment is to give you a grounded estimate for your situation before you spend anything.

If you want to know what your engineering work may be worth, start with a conversation.

This post is for informational purposes only and does not constitute tax or legal advice. Consult a qualified tax professional regarding your specific circumstances.

Author

Strata R&D Tax Group

← All insights