The Puzzle of Non-Refundable SR&ED Tax Credits

‍‍In a recent post I talked about what actually matters in the recent SR&ED changes.There’s a lot in there and much of it is genuinely meaningful for founders.

There's also one major change that we've had a lot of questions about, not surprisingly. There is confusion, interest and concern around it and that deserves a closer look.The new Pre-Claim Approval (PCA) process is getting positioned as a win for founders, more certainty, faster timelines, less risk.

On the surface, that sounds great. We're not sure it will play out that way for most companies.

‍What Pre-Claim Approval Is (Without the Spin)

‍At a high level, PCA lets you submit your planned R&D work to the CRA before you do it and get a decision on whether it qualifies.If you’re approved:

• You get a faster review later
• You have some upfront clarity
• You can reference that approval for future claims

It’s optional, nothing changes if you ignore it.But we argue that how it's described isn't the key here, it’s how it actually changes behaviour.

‍What the PCA Actually Introduces

‍This is the part people are dancing around. You are:

• Writing a technical narrative upfront
• Defining your uncertainties in advance
• (potentially), sitting with a CRA reviewer to walk through your work

That’s not just “guidance.” It’s a front-loaded eligibility review, now happening earlier than it used to.To be fair, this is solving a real frustration. SR&ED has always had a timing problem:

• You invest first
• You find out later if it qualifies (sometimes much later)
• You get reimbursed (ofter 18+ mths after the work is done)

The PCA flips that: “Tell us what you’re going to do, and we’ll tell you if it counts.”

We get it! If you’ve had a bad experience before, that’s appealing.

‍Where This (May) Break Down for Founders

1. You're Locking in Your Plan Too Early

‍R&D doesn’t always go to plan. In the real world it evolves, pivots and often fails in unexpected ways. When you define your work upfront, you’re:

• Creating a fixed narrative
• Increasing the risk of misalignment later
• Potentially needing to rework or re-justify

The system was designed to evaluate what actually happened. This moves you toward defending what you said would happen.

‍2. You Start Optimizing for Approval

‍Once founders know they’re being evaluated early, behaviour changes. We predict that we'll see:

• Safer project framing
• Less technical ambition
• Fewer edge cases

Not because the work is changing but because the needs to frame it might and therefore, the presentation changes. We'd argue that over time, this (might) quietly reduce the strength and size of claims.

‍3. It Adds Process at the Wrong Time

‍PCA requires a lot from a company, all before they've created real R&D output.

• Documentation
• Internal alignment
• Time with reviewers

For early and growth-stage teams, that’s backwards. Our advice is always to spend time and energy building, selling and networking. You should be spending that energy on building, not pre-justifying.

‍4. It (May Not) Remove as much Risk as Advertised

‍This is the important part. The PCA is meant to help assess technical eligibility. It does not fully protect your claim. You still may run into issues down the line with:

• Costs
• Execution differences
• Supporting documentation

What that really means is that you're not eliminating risk, you’re just shifting where it shows up.

‍Who This Might Work For

‍It's not all bad news! And of course with any new change, there will be a learning curve that we all will ride along together.We believe there are a few cases in which the PCA could make great sense:

• First-time claimants who want guardrails
• Teams with highly ambiguous work and no internal experience
• Companies that are extremely risk-averse

We support all of these types of clients often and in those situations, the PCA may prove very useful.We would just argue that this is not the profile of most scaling companies.Ultimately, I stand by our first post. The companies that will be able to maximize their SR&ED claims annually:

• Do real technical work under uncertainty
• Document it properly as they go (we can help with this!)
• Build a clear, defensible narrative, factoring in all of the evolutions, pivots and fails as noted above; often the strongest part of a SR&ED claim!

If you have questions about the PCA or are considering taking advantage of it for your business, we're happy to chat and help you determine if it might be the right fit for you.

Last week, Canada passed the most significant updates to the SR&ED program in over a decade. After years of advocacy from many groups and Canadian leaders, major reforms to Canada’s Scientific Research and Experimental Development (SR&ED) program take effect today.

You’ve likely already seen the headlines and all the "hot takes" but the real question is what these changes mean for how you build your company.

For years, SR&ED hasn’t kept pace with how modern companies operate. Limits didn’t move with inflation, capital-heavy innovation was at a disadvantage, and many companies outgrew the program earlier than they should have. What we’re seeing now is less about generosity and more about the government catching up.

Arguably the most active group in pushing for these updates is the Canadian Council of Innovators and as members of CCI, we’ve had a front-row seat to many of the conversations shaping these changes.

The direction is clear: this is about making SR&ED more relevant to how companies actually build today.

The Updates

On paper, the changes are straightforward:

• capital expenditures are eligible again,
• expenditure limits for CCPCs have doubled from $3M to $6M,
• taxable capital thresholds have expanded,
• and certain public companies can now access refundable credits (Canadian-listed firms, that are Canadian-controlled and their Canadian-based subsidiaries)

Individually, these look like incremental improvements. Taken together, they materially increase the amount of non-dilutive capital available to innovative companies.

For some businesses, refundable SR&ED can now reach over $2M annually. That’s no longer a marginal benefit, it’s meaningful runway.This shift also changes who benefits most. Bringing capital expenditures back into the program is particularly significant for companies investing heavily in infrastructure, equipment, or technical systems. Hardware, manufacturing, and deep tech businesses, historically disadvantaged under SR&ED, are now much better positioned to benefit. At the same time, higher limits and expanded eligibility mean the program extends further into the growth stage, and the belief that this is just for "early stage" companies can be rebutted.

I think these changes should change behaviour in material way.

The Mindset Shift

Historically, SR&ED has been treated by most companies as a retrospective exercise. You build, you document, and you file. You optimize around the edges.

We tell our customers (and have been for some time!) that this is an outdated and ineffective approach.  And now, when the program becomes this material, and when capital and scale are more meaningfully included, it starts to influence decisions upstream. Hiring plans, technical direction, capital allocation, and the timing of investment all become part of the equation. SR&ED moves from something you deal with after the fact to something you can design around and plan for in a strategic way.

New: The Pre Claim Approval Process

One area where we see a lot of optimism, and where we take a more cautious view, is in the new Pre-Claim Approval process (PCA). In theory, this offers certainty before work begins, along with faster processing and reduced audit risk. On the surface, that’s appealing. But innovation rarely follows a fixed path.

Projects evolve, scope shifts, and technical challenges change direction. Locking in assumptions early may limit flexibility later, whether that means under-claiming or constraining how work develops. Certainty has value, but in this context, it might come at a cost. For many companies, especially those iterating quickly, flexibility could bemore important.That doesn’t mean pre-claim approvals won’t have a place. It does mean they should be approached thoughtfully, not automatically and we will continue to watch this space.

Final Thoughts

Overall, the risk is that companies interpret these changes too narrowly. It’s easy to see this as simply “more funding” or a larger claim opportunity. In reality, the bigger shift is strategic.

SR&ED is no longer just about maximizing what you did, it’s increasingly about shaping what you choose to do.The companies that benefit most won’t be the ones that file the "best" claims. They’ll be the ones that plan more deliberately, align their technical and financial decisions earlier, and treat SR&ED as part of how they build, not just how they report.

I think we can all agree that the Canadian government has made it easier to fund innovation with these updates. This is a meaningful step forward but it doesn’t change the fundamentals. Building a great company still comes down to execution, judgment, and timing. These program improvements have been a long time coming and as CCI noted, "now the focus is on making sure it works in practice and continues to improve."

SR&ED is a more powerful tool than it was before. What matters is how you choose to use it.

‍

The build vs. buy decision in non-dilutive funding

Managing non-dilutive funding internally requires a trade-off between saved fees and lost engineering velocity. While the "0% fee" of a Do-It-Yourself (DIY) approach is appealing, it often fails to account for the substantial internal cost of diverted technical talent and the unhedged risk of compliance reviews. This decision is rarely just about the invoice at the end of the process; it is a calculation of internal hours versus the value of specialized expertise in maximizing claim size and defensibility.

For the busy executive, the instinct to keep the process in-house typically stems from a mandate to preserve cash flow and reduce administrative overhead. However, treating the annual SR&ED tax credit or any number of grant applications as a purely administrative task underestimates the technical rigour required by the Canada Revenue Agency (CRA) for SR&ED and various funding bodies (for grants).

When technical leaders divert their focus from product development to writing and compliance, the company incurs a hidden cost that does not appear on a balance sheet but impacts the product roadmap significantly.

We view this as an engineering resource allocation problem rather than a simple accounting decision. You must weigh the certainty of professional fees against the variable and often higher costs of internal time, potential claim reductions, and the significant disruption of a defensibility audit. The pragmatic approach involves analyzing where your team adds the most value: writing code that generates revenue or writing technical narratives to reclaim costs.

The math: Internal engineer hours vs. Agency fees

Quantifying the true cost of a DIY claim reveals that a robust filing typically costs $12,000+ in direct engineering time plus significant opportunity costs, often exceeding the perceived savings of avoiding agency fees. A defensible claim typically requires 40–80 hours of time from senior technical leaders, such as the Chief Technology Officer (CTO) or Vice President (VP) of Engineering, to scope projects and draft narratives that meet legislative requirements.

Calculating opportunity cost

If a CTO spends 80 hours on claim preparation and compliance, they are not shipping product. With an estimated internal cost of $150/hour for senior engineering time, the direct investment is $12,000. However, this figure only accounts for the base salary and overhead. It does not factor in the context switching required to move from high-level architecture or team management to retrospective technical documentation.

The indirect cost, lost velocity on the roadmap, can be significantly higher.

When a technical lead pauses development to compile timesheets, comb git commits for evidence, and draft technical narratives, the entire engineering team may experience a slowdown. In a fast-moving market in 2026, delaying a feature release by two weeks to accommodate a tax filing deadline can result in missed market opportunities that far exceed the cost of the filing itself.

On the grant side of things, we see that often poor ROI on successful grants can be associated with internal attempts, where high-value hours are spent on low-leverage administrative drafting rather than core innovation.

Comparing the agency model

Consultants typically charge a success-based fee ranging from 10% to 20% of the claim value. This model aligns incentives: the partner is paid only when funding is received, offloading 80–90% of the heavy lifting. By engaging an advisory partner, the internal time commitment often drops to 5-10 hours of document gathering, high-level interviews and reviews. The advisor handles the labour-intensive tasks of mapping financial data to technical activities and ensuring the language meets specific CRA criteria.

The reality of "free" filings

There is no such thing as a free claim. You are either paying an external expert from the refund or funding the process with expensive, pre-tax engineering payroll. When companies attempt to save the advisory fee, they often effectively "pay" that amount or more through inefficient internal processes. Furthermore, internal teams lacking specialized knowledge may inadvertently leave eligible expenditures unclaimed, effectively paying an "ignorance tax" that reduces the net benefit of the program.

The risk: Audit risk in non-dilutive funding

Compliance exposure is the single largest variable that DIY cost models fail to capture, with anecdotal industry data suggesting significant review and audit rates for Scientific Research & Experimental Development (SR&ED) claims.

While being selected for a review is not necessarily an indication of wrongdoing, it is a procedural reality that demands significant time and resources to manage.

The cost of defence

Defending a claim without support is a massive distraction. A typical review process can require 200+ hours of preparation, evidence gathering, and meetings with CRA officials. This "distraction tax" can paralyze a technical team for months. The CRA review process involves deep dives into technical obstacles, contemporaneous evidence, and financial allocations. Without a partner to filter requests and organize data, the technical team must educate themselves on tax law in real-time while trying to defend their engineering work.

Advisory insulation

Firms that employ ex-CRA auditors build defence into the process from day one. By preparing claims that anticipate specific  SR&ED advisory services  triggers, they reduce the likelihood of extensive inquiries. An experienced advisor understands how to frame "technological uncertainty" and "systematic investigation" in ways that align with  Section 37 of the Income Tax Act , which governs the eligibility of scientific research and experimental development.

Pricing the risk

When making the decision, you must price in the probability of a significant time penalty.  CRA program service standards  indicate that refundable claims selected for review are processed within 180 days, but the internal disruption can last much longer. For a company focused on cash flow, the delay in funding combined with the resource drain of a defence can be damaging. An advisory partner provides a layer of predictability, ensuring that if a review happens, it is a managed process rather than a crisis.

Next step: Review your past three filings for documentation gaps that could trigger a compliance review.

The hidden friction of DIY compliance

Issues with retroactive documentation is the primary cause of claim denial and compliance issues, as the CRA expects evidence generated at the time the work was performed rather than reconstructed logs. DIY teams often scramble to reconstruct logs at year-end, a practice that creates inconsistencies during reviews and weakens the defensibility of the claim.

Best practice is to gather evidence at the time the work is performed. Your advisors can recommend ways to implement systems to capture this evidence throughout the year, ensuring data is defensible by default. Relying on git commit logs or Jira tickets alone is often insufficient; the CRA looks for evidence of the process of experimentation, such as hypothesis testing, analysis of failures, and iterative conclusions.

Navigating government funding rules

Stacking funding requires precision. For example, understanding how the Industrial Research Assistance Program (IRAP) interacts with tax credits is critical. Misclassifying "government funding" or failing to account for it correctly can lead to accidental double-dipping and mandatory repayments. Distinguishing between "government assistance" (which grinds down your SR&ED claim) and a "bona fide loan" (which typically does not) is a technical distinction with massive financial implications.

Systematic evidence gathering

Experienced advisors transform compliance from a year-end panic into a background process. This ensures that if a review happens, the data is already organized and prepared for review. By integrating funding strategy into the company's operational rhythm, the "tax season" becomes a review of existing documentation rather than a creation of new documents.

Next step: Implement a monthly technical meeting log to capture contemporaneous evidence of hypotheses and failures.

Securing capital without stalling the roadmap

Non-dilutive funding should serve as fuel for your innovation, not a brake on your development velocity. The goal is to capitalize the company and protect your cap table while keeping your most valuable assets, your engineers, focused on solving technical problems. Choosing an advisory partner brings the "adult in the room" to your funding strategy, replacing guesswork and retroactive scrambling with predictability, stability, and a process that preserves your equity.

When you calculate the cost of lost engineering hours, the risk of a 200-hour audit defence, and the potential for missed funding opportunities, the "savings" of a DIY approach often evaporate. A strategic partnership allows you to maximize your non-dilutive intake while maintaining the focus required to grow your business.

If your R&D expenditures are scaling, it is time to evaluate whether your current filing method is costing you more than you realize.  Book a discovery call to audit your funding strategy  and discuss how to maximize your return while minimizing your risk.

FAQs

Q: What is the typical success fee for an SR&ED consultant?

A: Fees generally range from 10-20% of the approved claim. This model ensures the partner is only compensated when you successfully receive funding.

Q: Does hiring a consultant guarantee I won't be audited?

A: No partner can guarantee a zero-audit rate, as the CRA uses random selection alongside risk-based selection. However, working with a firm that utilizes ex-CRA auditors ensures your claim is constructed to minimize compliance inconsistencies and that you have professional representation if a review occurs.

Q: Can I just write the technical part and hire an accountant for the numbers?

A: Similar to building a product without a blueprint, this hybrid approach creates risk. A common cause for reviews is a misalignment between the technical narrative and the financial schedules. An integrated partner ensures these two halves of the claim tell a consistent story, preventing discrepancies that trigger reviews.

Q: How much time does my engineering team need to commit if we use an advisor?

A: A streamlined advisory process typically requires some up front work in the first year from your technical lead for interviews and final review, and less as subsequent years go by. This is compared to the 40–80 hours required for a DIY filing. This allows your senior technical talent to remain focused on product development.

Please note that this content is for general informational purposes only and does not constitute tax or legal advice. We hope this analysis provides a helpful overview for founders navigating the 2026 funding landscape.

‍

‍