What are the Biggest Mistake When it Comes to SR&ED?

The biggest mistake founders make with SR&ED is treating it as a year-end paperwork exercise instead of a year-round discipline. 

SR&ED is genuinely the largest source of non-dilutive R&D funding available to Canadian startups, and that's even more true after the 2026 SR&ED Enhancements, which doubled the expenditure limit for the enhanced 35% refundable credit to $6M and widened the phase-out range to $15M–$75M in taxable capital. That's real money that can have a powerful impact, but we still see the same pattern trip up first-time claimants over and over: They wait until tax season, then try to reconstruct months of technical work from memory. 

This strategy fails because a defensible SR&ED claim needs two things at once: The technical narrative (what uncertainty were you resolving, what did you try, what didn't work) and the financial substantiation (salaries, contractor time, materials mapped precisely to that work). The problem is those two things usually live in different people's heads. Your engineering lead can tell you exactly why the caching layer redesign was genuinely uncertain, while your finance lead can tell you exactly what was spent. Almost nobody can do both, however, and by the time you're assembling the claim retroactively, the connective tissue between the two is gone. You're left guessing, and CRA reviewers can tell the difference between documentation and reconstruction. 

The smart approach is contemporaneous tracking instead of retroactive assembly. 

Practically, that means: 

• Capture technical notes as you go, in whatever tool your team already lives in (tickets, PR descriptions, sprint notes). You're not writing the claim yet, just leaving a trail of what problem you were solving and why the answer wasn't obvious. 

• Tag time and cost data to projects in real time, not in April. Even a rough weekly estimate of hours per initiative beats a reconstructed guess. 

• Loop in someone who can translate between the two languages, whether that's an internal hire, an advisor, or an SR&ED partner, before you're staring down a filing deadline. 

Do this consistently and the claim itself becomes almost mechanical to assemble each year, because you're not starting from zero every cycle. Skip it, and even a technically strong R&D program can end up under-claimed, or worse, indefensible if the CRA asks questions. 

One more thing worth knowing given the recent changes: Capital expenditures are SR&ED-eligible again for property acquired after December 15, 2024, after being excluded since 2014. If your team wrote that category off as ineligible, it's worth a second look. 

Happy to go deeper on any part of this if useful to the group. 

Great insights Paul! Thanks for sharing. Planning it early and having treating as year-round is the key as SHRED can greatly help with early-stage product development.