The North American seed market didn’t dry up in 2026. It narrowed. Capital efficiency stopped being something founders discussed after the round closed and became the filter that decides whether the round happens at all. Fewer checks are going into a smaller set of companies, and the ones clearing the bar are built differently than the ones that raised in 2021.
That is not a reason to despair. It is a reason to change how you design the company on day one — headcount, spend, and structure — so that the raise becomes an accelerant rather than a life-support system. Here is what the data actually says in both markets, and what to do about it.
The 2026 funding picture in Canada and the United States
Start with the American side. Carta’s State of Pre-Seed report for Q1 2026 shows roughly 3,000 U.S. startups raising pre-seed capital in the quarter — a stable total pool, but an increasingly uneven distribution of it. Rounds between $1 million and $2.5 million made up 24% of all pre-seed rounds in Q1 2023; by Q1 2026 they were 18%. Meanwhile, AI startups now capture roughly half of all pre-seed dollars, up from about 30% a few years ago. The middle is thinning while the extremes grow.
Canada shows the same concentration with sharper edges. RBCx’s 2026 mid-year market check-in found that among the pre-seed and seed companies it banks, 162 were actively raising a combined $510.7 million in January 2025 — but by March 2026 only 61 were raising, seeking roughly $189.8 million. On the fund side, the top five Canadian funds captured 46% of all capital raised in 2023 and 80% by 2025. Emerging managers, historically the most willing to back first-time founders, raised roughly $2.8 billion over three years against an expected $4.3 billion — a 36% shortfall. The CVCA recorded the lowest quarterly deal count since 2017 in Q1 2026: 104 deals totalling $936 million, with a single growth-stage deal completed.
Two conclusions follow. First, capital exists but reaches fewer companies. Second, a thin growth-stage market raises the bar at entry, because early investors underwrite the next round as much as this one. If your plan assumes a Series A in eighteen months, you are underwriting a market that may not show up.
Capital efficiency is a design decision, not a spending decision
Most founders treat efficiency as restraint — spend less, hire slower, negotiate harder. That’s cost control, and it caps the downside without changing the shape of the business. Real capital efficiency is architectural. It’s decided when you choose what the company does itself, what it buys, what it automates, and which roles you make permanent.
The measurable proxy is revenue per employee. SaaS Capital’s 2026 benchmarks, drawn from a survey of more than 1,000 private SaaS companies, put the median at $141,125 per employee, up from $129,724 the year before. The breakdown is more instructive than the headline:
- Companies at $1M–$3M ARR run a median of $109,644 per employee.
- Equity-backed companies at $5M–$10M ARR sit at $152,295.
- Bootstrapped companies at the same $5M–$10M band reach $177,240.
Bootstrapped companies out-earn equity-backed peers per head at every ARR level. That gap isn’t a moral lesson about venture capital — venture-backed companies typically grow faster — but it does tell you what discipline produces when capital is scarce by default rather than by choice. Pick a target from that table before you write your first job posting, then hire against it.
A four-part operating design for lean company building
1. Sequence hires by irreversibility, not by urgency
Every early role falls somewhere on a reversibility scale. A fractional controller, a contract designer, an agency running paid acquisition — all reversible within a month. A VP of Sales with a team, a second engineering pod, an office lease — effectively permanent, because unwinding them costs morale and momentum as well as money.
Fill the reversible roles first and keep them reversible until the underlying function has a repeatable, measurable output. Convert to full time only when you can name the metric the role owns and show twelve weeks of data on it. This single rule prevents most of the headcount regret we see in early-stage companies.
2. Buy the boring, build the differentiated
Draw a hard line between the parts of your product that customers pay for and the parts they simply expect. Authentication, billing, notifications, log aggregation, document storage, and analytics plumbing are expected. Building them in-house consumes senior engineering capacity that should be pointed at the thing no competitor has.
Audit this quarterly. The honest question is not “could we build this?” but “if we built this, what would we not build instead?” Most teams discover they are maintaining infrastructure that costs more in engineering salary than the vendor bill they were avoiding. This is the same lens we apply when we structure and build companies — the differentiated surface area should be small, deep, and defensible.
3. Run non-dilutive capital in parallel with equity
Both countries subsidize R&D directly, and both programs are underused by early-stage teams who assume the paperwork isn’t worth it. In Canada, the Scientific Research and Experimental Development (SR&ED) program administered by the CRA offers refundable credits to qualifying Canadian-controlled private corporations — meaningful because refundable means cash, not just a deduction against profits you don’t have. In the United States, the federal research credit allows certain qualified small businesses to apply the credit against payroll tax liability rather than income tax.
The operational point: both require contemporaneous documentation. Retrofitting a claim from memory eighteen months later is where most of the value gets lost. Set up time tracking and technical narratives from the first sprint, not at year end.
This isn’t tax, legal, or financial advice, and eligibility rules change. Work with a qualified accountant in your jurisdiction before relying on either program.
4. Put governance in before you need it
Clean governance is a capital efficiency measure because it removes friction at exactly the moment friction is most expensive. A cap table with unresolved founder equity, undocumented contractor IP assignments, or handshake advisor grants will cost you weeks of diligence and real negotiating leverage. In a market where investors have their pick of deals, an unclean company is an easy pass.
The minimum viable set, in place before you raise: incorporation in a defensible jurisdiction, vesting on all founder shares, signed IP assignment from every person who has ever touched the codebase, a documented option pool, and board mechanics you can actually operate. None of this is expensive early. All of it is expensive late.
What “default alive” looks like on a 2026 cap table
Paul Graham’s framing — is the company profitable before it runs out of money on current growth — is more useful now than when he wrote it, because the bridge round that used to rescue a default-dead company is harder to find. Practically, that means three numbers on the wall:
- Runway in months, calculated on committed spend, not optimistic revenue.
- Months to default alive at current growth and current burn — if that number exceeds your runway, you have a structural problem, not a sales problem.
- Revenue per employee, tracked monthly against the benchmark for your ARR band.
Founders who can present those three numbers with confidence raise on better terms than founders presenting a hockey stick, because the first set is verifiable and the second is a hope. In a concentrated market, being underwritable beats being exciting.
How we think about it
Protocol 42 structures, builds, funds, and operates a small number of companies each year. We deploy capital selectively and at our discretion, which means we sit on both sides of this equation: we build the operating design and we carry the consequences of it. That vantage point is why we’re blunt about lean structure — not as an austerity posture, but because companies designed for efficiency have more strategic options than companies designed for scale they haven’t earned yet.
The market that funded growth at any cost is gone for now. The market that funds companies with a credible path to standing on their own is very much open. If you’re building one and want operators in the room rather than advisors on a call, start a conversation with us.