The Canadian funding programs that publish their approval rates

Published 31 July 2026 · catalogue of 785 programs, read the same day.

Why only 21 programs out of 785

Canadian funders publish approval data inconsistently and usually by accident. Federal evaluations, departmental performance reports, access-to-information releases and the occasional press release contain the counts; program webpages almost never do. A funder is much more likely to tell you it received “strong interest” than to tell you it funded 27 of 567.

Our catalogue carries an approval-rate note on 489 of the 785 programs, and 348 of those notes state a percentage. But most of those percentages are our analysts’ estimates, reasoned from budget size, award counts and observed demand. Only 21 records cite a published numerator and a published denominator — the only case where the rate is an observation rather than an inference. Those 21 are what this article claims. The estimates appear once, near the bottom, clearly labelled as estimates and never attached to a program name.

Under 10%: five programs where nine in ten applicants get nothing

These are the programs where the published numbers are genuinely brutal. Note what they have in common: national scope, large per-award amounts, and a pool far wider than the budget.

Swipe to see all columns →

† The denominator is not applications. SITES compares funded projects to the ~2,757 eligible Indigenous tourism businesses nationally; AgriInnovate’s 557 are initial pre-screenings, not completed applications. Both are still the funder’s own published pair of numbers, so we show them — labelled.

The Black Entrepreneurship Loan Fund is the one to sit with, because it moved. In FY2021–22 it approved 118 of 2,838 applications, a 4.2% rate. The following year it approved 180 of 1,246 — 14.4%. Nothing about the applicants changed that much in twelve months. The pool shrank as the launch rush faded, and the odds more than tripled. Timing inside a program’s life cycle is a real, underrated variable.

10% to 19%: the band most competitive programs actually live in

This is where the serious, juried, nationally-advertised programs cluster. If you are applying to something with a fixed deadline, an expert panel and a scoring rubric, this is probably your real neighbourhood.

Swipe to see all columns →

† SSHRC’s figure is end-to-end across two stages: 22 of 81 Stage 1 applications were invited forward (27%), and 15 of those 22 were funded (68%). A single-stage reading of either number badly misstates the odds.

CIHR’s number carries the most uncomfortable detail in this whole census. Its own reporting notes that over 90% of reviewed applications are judged fundable, while 13.6% get funded. That is not a quality filter. It is a queue with a budget line drawn across it, and it is the honest picture of most oversubscribed national programs.

20% to 39%: the programs where a strong application genuinely decides it

Roughly one in three or one in four. At these odds the marginal quality of your application stops being noise and starts being the deciding variable.

Swipe to see all columns →

† Two denominators here are not applications. IRAP’s 9,187 is total clients reached in the year, not proposals submitted — many of those clients never reach a funding decision, so the true application-to-approval rate is higher than 34.1%. AgriStability is an entitlement program: 1,122 of 4,211 enrolled producers received a payment because their margins declined, not because a panel picked them.

Futurpreneur’s evaluation adds a figure nobody else publishes and everybody should: a further 38.7% of applications were assessed as viable but went unfunded for capacity reasons. Read together, roughly three quarters of complete applications were either funded or judged fundable. The binding constraint on that program is money, not merit — which is a completely different thing to be told than “36.6%.”

Above 40%: two programs, and neither means what it looks like

Swipe to see all columns →

† Newfoundland and Labrador’s restaurant program received only 21 applications in its 90-day window and deemed 11 eligible — that 52.4% is an eligibility screen, not a competition, and six of the guarantees issued expired unused. AgriMarketing’s 81% is the published rate for its associations stream against a published count of 124 applications, from the Growing Forward 2 era; the numerator is derived from that rate rather than separately published, and only national industry associations can apply. Its historical SME component ran at 54% of 314 applications.

A high approval rate is not the same as easy money

Before you use any of these numbers to choose a program, know what they can and cannot tell you. An approval rate answers exactly one question — what share of a pool got a yes. It does not tell you how much they got, whether the money covered the project, or how narrow the door was before anyone applied.

Contests are the least honest odds in Canadian funding

Branded small-business contests — the ones with a public vote, a photogenic winner and a five-figure cheque — publish winner counts constantly and applicant counts almost never. Our catalogue holds several that describe their pool only as “thousands of nominations” for five prizes, or “hundreds of applicants” for ten. None of them reaches the bar for this census, and that is not an accident of record-keeping.

Treat them accordingly. A contest is a marketing surface with a prize attached; the selection often turns on how many people you can mobilise to vote for you, not on the strength of the project. The realistic odds are worse than the worst assessed program in the table above, and the effort is not transferable to your next application the way a real funding application is.

What to do with odds this uneven

Here is the practical read. Because the spread between programs is 200-fold and the spread between two versions of the same application is nowhere near that, the single highest-leverage decision you make is which programs you apply to at all. Most founders get this backwards: they find one program, pour everything into that one application, and never learn that they cleared the gates on eleven others.

Three things follow from the data on this page:

How this census was built

The whole value of a number like “21 of 785” is the rule that produced it. Ours is below in full, including the judgement calls we made and the records we excluded.

Method note: what counts as a published approval rate

The inclusion rule. A program enters the census when a published source gives both a count of applicants funded and the size of the pool they were selected from, as specific figures. A percentage on its own does not qualify. A count of winners with no denominator does not qualify. A denominator described as “thousands,” “hundreds,” or a range such as “an estimated 50–70 applications” does not qualify.

One record enters on a variant of the rule and is marked in its table: AgriMarketing publishes a rate (81%) against a published application count (124), so the numerator is derived rather than separately stated. We show the derivation instead of hiding it.

Where the numbers live in our data

Every figure above is quoted from one of two fields on the program record in our catalogue: approvalRate or competitivenessNotes. Both trace back to departmental evaluations, performance reports, funder press releases, or in a few cases reporting by CBC and other outlets. Each program links to its own profile page, where the source URL and our last-verified date are shown.

The denominator caveat, program by program

Five of the 21 pairs count something other than submitted applications. We kept them and flagged them rather than quietly dropping them, because they are the funder’s own published pair — but the comparison is not like-for-like:

Swipe to see all columns →

Program	What the denominator actually counts
Indigenous Tourism Fund — SITES	Eligible Indigenous tourism businesses nationally (~2,757), not applications received
AgriInnovate	Initial pre-screenings (557), not completed applications
NRC IRAP	Total clients reached in the year (9,187); many never submit a proposal
AgriStability	Enrolled producers (4,211); payment is formula-driven, not competitive
NL Restaurant Loan Guarantee	All 21 applications received; 11 were deemed eligible, which is a screen rather than a selection

Deduplication

Several programs appear in our catalogue more than once — an umbrella record plus a stream record, for instance. Where two records carried the same published pair (Strategic Response Fund and its steel stream; Innovative Solutions Canada and its Challenge Stream; Futurpreneur’s core and newcomer records; Telefilm’s umbrella and Talent to Watch), we counted the program once and cited the more specific record. Twenty-one is the count of distinct programs, not of records.

What we excluded, and why it matters

Roughly seventeen further records came close and were cut: contests reporting winners against “thousands of nominations,” superclusters comparing funded projects to their member base, accelerators quoting cohort sizes, and a set of programs whose denominator was explicitly an estimate. Each exclusion makes the census smaller and the claim safer. If we had allowed estimated denominators, this article could have shown a table of about 40 programs and roughly half of it would have been guesswork wearing a decimal point.

The 324 estimates, and what they reveal about estimating

Our catalogue also carries a stated percentage on 324 further programs. Those are analyst estimates — reasoned from budget size, published award counts, intake frequency and observed demand — not published counts, and we do not attach them to program names here. In aggregate they are still interesting, mostly as a warning:

Swipe to see all columns →

Estimated approval band	Programs	Share
Under 10%	23	7%
10–24%	35	11%
25–49%	126	39%
50–74%	94	29%
75% and above	46	14%

82% of our estimates sit at 25% or higher. Only 38% of the published rates do. Either the programs that publish are unrepresentative — large national competitions are exactly the ones with evaluations — or estimating approval rates systematically produces optimistic numbers. Both are probably true. Treat any estimated approval rate, ours included, as directional at best, and treat the 21 published ones as the only ones you would bet on.

Counts read from our catalogue of 785 programs on 31 July 2026. Program status changes; the profile page linked from each row carries the current status and our last-verified date.