title: "State of CSBFP Lending — methodology and what Capital Toolkit will track" description: "Capital Toolkit will publish an annual State of CSBFP Lending report — anonymized observations from Canadian small-business financing files. This post lays out the methodology: what the report intends to track, how the data will be collected, why each metric matters, and when the first full data drop arrives." date: "2026-05-22" author: "Capital Toolkit" tags: ["csbfp", "state of lending", "methodology", "canadian financing", "small business", "annual report"] videos:
- understanding-the-csbfp
- sbl-business-funding
The Canada Small Business Financing Program funds billions of dollars of Canadian small-business loans every year, but the file-level data on how those loans actually move through the system — how long they take to fund, which uses get approved at what rate, which documents trigger lender pushback, which sectors cluster — sits inside individual lender files and is never published at the program level.
Capital Toolkit is in the file-preparation seat. We see every document that goes into an application package. We see the outcome of the lender’s review. That position will, over time, generate practitioner-level data that nobody else holds, and we intend to publish it.
This post is the methodology document for the State of CSBFP Lending annual report — what the report intends to track, how the data will be collected, why each metric matters, and what the first full data drop will look like. The methodology lives at this URL permanently; each year’s data drop will be linked from here.
Why this report will exist
The federal Innovation, Science and Economic Development Canada (ISED) publishes program-level statistics on CSBFP — broadly, total dollars loaned and number of loans, with some sector aggregation. Those numbers are useful but coarse. They do not tell a Canadian small business owner what to actually expect from an application: how long it will take, what documents will get flagged, which sectors are clustering at which lenders, what the typical revision cycle looks like before a file lands at the lender.
The State of CSBFP Lending report is intended to be the practitioner-level view. It will be built from anonymized observations across the files Capital Toolkit prepares and places, plus whatever decline-side information lenders provide on a file-by-file basis. It is not a substitute for ISED’s program-level reporting; it is a complement, focused on what a borrower or advisor actually needs to know to plan a file.
What the report intends to track
The report is organized into six metric categories. For each, we will publish (a) the headline figure for the year, (b) the methodology paragraph explaining how the figure is constructed, and (c) the practitioner takeaway — what an owner or advisor should do with the number.
1. Time-to-funding. Median and 25th/75th percentile days from completed application to funded loan, segmented by loan use (real property, equipment, leaseholds, intangibles, working capital). Time-to-funding is the single most predictive variable for whether a borrower’s plans line up with reality — a six-week real-property close needs a very different timeline than a two-week equipment refinance.
2. Document-revision cycles. Average number of document revisions a file goes through between Capital Toolkit’s first review and lender submission. High revision counts signal where owners are systematically misjudging what lenders want; low counts signal categories where the canonical documentation pattern is already widely understood.
3. Lender placement patterns. Anonymized distribution of files across participating CSBFP lenders, broken out by loan use and loan size. This data reveals which lenders are actively underwriting which kinds of files at the time of the report — useful information that is not published anywhere else.
4. Rejection-reason breakdown. For files that get declined at the lender stage, the distribution of rejection reasons across the seven categories documented in 7 reasons CSBFP applications get rejected. This shows whether the binding constraint in a given year is documentation gaps, structure issues, credit problems, or genuine ineligibility — and how that mix changes over time.
5. Use-of-proceeds mix. Share of total loan dollars going to each eligible CSBFP use category (real property, equipment, leasehold improvements, intangibles, working capital). The mix is a leading indicator of where Canadian small businesses are deploying capital — outfitting, expanding, building.
6. Loan-size distribution. Distribution of approved loan amounts within the program’s sub-limits. This reveals where the practical ceiling actually sits for most files (often well below the $1,000,000 statutory maximum), which is useful context for an owner sizing their own ask.
How the data will be collected
Three sources will feed the report.
Capital Toolkit file data. Every file that runs through the Capital Toolkit qualification, document review, and placement workflow will be anonymized and indexed against the metric categories above. Personally identifying information is stripped before the file enters the dataset; what remains is shape data — loan use, dollar amount range, sector, days at each stage, number of document revisions, eventual lender, approval or decline. The aggregate-research use is disclosed at the point of collection in our privacy policy and PIPEDA page; borrowers may opt out of inclusion at any time by emailing privacy@capitaltoolkit.com.
Lender feedback, where lenders provide it. Some lenders return structured post-decision feedback on declined files — the specific underwriting issue, the documentation gap, or the structural conflict that drove the decline. Where this feedback is provided, it will be anonymized and bucketed against the rejection-reason categories. Where it is not, the rejection-reason data will be reported based on Capital Toolkit’s own observation rather than lender-reported reasons, and the report will note which is which.
Public CSBFP program data. ISED’s published program totals will provide the program-wide denominator against which Capital Toolkit’s volume share can be estimated. ISED’s aggregate publications are not granular enough to validate the report’s six specific metric categories — there is no public ISED data on time-to-funding or revision cycles — but they will anchor the report’s position within the broader program.
No individual borrower’s data will appear in published material, ever. All figures are aggregates with a minimum sample-size threshold before publication; categories with too few files for safe aggregation will be reported as “sample too small” rather than published with a misleading figure.
What the first data drop will look like
The first full State of CSBFP Lending report will be published once Capital Toolkit’s file dataset reaches the threshold where each metric category has enough underlying files to publish defensibly. The methodology above is the scaffolding; the data fills in once the sample is large enough to be representative rather than anecdotal.
A few honest framing points for the first drop:
- The numbers will be Capital Toolkit’s files, not the entire CSBFP universe. We are one participant in the program; the report is a practitioner snapshot, not a census.
- Year-one numbers are observational. Trends require year-two and year-three data to surface; comparisons between years only become meaningful starting with the second annual drop.
- Where Capital Toolkit’s numbers conflict with ISED’s program statistics, ISED is the authoritative source for the program as a whole. Our numbers describe our slice of it.
Why this matters for owners and advisors
Most CSBFP guidance available to a Canadian small business owner today is either (a) program-rules summary, which tells you what is allowed but not what actually happens, or (b) lender-marketing collateral, which tells you what the lender wants you to think happens. Neither tells you what to expect.
Time-to-funding, document-revision rates, lender placement patterns, and rejection-reason breakdowns are the kinds of numbers that change how a file is prepared. Knowing the median time from submission to funding on an equipment file lets an owner plan supplier conversations realistically. Knowing the rejection-reason mix for a given year lets a CPA or broker put the right effort into the right gap on the file. Knowing where loan sizes cluster within the sub-limits lets an owner size their ask without leaving room on the table.
The annual report exists to publish those numbers — once the sample is large enough to support them.
Subscribe (and what to expect)
The first full data drop will be published as a follow-up blog post linked from this methodology document. There is no separate mailing list; new posts appear on the blog index as they ship.
In the meantime, the file-level observations and program-mechanics analysis from each new Capital Toolkit blog post — comparison posts, definition pages, and how-to guides — fold into the dataset that this annual report will eventually summarize. The reading list lives at the blog index; the program-mechanics pillar is at the CSBFP overview.
Written by Capital Toolkit