BDC LIFT Track A vs Track B — A Decision Guide with Worked Examples

BDC LIFT launched as a single program in April 2026 and quickly grew two distinct tracks as the funnel revealed two different SME shapes asking for it. Track A serves AI-only projects across every sector. Track B serves AI-plus-equipment projects in seven priority sectors. The cheque sizes, eligibility rules, and application emphasis are different enough that picking the wrong track adds months and risk to the file. This post is the decision framework we walk clients through in the first 20 minutes of a LIFT call.

1. The 30-second test

Before any of the detail, the cleanest decision rule:

That rule disambiguates about 85% of LIFT inquiries inside half a minute. The remaining 15% are hybrid cases (some equipment, some software-only) where the decision deserves a more careful look. We'll cover those in section 6.

The structural reason the rule works: Track A is designed for software-and-services AI projects with a $2M cap. Track B is designed for capital projects where AI is the productivity driver but the dominant spend is on physical assets, with a $5M cap and a requirement that at least 25% of the project be AI-specific. The two tracks are not competing — they're serving different project shapes.

2. Track A — the structure

Track A is the broad, sector-agnostic track. The shape:

• Cap: $2M per loan.
• Floor: $25K. Below that, BDC's standard small-business products are usually a better fit.
• Sectors: All. Trades, professional services, retail, hospitality, agri-food, manufacturing — any sector that wants to adopt AI for operational work.
• Eligible costs: AI software (build or licence), integration labour, cloud infrastructure, internal staff secondment, training, change management, project-management overhead, contingency.
• Rate: 2.25% with a named Canadian integrator. Without a Canadian integrator, BDC's standard rate applies (typically 6.5–7% in mid-2026).
• Term: Five-year amortisation. Up to 24 months principal postponement (interest-only).
• Advisory plan: Required. This is the document where Track A's "we want to see a real plan, not just a loan request" posture lives. Covered in our application walkthrough.

Track A is the workhorse track. About two-thirds of the LIFT files we touch run through Track A. It's the right product for trades businesses automating dispatch, clinics building intake automation, professional-services firms shipping internal AI agents, and any SME whose AI project is software-and-services with no significant equipment line.

Full Track A breakdown: our Track A page.

3. Track B — the structure

Track B is the larger, sector-restricted track. The shape:

• Cap: $5M per loan.
• Floor: Typically $250K. Below that, Track A is usually a better fit even for sectors that qualify.
• Sectors (seven): Advanced manufacturing, agri-food processing, transportation and logistics, life sciences and biomanufacturing, clean technology, natural resources processing, and industrial infrastructure. Other sectors are excluded; if you don't fit one of the seven, Track B isn't open to you.
• AI minimum: 25% of the project budget must be AI-specific spend (software, AI model development, integration labour for AI components, training data acquisition, etc.). The other 75% can be equipment — sensors, robotics, vision systems, automation hardware, vehicles, processing lines.
• Eligible costs: Everything Track A covers, plus the equipment and capital-asset portion.
• Rate: 2.25% with a named Canadian integrator. Standard rate otherwise.
• Term: Five to seven-year amortisation depending on the asset profile. Up to 24 months principal postponement.
• Advisory plan: Required, with a heavier emphasis on capital-project execution risk and integration with existing operations.

Track B is the right product when the project is fundamentally a capital project — new equipment, new production line, new fleet — and AI is the productivity driver that makes the capital project worth doing. Examples: a food processor adding a new packaging line with vision-AI defect detection; a logistics operator adding instrumented trailers with predictive-maintenance AI; a manufacturer adding robotic cells with adaptive AI control. The hardware is the biggest line; the AI is what makes the hardware return.

Full Track B breakdown: our Track B page.

4. The side-by-side

5. Worked example 1 — HVAC contractor, $5M revenue, $400K Track A

Composite scenario from real-shape work in BC and Alberta.

This is a clean Track A file. The application takes 6–8 weeks end-to-end. The economics pay back in roughly 16 months. No reason to consider Track B for this project — the sector wouldn't qualify, and the absence of equipment makes Track A the obvious fit. The HVAC-specific angle is at our LIFT-for-HVAC page.

6. Worked example 2 — Manufacturer, $12M revenue, $2.5M Track B

Composite scenario from real-shape work with a southern Ontario food processor.

This is a textbook Track B file. The application takes 10–14 weeks because the underwriting load is heavier (capital-project execution risk, vendor due diligence on the line equipment supplier, environmental and operational permitting if applicable). But the economics are strong — the AI productivity layer turns a capacity expansion that would have paid back in 4 years into one that pays back in 2.5.

7. The hybrid case — when you do both sequentially

A small but real subset of SMEs end up running both tracks sequentially. The shape:

• Year 1: Track A loan for the AI software and integration work. Get the AI capability built, tuned, and proven against the existing operation.
• Year 2 or 3: Track B loan for a capital expansion (new line, new equipment) that leverages the AI capability built in Year 1.

This pattern fits SMEs who are not yet ready to commit to a $2M+ capital project but want to start with the software side. Get the AI working on the existing line; prove the throughput and defect gains; then justify the capital expansion to the board (or to the lenders) with real numbers from Year 1.

BDC's view on this pattern, in our experience: supportive. The two loans are evaluated independently. The first loan being in good standing makes the second loan easier to approve. The Track B underwriting team can see the Track A track record and weight the project's execution risk accordingly.

One caveat: the same dollar of project cost can't be funded by both loans. If your Year 1 Track A funded $400K of integration work, that integration work is "done" and out of scope for Year 2's Track B. The Track B loan funds new costs — the line, the new sensors, the additional AI development specific to the new capacity.

8. The decision matrix

For SMEs sitting between obvious cases, this matrix is the framework we use:

9. Application differences between the two tracks

The document pack is similar but the emphasis shifts. The six-document framework from our application walkthrough applies to both tracks, with the following differences for Track B:

• AI scope memo is longer. Track B's underwriters want to see the AI component of the capital project explained separately and the 25% AI-spend calculation clearly attested. Plan 4–6 pages instead of 2–4.
• Use-of-funds is more detailed. Equipment vendors are named, quoted, and the line-item budget is finer-grained. Underwriters validate equipment quotes; vague "equipment: $1.4M" lines get bounced.
• Advisory plan covers capital-project execution risk. Vendor selection, delivery timelines, installation logistics, commissioning plan. The AI plan sits inside a broader capital-project plan.
• Sector qualification disclosure. Track B's seven-sector list is interpreted by underwriters case-by-case. The application needs to make the sector fit explicit — "this is agri-food processing because X, Y, Z" — not assumed.
• Security and collateral are heavier. Equipment becomes part of the security package. The line itself, when purchased, will sit in BDC's security position. Equipment-supplier security positions (if the vendor offers financing) need to be subordinated explicitly.

None of this is unsurmountable, but it explains why Track B applications take 10–14 weeks instead of 4–8. The work is real and the underwriting load is heavier.

10. Sector qualification for Track B — the practical interpretation

The seven Track B sectors are interpreted with some flexibility by BDC underwriters, but the lines do exist. Practical notes:

• Advanced manufacturing. Any SME that physically manufactures or processes goods. Custom fabrication, plastics, electronics assembly, food processing, beverage production — all qualify.
• Agri-food processing. Distinct from advanced manufacturing for some purposes but overlapping. Food processors and beverage producers can usually be framed under either; agri-food framing sometimes opens additional provincial stacking options.
• Transportation and logistics. Trucking, warehousing, distribution, last-mile delivery, intermodal operators. AI applied to fleet management, route optimisation, instrumented vehicles.
• Life sciences and biomanufacturing. Pharma, medical devices, bioprocessing. Niche but real.
• Clean technology. Clean energy, water and waste, environmental services with AI-driven monitoring or optimisation.
• Natural resources processing. Forestry, mining services, oil and gas services with AI in production or processing operations.
• Industrial infrastructure. Utilities, large-scale industrial automation, infrastructure operators with AI-driven asset management.

If your sector doesn't fit any of these clearly, Track B is closed and you're working within Track A's $2M cap. Some SMEs try to stretch their sector framing to fit Track B; this rarely works — underwriters pattern-match the actual operations against the sector list and reject stretched framings. Don't waste a quarter trying.

11. The Creatrixe role across both tracks

Our scope shifts slightly between tracks. On Track A, we are the named Canadian integrator: scope, build, ship. On Track B, we are usually one of several named vendors — equipment-line OEM, AI integrator (us), MES/ERP integrator, sometimes a sensor or vision-system specialist. The Advisory Plan covers our piece; the rest is the SME's broader vendor governance.

What this means practically: on Track A files, we are 60–80% of the project budget. On Track B files, we're usually 15–40% — the AI-specific portion plus integration. Both are reasonable shapes. The qualification for the 2.25% rate is "the project includes a Canadian integrator," not "the Canadian integrator is most of the spend." We have the conversation with BDC underwriters about this on Track B files regularly; it doesn't compromise the rate qualification.

12. The honest closing

For most Canadian SMEs adopting AI in 2026, Track A is the right answer. It's the broad, sector-agnostic track designed for software-and-services projects up to $2M, which is where the majority of operational AI work lives. Trades, services, professional firms, smaller manufacturers — Track A fits.

For SMEs in one of the seven priority sectors doing a real capital project with AI as the productivity driver, Track B is the right answer. The $5M cap, the equipment-eligible spend, and the 7-year amortisation are structurally fitted to capital-project execution.

For SMEs sitting between the two — under $2M, but with some equipment in the mix — Track A almost always wins on application overhead. The exception is when the equipment portion is large enough that the project genuinely needs the $5M cap, in which case the heavier application is worth it.

If you'd like a 30-second sanity check on which track fits, the LIFT calculator gives you an envelope estimate. For a deeper conversation, we run 30-minute scoping calls where we walk your specific project shape and tell you honestly which track to apply for — or whether LIFT isn't the right product for your project at all.

About this post

Creatrixe is a Burnaby, BC-based AI consultancy named as Canadian integrator on BDC LIFT files across both Track A and Track B. We work with SMEs in trades, services, manufacturing, agri-food processing, and logistics. Worked example numbers reflect real-shape composite scenarios. Program details accurate as of publication; track caps and sector definitions may shift between BDC fiscal cycles.