What a strong SR&ED narrative actually looks like
The hardest part of a T661 is describing the uncertainty and the advancement well. These fictional examples put a weak version next to a strong one for the same project, so you can see what a reviewer is looking for.
Sub-second reconciliation at 10M+ transactions
A fictional SaaS team's reconciliation-engine claim — how the uncertainty and advancement read weak versus strong.
Weak vs strong AI & machine learningDefect detection with scarce labelled data
A fictional computer-vision team's claim — separating the eligible research from the routine model plumbing.
Weak vs strong ManufacturingHolding tolerance on a recycled-content polymer
A fictional manufacturer's process-development claim — where the eligible work is the process, not the product.
Weak vs strong BiotechnologyA stable assay for a degradation-prone biomarker
A fictional biotech's assay-development claim — separating experimental development from routine testing.
Weak vs strong Food scienceRemoving an emulsifier without losing shelf stability
A fictional food company's reformulation claim — the line between recipe development and eligible SR&ED.
Weak vs strong Electronics & hardwareThermal stability in a compact RF sensor
A fictional hardware team's claim — the design work that pushed a physical limit, not the reference-design parts.
Weak vs strongThis is a fictional educational example. It is not a customer claim and is not tax advice. It illustrates how to structure and reason about a SR&ED narrative — not text to copy into a claim.