The four tests
- Technological uncertainty.
At the start, a competent professional in your field could not have known how to do it from published knowledge. This is the "why" test. Hard, expensive or late is not the same thing: difficulty, cost and schedule pressure do not count.
- Systematic investigation.
You worked through it as an experiment: a hypothesis, a test, a measurement, analysis, an adjustment, and again. This is the "how" test. Trying things until something worked, with no record of why, is weak.
- Work performed in Canada.
The people doing the experimental work did it here. Work done abroad generally does not count. The one narrow exception: your own Canadian-resident employees, for time abroad that supports the work done here, capped at 10% of your Canadian SR&ED salary.
- A real expenditure base.
There were costs to claim against: salary paid to the people doing the work, materials consumed, Canadian contractors. A pool too small to produce a meaningful credit fails this test in practice.