A methodology that measures practice, not opinion
This is the load-bearing difference. The benchmark is built on behavioural data, actual diagnostic answer patterns, not a sentiment survey of what people believe about their own quality practice.
Sentiment
Ask practitioners what they think about their quality practice. Useful, but it measures opinion about practice, and treats any one national market as a single cut among dozens.
Behaviour
Generated from real diagnostic answer patterns that surface contradictions and urgent findings, reading what a team's answers reveal rather than a single self-rating. Canadian by design, with provincial and sector texture no global survey provides.
How a reading is produced
Diagnostic
An organization completes the CareLogic diagnostic, with adaptive questions across People, Processes, and Products.
Scoring
Responses are scored against the CareLogic model, consistent across organizations and comparable year over year.
Aggregation
Scores are aggregated into the cohort at province and sector level, never as individual organizations, always above the k-anonymity threshold.
Live, reviewed yearly
The benchmark updates the moment a diagnostic clears the k-anonymity threshold. It's live, not a yearly snapshot. Once a year, the advisory board reviews the methodology. The annual report names the adjustments it makes, along with the topics the board raised that we chose not to act on.
How this differs from existing QE benchmarks
We name them because credible readers already know them. The World Quality Report, ISTQB surveys, and DORA reports are credible global instruments, and each treats Canada as one market among many. This benchmark is Canadian-first, with provincial granularity, a behavioural basis, and a placement each respondent can see.
The dashboard is live, but once a year we also publish a State of Canadian Quality Engineering report drawn from this same data: the year's key trends and movements, with methodology disclosure and named advisory-board perspectives. The dashboard tells you where you stand today; the annual report tells the story the data reveals across the year.
Sectors follow NAICS, Statistics Canada's North American Industry Classification System (SCIAN in French), the Canadian standard for industry classification. The Technology grouping aligns with Statistics Canada's ICT sector definition. Groupings are broad by design: each bucket reaches the k-anonymity threshold faster, and as the cohort grows, buckets subdivide into finer NAICS sub-sectors, with no re-survey, because the detailed code is captured from day one.
How we keep the comparison fair across sample sizes
A score from a few organizations is less certain than the same score from many. So we don't present them as equally solid. Each cohort's score is blended with the national picture, weighted by how many organizations stand behind it: a cohort with few participants leans on the national number until it has enough of its own, while a large cohort stands on its data alone.
Under 10 organizations, no score is shown at all. Between 10 and 29, the score is shown as provisional (hatched). At 30 or more it renders solid. This is why provinces look similar early on, then separate as the data grows. That restraint is the method working, not a gap.