Lire cet article en français →
Wednesday morning, 8:15 a.m. The occupational health and safety coordinator at an IT services firm in Quebec City — 95 employees, three development teams, a fast-growing client support department — opens three browser tabs. He spent the previous evening searching for psychosocial risk self-diagnostic tools available online. He finds one published by a professional order, a second offered by a workplace health organization, and a third translated from English by a sectoral association. He downloads them. He distributes them to his team over the course of the week. Eighty-two people respond. The scores look good. He tells himself the work is done.
Six months later, he brings in a consultant to start over from scratch. Not because the tools were bad. Because the results had never said what he thought they were saying.
What scores don’t contain
Psychosocial risk self-diagnostic tools available online generally measure perceptions: the felt level of overload, role clarity, perceived support from the immediate supervisor, recognition. These are useful indicators. They give a subjective snapshot at a given moment in time. But Quebec’s Law 27 (the Act to modernize the occupational health and safety regime) does not ask for a subjective snapshot. It requires the identification of risks and an analysis of the organizational factors that generate them. Those are two different exercises.
At the Quebec City firm, overload scores were high in two of the development teams. What the tool could not reveal: the overload was structural, embedded in a client mandate estimation process that systematically underestimated delivery timelines. No checkbox captures that. The tool measured the effect; no one had yet named the cause. When the consultant arrived six months later, he conducted semi-structured interviews with team leads and analyzed project management data. The organizational cause emerged in under three weeks. It could have emerged from the start.
The real cost is not that of a consultant hired twice. It is the cost of the six months during which the organization believed it had acted. Two employees left during that period. A third was on partial leave. Management had a survey report sitting in a binder. That report could not be used to defend against a claim, because it identified no modifiable organizational factors and documented no corrective plan. That is not compliance. That is the appearance of compliance — which, in the event of an inspection or a dispute, can worsen the employer’s position rather than protect it.
What the tool does well, and why it deserves to stay in the process
The honest objection is this: self-diagnostic tools are not useless, and saying otherwise would be inaccurate. They create space for conversation, especially in environments where psychosocial risks have not yet been named out loud. They give employees a first language for describing what they are experiencing. They allow an OHS coordinator to prioritize: if social support collapses in one unit while holding steady everywhere else, that is a signal that even a basic analysis makes visible. In an organization that has not yet done anything, distributing a validated questionnaire is always better than distributing nothing.
The problem is not the tool. The problem is the confusion between measuring and analyzing. Measuring means collecting perceptions. Analyzing means tracing those perceptions back to management practices, work organization models, and internal policies that produce them. Organizational analysis requires an outside perspective — not because internal coordinators lack competence, but because they are inside the very system they are trying to read. That is a structural limitation, not a personal one.
The OHS coordinator at the Quebec City firm — I asked him directly during an informal conversation — said this: “I should have asked the consultant, before signing anything, what he was going to do with the data I already had.” That is the right question. A consultant who sets aside your existing data and starts from zero without explanation is billing you twice for the same work. A consultant who knows what to do with your incomplete scores, who can extract what is salvageable and identify what is missing, reduces the total cost and produces a deliverable that holds up before an auditor.
CSS Consult begins every mandate with a review of data already available — internal surveys, absence data, incident reports — before proposing any new data collection. The question that should stay with you after reading this article is: if your management had to justify today, with supporting documents, that the organizational causes of overload in your teams have been identified and corrected, what could you put on the table?
What gets measured gets prevented. What gets proven remains.