How do workplace investigations go wrong — the common failure modes?
Investigations fail in predictable ways — pre-judging, vague allegations or mandate, delay, leading interviews, demeanour-based credibility findings, poor documentation, scope creep, and conflating fact-finding with discipline — each the mirror image of a procedural-fairness test.
The procedural-fairness tests set out in the procedural-fairness note describe what a defensible investigation looks like. This note is their mirror: the recurring ways investigations fail. Each failure maps to a fairness test it breaches.
- Pre-judging / a “closed mind.” Treating the investigation as a formality to confirm a conclusion already reached. Canadian courts test impartiality by asking whether the investigator approached the case with a “closed mind” (Southern Chiefs Organization Inc. v. Dumas, 2016 FC 837). In Elgert v. Home Hardware Stores Limited, 2011 ABCA 112, the Alberta Court of Appeal upheld significant punitive damages where the investigator had no relevant training and a relationship with the complainant’s father. (Breaches: impartiality.)
- Vague allegations or an inadequate mandate. Investigating “a feeling” — “bullying,” “toxic behaviour” — rather than specific conduct, dates, and context. If allegations are not framed as testable behaviour, the respondent cannot fairly answer and the evidence cannot be tested. (Breaches: notice, appropriate scope. See the investigation planning and scope stage.)
- Delay. Sitting on a complaint signals it is not taken seriously and lets evidence disappear and stories align. Ontario’s guidance states: “It would be reasonable to complete the investigation as soon as possible within 90 days or less unless there are compelling reasons why a longer investigation is needed (e.g. there are multiple witnesses, a key witness is unavailable due to illness, etc.).” (Breaches: thoroughness/timeliness.)
- Leading or sloppy interviews. Feeding the answer, or letting a hostile witness hijack the interview, produces unreliable evidence. (Breaches: thoroughness; see the interviews note.)
- Demeanour-based credibility findings. Deciding who is lying from body language, nervousness, or eye contact. The research is clear that “body language is not a reliable predictor of honesty/dishonesty,” and courts criticize demeanour-only findings. (Breaches: thoroughness; see the evidence-and-credibility note.)
- Poor documentation. Conducting interviews but failing to record questions, answers, and reasoning — “you can’t defend conclusions you didn’t record.” (Breaches: documentation.)
- Scope creep. Letting “one more thing” expand a focused complaint into an unbounded inquiry, blowing timelines and muddying the report; exceeding the mandate is itself a fairness breach (Shoan v. Canada (Attorney General), 2016 FC 1003). (Breaches: appropriate scope, timeliness.)
- Conflating investigation with discipline. The investigator purporting to decide the penalty, or the employer disciplining before the investigation concludes. Findings of fact and the discipline decision are separate steps (see the findings-and-report note and the progressive-discipline note); acting on instinct before fact-finding is a recognized source of bad-faith and constructive-dismissal liability.
A distinct, costly error sits slightly apart from the fairness tests: importing a criminal or heightened standard of proof. Requiring near-certainty before substantiating a serious allegation contradicts F.H. v. McDougall, 2008 SCC 53 — the standard is always the balance of probabilities (see the findings-and-report note).
For an Ontario SMB, the highest-frequency, highest-cost failure is the first one — bias, real or perceived — closely followed by delay and poor documentation. If the in-house candidate to investigate has any relationship to the parties or any stake in the outcome, that alone can justify an external investigator; the internal-vs-external decision is set out in the procedural-fairness note. This is general information, not legal advice.
Last reviewed .
Confidence: Industry consensus
Related notes
- What makes a workplace investigation defensible and procedurally fair? — A defensible workplace investigation is one that satisfies procedural fairness (natural justice): an impartial, unbiased investigator; notice to the respondent of the allegations against them; a genuine opportunity for both sides to be heard and respond; a thorough, timely, well-documented process; and an appropriately scoped inquiry — with an external investigator preferred where impartiality, seniority of the respondent, complexity, or litigation risk are in play.
- How do you conduct investigation interviews: order, rights, and technique? — Interview in a defensible order (commonly complainant, then respondent, then witnesses), put the specific allegations to the respondent so they can answer, allow a support person, ask open non-leading questions, and keep contemporaneous notes.
- How do you gather and weigh evidence — and assess credibility? — Weigh documentary and testimonial evidence for corroboration and relevance, and assess credibility against the preponderance of probabilities a practical person would accept (Faryna v. Chorny; Bradshaw v. Stenner) — not demeanour alone.
- How do you make findings and write the investigation report? — Apply ONE civil standard — the balance of probabilities (F.H. v. McDougall) — make findings of fact (and policy breach where the mandate asks), and leave discipline to the employer; never import a criminal or heightened standard for serious allegations.