Completed internal investigations are often reviewed at a difficult moment: a finding has been challenged, a decision is about to be made on the strength of it, or the matter has become litigation. The question is no longer only what the report concludes, but whether the work behind it will hold up.

A useful review does not start by re-investigating. It starts by testing whether the investigation that was done can support the findings it reached.

Five things to test

  1. The question that was actually investigated. Compare the allegations as made with the scope as investigated. Narrowed or reframed questions are common, and a finding can be sound on the question asked while leaving the real issue untouched.
  2. The sources that were collected, and those that were not. List every document, message, recording, and device referenced. Then list what a reasonable investigator would have sought. The gap between the two is often where a report is weakest.
  3. How interviews were sequenced and recorded. Who was interviewed, in what order, and how accounts were captured all affect reliability. Look for whether key witnesses were given the opportunity to respond to material evidence.
  4. How conflicting accounts were resolved. When witnesses disagree, the report should explain why one account was preferred, with reference to corroboration. "Found more credible" without a stated basis is a vulnerability.
  5. Whether each finding traces to the record. Take the findings one at a time and follow each back to its supporting evidence. Findings that rest on inference should say so.

What a review should produce

The result should be a clear account of which findings are well supported, which rest on incomplete work, and what additional steps, if any, would materially change the analysis. That lets counsel and leadership decide whether to rely on the report, supplement it, or commission independent work.

An independent review is not a verdict on the original investigator, and it does not determine legal liability. Its purpose is narrower and more useful: to establish how much weight the existing record can bear.

This analysis is general and methodological. It is not legal advice and does not draw on any client matter.