Prepare management interviews that lead to useful ESG follow-up
A management discussion should leave the review team with a clearer account of how a process works, what the terms mean and which records can test that account. Questions that suggest the desired answer make that harder. A confident explanation is still an attributed statement until the relevant evidence has been examined.
This guide helps teams prepare management discussions within an agreed ESG review. The question sequence and fictional example are practical suggestions, not an OECD or IFC interview protocol. They do not establish compliance, provide assurance or replace specialist assessment.
Choose the question and the people who can answer it
Before the meeting, write one review question in plain language. For example: “What supports the statement that new suppliers are reviewed before the first purchase order?” Identify the entity, period and supplier categories covered. Use the first-review diligence checklist to settle the wider scope before preparing detailed questions.
Invite people who can explain the process and the relevant records. A policy owner may know the intended steps; someone operating the process may know how exceptions are handled. Ask each participant which parts they know firsthand and which rely on information from others. Share the topics and relevant document references in advance without supplying preferred answers.
Agree who will take notes, how statements will be attributed and who may receive the notes. Keep the discussion within authorised access arrangements. Use coded examples and appropriately redacted records where possible; avoid collecting unnecessary personal information.
Move from an open account to precise follow-up
Prepare prompts in this order, then follow the answer rather than reading every question mechanically:
Understand the process. “Please walk us through how a new supplier moves from an initial request to its first order.” Let the participant explain before naming the control you expect to find.
Clarify the terms. “When you say a review is complete, what has happened? What does ‘before use’ mean here?” Separate an approval, a system-entry date and a purchase-order date.
Locate the records. “Using one coded case, what records, if any, show each step and its date? Where are they kept?” Identify references for later examination rather than treating a verbal walkthrough as corroboration.
Explore variation. “What circumstances, if any, change that sequence? How are those cases handled and recorded?” Leave room for no exceptions, an exception or an unknown answer.
Test the boundary. “Which sites, supplier categories and periods does this explanation cover? What falls outside it?” Ask whether the selected example is typical and what would help test that description.
Agree the next check. “What should we examine to resolve the remaining question? Who can provide it, and when?” Confirm an achievable response date with the responsible person.
Ask one question at a time. Replace “You always finish the checks first, don't you?” with a question about the actual sequence. If the answer is general, ask for a specific example. If the participant does not know, identify who may know rather than pressing them to guess.
Clarify a reassuring answer with a fictional example
All people, statements and figures below are fictional. No supporting records have been examined in this example.
A manager says, “All suppliers are checked before we use them. We completed 40 reviews in January to June.” The interviewer asks what the count represents. The manager explains that it covers 40 review events: 25 new-supplier reviews and 15 renewals. It is not presented as a count of 40 distinct suppliers.
The interviewer then asks, “For the 25 new-supplier reviews, how does the completion date compare with the first purchase-order date?” The manager reports 21 completed before the first order, three after and one whose timing is unknown.
The reported totals reconcile: 40 = 25 + 15, and 25 = 21 + 3 + 1. That arithmetic checks the consistency of the account. It does not verify the reviews, their timing or the underlying practice.
Useful next questions are:
“Does the January-to-June total include every review completed in that period? How does it relate to all suppliers first used in that period, including any without a completed review?”
“What event does each completion date represent: the review decision, its approval or entry into the system?”
“What explains the sequence in the three cases reported as after the first order? Which records could support that explanation?”
“What is missing for the case with unknown timing, and how might it be established?”
“How are renewals triggered and timed? Should they be examined separately from first-use reviews?”
The reported 21 cases do not establish that every new supplier was checked before use. The three reported later cases need their meaning and context examined; the unknown remains unknown. Do not turn these uncorroborated counts into a compliance rate or silently remove the unknown case from the denominator.
Compare accounts without coaching a resolution
If a supplied record appears inconsistent with an answer, identify the exact reference and ask for an explanation: “This row labels the review date as 12 June and the first-order date as 9 June. How should we understand those fields?” Do not offer “Was that just a late system update?” as the expected answer.
Ask what evidence would distinguish the possible explanations. Keep a disagreement or unresolved definition visible. If someone corrects an earlier statement, note the correction and its basis rather than rewriting the original account as though the uncertainty never existed.
Use the interview within an appropriate review method
IFC's ESMS Implementation Handbook describes interviews alongside observation, measurement or testing, and document review as monitoring methods. See the monitoring discussion, printed pages 59–60. Our practical implication is to plan the evidence check after an interview instead of expecting the conversation alone to establish performance.
The OECD guidance says assessment methods should fit the risk. It also notes that sensitive worker interviews may need trusted third parties because workers may be uncomfortable responding honestly around management. See Question 25, printed pages 66–67. This management-discussion guide is not a worker-interview or grievance-investigation procedure. Do not use a manager's account as a substitute for affected people's perspectives.
End with a specific evidence request
Read back the important definitions, the participant's account and the unanswered questions. Invite corrections. Ask for the precise record references, responsible owner and agreed response date needed for each next check. In the fictional example, that could include coded case references, the meanings of the date fields and records supporting the reported sequence, plus a reconciliation to the full first-use population.
Record agreement about what was said separately from whether it is supported. An interviewee confirming the notes improves their accuracy as a record of the discussion; it does not independently verify the account. Leave evidence-dependent questions open until the agreed review has addressed them.
Nitium's ESG due diligence support covers agreed record review, findings, evidence gaps, follow-up questions and next steps. Discuss the scope of your review. Legal opinions and independent assurance are outside this advisory scope.
Primary references checked 13 October 2026. Adapt the preparation to the decision, operating context and agreed review limits.
Comments