Rebutting each example
Dismantling the three cases one by one wins you the facts and loses you the substance. The person did not come for a project explanation; they came to tell you they feel treated differently — and you have just confirmed it.
An accusation of favouritism should first be treated as information about your criteria, not as an attack. Three steps: receive the perception without validating or denying it, make the rules for assigning work and visibility explicit, then genuinely correct whatever the review reveals — which is usually some part of the claim.
Adrien VasseurUpdated
Karim asks for a meeting and puts it plainly: «the interesting work always goes to the same people.» He cites three specific cases, two of which you had forgotten. You know there is an explanation for each, and you can feel the bad faith. You can also feel, more uncomfortably, that he is not entirely wrong: you hand urgent work to whoever replies fastest, and that has in fact concentrated on two people.
This is an accusation about your integrity, and the first reflex — to justify yourself — is also the most destructive. Perceived favouritism is rarely built on deliberate favours: it emerges from a series of rational micro-decisions taken under time pressure, which produce a pattern that is visible from outside and invisible from inside. So you have to honestly examine choices you never framed as choices, while holding a conversation in which any defence sounds like an admission.
Dismantling the three cases one by one wins you the facts and loses you the substance. The person did not come for a project explanation; they came to tell you they feel treated differently — and you have just confirmed it.
«There’s no favouritism here» closes the conversation and costs you the information. You do not get to decide how others perceive your decisions; you only get to decide whether to look at them.
Handing the next good project to the complainant to make the subject go away is spotted instantly by the whole team and confirms that allocation depends on pressure rather than criteria. Expect three similar requests within the month.
«If you’d delivered project X on time you’d have had the next one» converts a fairness conversation into a surprise appraisal. Even when true, it is the wrong moment, and it destroys any chance of future feedback reaching you.
An opening that holds: «I don’t see it that way, and I take it very seriously that you do. Give me the examples — all of them.» You concede nothing factual; you concede that the question is legitimate, which is enough to bring the temperature down.
Go back over twelve months: who got which projects, which committee presentations, which training. Most managers find a concentration they had not seen. Do this work before the second conversation, so you arrive with findings rather than an opinion.
The structural remedy is explicitness: what criteria govern visible assignments, in what order, with what rotation. Announced to the whole team, not just to the complainant. Favouritism thrives on the implicit; goodwill alone rarely dissolves it.
If there is a drift, say so simply: «You’re right about one thing: I got into the habit of giving urgent work to whoever answers fastest, and it concentrated. I’m changing the rule, here’s how.» A measured acknowledgement restores more credibility than a flawless defence.