Leadership influence mistakes worth avoiding early

People infer rules from what leaders do, and the inference is always broader than the action. Make one exception, and the team concludes exceptions are available.

Reverse one decision quietly, and the team concludes the record is unreliable. Each mistake below is paired with the general rule a reasonable observer would take from it, because that inference is the actual cost, not the incident.

What to take away

Mistakes about reasoning

1. Announcing a decision without its reasoning. The conclusion: the reason cannot be said out loud.

There is a second and larger cost. Without the reasoning, nobody can tell which circumstances the decision depended on, so nobody can tell when it is legitimate to reopen it. The result is that it gets reopened on any change at all: a new hire, a shifted date, a bad week.

Decisions that carry their reasoning close themselves. Decisions that do not are re-litigated indefinitely. The re-litigation is not disloyalty, it is the only available way to find out whether the reason still holds.

2. Asking for input on a decision already made. The conclusion: consultation here is theater.

The consultation would have been fine as an announcement. Dressed as a question, it teaches people that their input is decorative, and the cost lands next time: when you genuinely need to know something and the responses come back thin and safe.

3. Treating agreement as agreement. The conclusion: nothing, out loud. They simply stop telling you things.

A room agreeing with the person who runs their performance review is not evidence. If you want to know what people think, the disagreement has to be cheaper for them than the silence, which usually means asking individually, asking specifically, and going last.

Mistakes about consistency

4. Giving different answers to the same question.

This one propagates fast because people compare notes. Once the team believes outcomes are determined by routing rather than by merit, every request becomes a shopping exercise and your ability to hold a standard is gone.

5. Rewarding the visible fix over the invisible prevention. The conclusion: let it break, then fix it loudly.

Weekend rescues are legible and prevention is not. If the only work that gets named is the dramatic kind, you have priced the incentives, and the people quietly stopping problems will either start letting them happen or leave.

6. Protecting a poor performer quietly. The conclusion: the standard is optional.

The intention is usually kindness, but it is also what the formal machinery exists to avoid. A performance management cycle is slow and uncomfortable precisely so the standard is applied in the open.

The quiet version moves work to whoever is competent and available. Those people watch the arrangement continue, rarely complain, and adjust their effort to match what is evidently required.

7. Reversing a decision without saying it is a reversal. The conclusion: the record cannot be trusted.

Changing your mind is fine and often correct. Presenting the new position as though it had always been the position is what costs, because it means every current statement might silently be replaced later. Say it changed, say what changed it.

Mistakes about what you teach

8. Passing your own pressure downward unmodified. The conclusion: bad news should reach you as late as possible.

If the response to a problem is visible stress, the team's rational move is to solve things quietly and tell you once they are handled, which is precisely when you can no longer help. What you lose is early information, and you lose it in the cases where it matters most.

9. Making an exception without labeling it. The conclusion: this is the new rule.

Exceptions are legitimate and sometimes necessary. Unmarked, they become precedent, which is how courts work on purpose and how teams work by accident, and the next person to be refused the same thing has a fair complaint that you cannot answer. Say it is an exception, say what made it one, and accept that you will be asked again.

10. Explaining a decision only to the people who objected. The conclusion: objecting is how you get information.

Reasoning given privately to the loudest voices, and never to the rest, creates a tier of people who know why things happen and a tier who do not. The second group learns the route into the first.

The general rule

Assume every action is being read as a policy, because it is. Before doing something one-off, ask what a sensible person watching it would take to be the new rule, then either be willing to live with that rule, or say out loud why this case is different.

The mechanics behind these are in leadership influence, and for guidance on how to say the reasoning so it survives being forwarded to others, see executive communication.

Both second and third mistakes fail disclosure test, since neither survives being described to the room. Test is in positive influence. Reversing a delegated decision without saying so is the take-back problem, which is also fully explained in control and influence, where rest of the delegation mechanics live.

Common questions

Some reasoning genuinely cannot be shared. What then?

Say that there is a reason, that you are not able to give it, and that you know what withholding it costs. People can work with a stated limit. What they cannot work with is a decision that appears to have no reason at all, because the only way to test whether it still applies is to keep reopening it.

Is asking for input on something I have mostly decided always theater?

Not if you say where the room is. "The direction is fixed, the sequencing is genuinely open" is honest and useful. The failure is the open question that was never open, and it is detected quickly, after which real questions get thin answers too.

Number six sounds like it is telling me to be harsher.

It is telling you to be explicit. Protecting someone can be right; doing it silently is what sets the standard as optional, because the people absorbing the extra work can see the arrangement and not the reasoning. Name what is happening, to the people it lands on, and put a date on it.

How do I reverse a decision without looking indecisive?

Say what changed. A reversal attached to a new fact is a functioning process. A reversal attached to nothing looks like mood, and a reversal presented as though it had always been the position is the one that costs, because it makes every current statement provisional.

What if I have made most of these?

Most leaders have. The inference the team draws is updated by behavior rather than by acknowledgment, so pick the one that is currently doing the most damage, usually the third or the eighth, and change that one habit for a quarter.