Dealing With a Difficult Employee: 5 Steps Before a Warning
Difficult or just different? Five steps before any warning: state the facts, name the type, check the context, talk it through, agree a dated plan.
Key takeaways
- Write down observable facts before you judge; what looks like a performance problem may be a difference in working style.
- Each kind of difficulty has its own fix: adapt your style, remove the hidden cause, or write a performance plan.
- In the conversation, talk about the behavior and its effect, not the person; feedback that points at the self is more likely to backfire.
- A plan that works names a specific behavior, a measure, a deadline and a weekly check-in, not just 'do better'.
- Document every conversation after it happens, and check local labour law and HR before any warning or dismissal.
On the way to work you tell yourself: "I'll talk to them today." Then the meeting starts. They answer every suggestion with a sarcastic line, or arrive late and ask for a recap, or agree with you in the room and do the opposite after you leave. The notes pile up in your head until you reach the label that makes it easy: "the difficult employee."
But that label closes the door before you have understood anything. "Difficult" is not a fixed trait in a person. It describes a relationship between them, you, and the conditions they work in. Treat every kind of difficulty with the same remedy and you will either be unfair to someone who is merely different from you, or go easy on someone who is draining the team.
This article gives you five steps that come before any written warning: state the facts, name the type of difficulty, understand the context, have the conversation, and agree a plan with a deadline. At the end you will see when to stop trying, and how to document.
Step one: difficult, or just different from you?
Start by doubting your own verdict. Not because you are necessarily wrong, but because the mind rushes to read people. In 1977 the psychologist Lee Ross named this the "fundamental attribution error": we explain other people's behavior by their character and underweight the situation they are standing in.
A well-known demonstration is a quiz game where roles were assigned by chance. The "questioner" wrote hard questions from their own knowledge, and the "contestant" tried to answer. Even so, contestants and observers rated the questioners as more knowledgeable, though the role alone had given them the advantage. Does that sound like deciding someone is "irresponsible" because they were late, when you cannot see what weighed on their week?
The tendency weakens when the person judging has the time and attention to consider the situation. So write before you judge, and use the camera test: phrase every complaint as something a camera could record.
- "Unreliable" is a label, not a fact. "Arrived late to the Sunday meeting three weeks in a row without apologizing" is a fact.
- "Aggressive" is a label. "Interrupted a colleague three times in the last meeting and raised their voice" is a fact.
After this exercise one of two things happens. Either your list is almost empty, which means the irritation lies in their style rather than in what they do, and you need understanding, not discipline. Or you find countable events, and you move on knowing what you are going to say.
Step two: three kinds of "difficult," three different fixes
Mixing up the types is what ruins the remedy. This is a practical sorting I suggest here, not an official scientific classification.
Type one: a difference in temperament
They work at a different rhythm from yours: they want many details before deciding while you like speed, or they prefer writing while you prefer a call. They break no rule and miss no result, but they get on your nerves. The fix is to adapt how you communicate with them. The title of chapter two of my Arabic book The Blind Manager, "The Four Personality Colors: A User Guide for Every Employee" (chapter titles here are translated from the Arabic), sums up the idea: every employee comes with a different user guide.
Type two: a need nobody noticed
An employee who was energetic and now criticizes everything, or one who has withdrawn from discussions. In their 2000 paper on self-determination theory, Ryan and Deci write that people can be "proactive and engaged" or "passive and alienated" largely depending on the social conditions in which they develop and function. They postulate three innate psychological needs: competence, autonomy and relatedness. When these are met, motivation and well-being rise; when they are thwarted, both fall.
That does not mean every complainer is frustrated, but it gives you hypotheses to test. Perhaps the person who objects to every decision has not been consulted in months. Perhaps the one who withdrew does not feel their effort is seen. The fix is to remove the cause, not punish the symptom. Chapter four of the book, "Beneath the Surface: Why People Do What They Do," asks the same question: what stands behind what we see?
Type three: a performance or conduct problem
A real gap between what is required and what happens, or conduct that breaks the norms of respect inside the team, after you have confirmed that expectations are clear and resources are there. The fix is a direct conversation and a written improvement plan, which steps four and five cover.
The types can overlap in one person, so rule out the first two before you decide it is a performance problem. One exception: conduct that harms others, such as bullying, harassment or threats, does not wait for this sequence. Refer it to HR immediately under your organization's policy.
Step three: before the confrontation, check the context
Before you ask them, ask yourself:
- Did I say clearly what is required, by what standard, and by when?
- Do they have the tools, information and time they need?
- Has anything changed recently: their tasks, a colleague who left, a decision that affected their work?
- When did I last really listen to their view on a decision that concerns them?
- Have I let similar behavior slide with others, or even rewarded it?
If two or more answers point to a gap on your side or a recent change in their circumstances, part of the problem lies outside the person, and fixing it is faster and cheaper than any confrontation. Then prepare two hypotheses about the cause, no more, such as "they seem to have lost the sense that their opinion counts" or "maybe the load is more than they can carry." Do not present them as fact. Treat them as questions to test during the conversation.
Step four: the conversation, and how to enter the danger zone without exploding
This is the kind of conversation Kerry Patterson and his co-authors describe in Crucial Conversations: stakes are high, opinions differ and emotions run strong. The training page for the book at Crucial Learning lists skills including focusing on what you really want, restoring safety when the other person gets defensive, speaking honestly and respectfully, and listening to understand. The Blind Manager gives the subject a full chapter: chapter eight, "Dancing in the Danger Zone: How to Manage Difficult Conversations."
Here is a suggested order for the first session, in a private place with thirty minutes to spare:
- Open with a shared purpose: "I want you to succeed with us, and something is slowing us down that I'd like us to solve together."
- State the event, not the label: "For the last three weeks you have arrived ten minutes after the meeting started."
- State the effect: "So we re-explained twice and the decision slipped."
- Ask, then listen: "How do you see it?" Summarize what you heard before you reply, and acknowledge their feelings without agreeing with the behavior: "I understand the pressure was high, and I still need us to agree on a fixed time."
- Agree the next step and set the follow-up date before the session ends.
Keep to events and effects rather than character; there is research behind this. In 1996 Kluger and DeNisi ran a meta-analysis of 607 effect sizes drawn from 23,663 observations. Feedback improved performance on average, yet in more than a third of cases it lowered performance, and the more attention shifted from the task to the self, the weaker the effect. That suggests a sentence that points at the person ("you are careless") is more likely to stall improvement than one that points at the task ("the report arrived after the deadline").
Step five: a clear plan with a behavior, a date and a follow-up
The worst improvement plan is "do better." In their review of 35 years of goal-setting research, Locke and Latham (2002) report that specific, difficult goals consistently beat urging people to "do their best," because that phrase has no external yardstick and everyone interprets it their own way. They add that a goal paired with feedback on progress works better than the goal alone, and that performance levels off or falls once the limits of ability are reached or commitment to a very difficult goal lapses. So make it hard enough, but achievable, and something they accept.
In an SHRM Middle East article on documenting discipline (Hastings, 2010), one practitioner is quoted saying each stage should spell out the improvement needed, how it will be measured, a timeframe, and the consequences of not improving. That, together with the goal-setting research, gives you four elements for your plan:
- The behavior: a sentence describing something you could watch happen. For example: "Sends the status report every Thursday before noon."
- The measure: "Six consecutive reports on time."
- The deadline and follow-up: six weeks, say, with a short weekly check-in.
- Support and consequence: what you will provide (time, training, removing an obstacle), and the next formal step if the goal is not met.
Involve the employee in the how. Ryan and Deci's paper notes that threats, deadlines and imposed goals weaken intrinsic motivation, while choice and acknowledging feelings strengthen it. That does not mean giving up the standard or the date. It means you fix them, then let the employee propose how to get there.
When do you stop trying? Documentation, the warning and the decision
Waiting with no end date has a price, and the team pays it. In 2006 Felps, Mitchell and Byington reviewed about two dozen studies (the count comes from a University of Washington release) and described a negative team member as someone who persistently withholds effort, shows negative emotion, or breaks important interpersonal norms. Their review showed that teammates perceive inequity and trust less, then react defensively with outbursts or withdrawal, which affects cooperation and creativity. The university's release adds that negative behavior outweighs positive, so one or two good colleagues cannot cancel out one negative one.
It is a review, not a precise measure of the size of the effect, so do not trust a circulating figure about a "percentage drop in performance" until you have checked its source. The practical meaning is simpler: when you postpone the decision for months, the cost lands on the committed colleagues around them.
Meanwhile, document every step. The SHRM article lists practices that include:
- Record that the conversation took place: where, the date and time, who was present, what each side said and the outcome.
- Write it up after the conversation, not before, and never build a file after you have already decided to dismiss, which defeats the purpose of progressive discipline: giving the person a real chance to improve.
- Ask the employee to sign an acknowledgment of the warning and the policy. If they refuse, send a follow-up message confirming what was discussed and that they declined to sign.
- Involve HR at every stage; it offers guidance and checks that policy is applied consistently.
At the end of the plan there are three outcomes. If performance improved, close the plan with specific recognition. If it improved partly, adjust the support or extend once with a new date, without letting the plan turn into permanent surveillance. If it did not improve despite support, move to the next formal step under company policy, together with HR. And if the real issue is a mismatch between the person and the role rather than negligence, moving them to a better-fitting role may be a fairer decision than a warning.
The bottom line: diagnose before you punish
"The difficult employee" is usually not one type but three: a colleague different from you, a person who was never heard, and an employee who truly fell short. What separates them is what makes a warning fair or unfair, and what makes your team trust you or fear you. These five steps are not about avoiding a warning at any cost. They get you to the decision knowing you saw the whole picture and gave the person a real chance.
If you want to go deeper into personality styles, hidden motives and the art of difficult conversations, they are the subject of whole chapters in The Blind Manager. You can get to know the book on its page, where an abridged Arabic audio version is free to listen to.
Frequently asked questions
How do I tell a difficult employee from one who is simply different from me?
Write every complaint as something a camera could record: a dated event, not a label like 'unreliable'. If you are left with specific, countable events, it is a behavior or performance issue. If all you have is irritation with their style, it is probably a difference in working style that calls for adapting how you communicate.
What if the employee denies what happened?
Do not argue about intent. Lay out the dated facts, hear their version, agree on what you will both observe over the next few weeks, and write down what was said. If the denial continues while the problem does too, take it to HR.
When do I move from talking to a written warning?
After a documented conversation and a written improvement plan with a deadline and real support have not worked, and in line with your company policy and local labour law. Harmful conduct such as bullying or harassment goes to HR immediately.
Can one difficult person really hurt the whole team?
A 2006 review by Felps, Mitchell and Byington (about two dozen studies, according to a University of Washington release) suggests one persistently negative member can set off a downward spiral and that negative behavior outweighs positive. It is a review, not a precise measure, so do not rely on any single percentage.
Sources
- Fundamental Attribution Error in Psychology (on Ross, 1977 and the quiz-game study)
- Self-Determination Theory and the Facilitation of Intrinsic Motivation, Social Development, and Well-Being
- Crucial Conversations Training (course overview and skills)
- The Effects of Feedback Interventions on Performance: A Historical Review, a Meta-Analysis, and a Preliminary Feedback Intervention Theory (abstract)
- Building a Practically Useful Theory of Goal Setting and Task Motivation: A 35-Year Odyssey
- Discipline: The Fine Art of Documentation
- How, when and why bad apples spoil the barrel: negative group members and dysfunctional groups
- Rotten to the core: How workplace 'bad apples' spoil barrels of good employees
- Saudi Labor Law, Article 66: Disciplinary Penalties
- Saudi Labor Law: Disciplinary Rules (Article 71)
- Disciplinary rules and procedures
- Labour Law and its Amendments No. 8 of the Year 1996 (English text)
- Crucial Conversations: Tools for Talking When Stakes Are High (authors and first edition, 2002)
- Jordan's New Labour Law Amendments Impact the Employment Relationship (Law No. 14 of 2019)
