“They are slacking.”
“They probably do not care.”
Managers use explanations like these every day.
They may feel accurate. They may even begin with a legitimate observation.
But they are not performance findings.
They are interpretations.
“Slacking” does not identify what the employee failed to do, how frequently it occurred or what standard was missed.
“They do not care” goes further. It assigns an intention to the employee without demonstrating what actually happened.
An impression can tell a manager where to look.
It cannot tell them what happened.
A label gives the employee nothing to correct
Consider the difference between these statements:
“You have become careless.”
“Three required follow-ups remained incomplete this week despite the agreed process.”
The first describes the manager’s opinion of the person.
The second identifies an observable behaviour.
Only one gives the employee something specific to answer, understand and correct.
Vague judgments make accountability more emotional and less useful. The employee defends their character while the manager repeats their frustration.
Neither side is examining the work.
Evidence moves the conversation away from personality and towards performance.
It asks:
- What was expected?
- What actually occurred?
- How often did it occur?
- What consequence did it create?
- What conditions contributed to it?
- What must change?
These questions do not weaken accountability.
They make it defensible.
Put the concern under review before judging the person
When I suspect that an employee’s performance has changed, I first examine the relevant operational evidence available through our systems.
In our case, that often includes information from the electronic medical record and customer relationship management platforms.
These systems can reveal patterns that memory and frustration cannot reliably reconstruct.
They may show incomplete documentation, delayed follow-ups, repeated communications, missed tasks or breakdowns in the movement of information.
I then conduct a short, time-limited fact-finding review before reaching a formal conclusion.
The purpose is to test my anecdotal impression against reality.
Is the behaviour genuinely recurring?
Is it isolated to particular shifts, tasks or handoffs?
Did the employee receive the correct information?
Was ownership clear?
Is the problem beginning with this employee, or merely becoming visible through them?
That final question has changed my judgment before.
I had mistaken where the problem surfaced for where it began
One employee suddenly appeared to be losing concentration.
They began making uncharacteristic and avoidable errors. Matters that had already been communicated or resolved were repeatedly raised again.
From the outside, it looked like carelessness.
It would have been easy to conclude that the employee was no longer engaged or had stopped paying adequate attention.
Instead, I examined the operational trail and observed the pattern more closely.
What emerged was a web of fragmented communication and poorly coordinated movement of information between departments.
The employee was working at the point where those disconnected messages, changing instructions and unclear handoffs eventually collided.
What I was seeing was the outcome.
It was not necessarily the origin.
I had mistaken the point where the problem surfaced for the place where it began.
Once we looked deeper, the underlying issues began appearing and could be addressed individually.
Had I acted only on my original impression, I might have disciplined the most visible person while leaving the system producing the confusion untouched.
The employee’s performance still mattered.
But the evidence changed the diagnosis, and therefore changed the appropriate response.
Data should begin the investigation, not end it
Operational data is valuable, but numbers alone do not explain cause.
A system can show that a follow-up was missed.
It may not show whether the employee received conflicting instructions, whether responsibility shifted between departments or whether the workflow itself created unnecessary duplication.
Evidence-based management therefore requires more than opening a dashboard.
It requires combining:
- Relevant operational records
- Direct observation
- The employee’s explanation
- The sequence of communication
- The responsibilities of other teams
- The established workflow
- The frequency and consequence of the problem
Evidence is not synonymous with numbers.
It means that the conclusion is proportionate to what can actually be demonstrated.
An isolated error and a performance problem require different responses
Not every mistake should begin a disciplinary process.
People make errors. Systems fail. Instructions can be misunderstood, and unusual conditions can produce unusual outcomes.
The response should escalate according to severity, recurrence and the support already provided.
1. Inform
Describe what happened and restate the expected standard.
Give the employee an opportunity to explain anything management may not have seen.
2. Coach or retrain
If the problem involves understanding, skill or an unclear process, provide the necessary support.
Then verify that the employee can apply what was explained.
3. Issue a verbal warning
If the expectation is clear, the employee has been supported and the same problem continues, the conversation must become firmer.
The employee should understand that recurrence now represents a performance concern.
4. Proceed to formal action
A formal warning should not arrive as a surprise.
By this stage, the organization should be able to show the standard, the recurring behavior, previous interventions, operational impact and the improvement required.
5. Make a decision
If the gap remains after clarity, training, resources and a reasonable opportunity to improve, leadership must determine what it is facing.
The employee may have encountered a capability or role-fit limitation.
Alternatively, they may understand the standard but fail to apply it consistently.
Those situations may require different decisions, but neither should remain unresolved indefinitely.
Fairness and firmness belong together
Evidence does not mean giving employees endless opportunities regardless of their performance.
It does not mean explaining away every mistake as a systems problem.
Sometimes the evidence will show that the system failed the employee.
Sometimes it will show that the employee received clear instructions, adequate training and reasonable support but continued failing to meet the standard.
Both findings are useful.
Fairness requires leaders to identify the correct one.
Firmness requires them to act on it.
Accountability without evidence becomes authority acting on assumption.
Evidence without consequences becomes documentation without management.
Your team deserves more than your impression of their performance.
They deserve an accurate diagnosis.
Your organization deserves the accountability that can only follow it.
Are you addressing the person responsible for the problem, or merely the person through whom the problem became visible?