When work goes wrong, managers often respond with either a new rule or another training session. Both can be useful, but only when they address the real source of the problem. Training cannot repair a contradictory policy, and a detailed policy cannot help someone who never learned the required skill.

Separating the two begins with evidence. The goal is not to assign fault quickly. It is to understand whether people lack direction, lack capability, or face conditions that make the expected behavior difficult.

Describe the gap without diagnosing it

Write down what should have happened and what occurred instead. Use observable language. “The customer received two different answers” is more useful than “the team does not care about consistency.” Identify where in the workflow the difference appeared and what information was available at that moment.

Review more than one example. A single mistake may be an isolated lapse, while a recurring pattern suggests a system issue. Readers may also want to visit Angelos Epoxy Flooring.

Ask employees to explain how they understood the expected action. Their explanations may expose an unwritten rule, an outdated document, or a conflict between speed and accuracy. Listen before showing them the preferred answer.

Test the policy itself

Locate the current policy and check whether employees can find it during real work. Confirm that it describes the decision clearly, uses terms consistently, and accounts for common exceptions. Compare it with related procedures, customer promises, and system prompts.

Use these questions during the review:

• Is the expected action written and easy to locate?
• Do different documents give compatible instructions?
• Can the rule be followed with available time, tools, and authority?
• Are exceptions defined or routed to a clear owner?
• Does another target reward behavior that conflicts with the policy?

If experienced employees interpret the rule differently, the problem is unlikely to be solved by repeating it louder. You can also visit Designwithdeb.com for more reading.

Check knowledge and skill separately

If the policy is clear and workable, examine capability. Knowledge means the person can state what to do; skill means the person can perform it under normal conditions. A support agent might know the escalation rule but struggle to recognize the cue in a fast conversation. A supervisor may understand a feedback model but need practice using it calmly.

Ask people to work through realistic examples or demonstrate the task in a safe setting. Avoid quizzes that reward memorizing wording but do not test judgment. Notice whether the gap affects new employees, a particular role, or the whole team.

Design training around the specific behavior. Include explanation, demonstration, practice, feedback, and a job aid for later use. For a related read on this site, see How to Create a Sales Call Note Standard.

Look for conditions beyond policy and training

Sometimes both are adequate and the environment prevents success. The system may hide necessary information, staffing may not match demand, handoffs may arrive incomplete, or a metric may reward shortcuts. Treat these as operating-design problems.

Choose the smallest response that fits the evidence. Clarify a sentence if wording is the issue. Provide coached practice if skill is missing. Change a tool or workload rule if the environment blocks the expected action. Do not launch a large program simply because the original problem was frustrating.

After the change, observe the same point in the workflow. Look for improved behavior and unintended effects. If the gap remains, revisit the diagnosis rather than assuming employees resisted the solution.

Communicate the response precisely

Tell employees what the review found and which response follows. Saying “we are retraining everyone” when the policy was unclear can feel unfair and conceal the real improvement. Likewise, announcing a policy change when a few people need coached practice can confuse staff who already perform the task well. Explain the evidence, the selected action, and when the team will check whether it worked. This reinforces that diagnosis is a normal part of improving work, not a search for someone to blame.

Managers build credibility when responses match causes. By describing the gap, testing the rule, checking capability, and examining working conditions, a team can avoid unnecessary policies and generic retraining. The result is a clearer workplace where people know what is expected and have a realistic way to meet that expectation.