Before You Approve Training

• A training request arrives in response to a missed target or a manager's concern.
• The proposed solution — a course, a refresher, a programme — has arrived before the problem has been examined.
• Several causes could plausibly explain the gap: knowledge, skill, judgment, access, workflow, expectations or reinforcement.
• You are about to commit budget, employee time and manager support to a learning response.

Training is a reasonable response when a meaningful gap in knowledge, skill or judgment contributes to the problem, and people have a credible opportunity to use what they learn. That requires more evidence than a missed target or a manager's request. A request for training is a starting hypothesis.

1. 1

   Start with two questions

   Would better execution of this task materially improve the result? What would still prevent that execution if employees had the required capability?

2. 2

   Define the performance before choosing the solution

   “Improve customer service” is too broad to guide an investment. “Resolve an eligible refund during the first customer contact, using the correct policy and approval route” gives you something to examine. Write down the expected performance, the current gap and the consequence, then look for evidence from actual work: case records, errors, observation, system access and employee explanations.

3. 3

   Separate a capability gap from a barrier to using capability

   Ask employees to demonstrate the task under realistic conditions, and compare successful and unsuccessful cases carefully. Someone who handles a practice case correctly may be blocked at work by permissions or an unavailable approver — and experienced employees' results do not automatically prove that everyone else needs training.

4. 4

   Make the approval decision explicit

   Use the decision model below: proceed, fix, defer or stop — with the reason recorded and an owner for every dependency.

5. 5

   Keep the investigation proportionate

   A straightforward task with an observed skill gap may need a brief conversation and demonstration; a costly enterprise programme warrants more scrutiny. Collect the smallest amount of evidence that could change the decision.

6. 6

   Give the sponsor a decision they can act on

   An approval brief states the performance to improve, the supporting evidence, the barriers that must be addressed and the people accountable for them — including the full time commitment and how you will judge whether the response helped.

• A documented approval decision — proceed, fix, defer or stop — with the reason recorded.
• A sponsor-ready brief: the performance to improve, the supporting evidence, the barriers to address, the people accountable, the full time commitment, and how you will judge whether the response helped.
• A completed decision worksheet shared with the operational owner.

• Illustrative example — not a client case or a report of measured results.
• A service leader requests a refresher course because refund cases keep escalating. In this hypothetical review, observation reveals two patterns: some employees cannot distinguish standard refunds from exceptions; others make the correct decision but lack permission to complete the refund. A broad refresher would address only part of the problem.
• The proposed response has two owners: operations clarifies the approval route and resolves the access barrier; L&D develops short case practice for employees who need help identifying exceptions, supported by a maintained decision guide.
• Before rollout, the team records the current escalation pattern and agrees how to check both decision accuracy and avoidable escalations — and monitors inappropriate refunds, so that reducing escalations does not create a different problem.
• If results improve after both changes, the team describes the combined intervention and examines other changes during the period; it does not attribute the entire improvement to the learning component.

• Keep the investigation proportionate: the commitment includes employee time, manager support, maintenance and implementation dependencies.
• The strongest objection to diagnosis is delay. Address it by collecting the smallest amount of evidence that could change the decision — if nobody can say how another interview or survey would affect the recommendation, reconsider whether it is necessary.
• Tell the sponsor what would cause you to revise the recommendation. For example, if employees perform accurately once access is restored, the case for broad retraining becomes weaker.

• Neither diagnostic question, by itself, proves that training is needed.
• A person who can explain a policy may still struggle to apply it to an ambiguous case — test under realistic conditions.
• Experienced employees' results do not automatically prove that everyone else needs training.
• Required training needs a separate check: this guide does not determine whether an obligation applies or authorise its cancellation. Confirm the requirement with the accountable owner, then diagnose what else is needed for reliable performance.
• Do not attribute the entire improvement to the learning component when the response combined several changes.