From course request to a bounded diagnostic brief
5 min read
Lesson objectives
- Turn an initial training request into a bounded, testable L&D intake brief and identify who can authorise each next decision.
Please sign in to use all the tools
Once you sign in, the reading tools are yours: larger or smaller text, typeface, line spacing, focus mode, highlighting and bold, and notes saved to your account and gathered in My study.
An L&D specialist often receives a short message: “We need customer service training.” That message proposes a solution; it does not specify the problem the solution should address. The first professional task is to turn it into a testable diagnostic brief. The brief answers six questions: who requested the work and who can decide; which role and task are affected; what workplace behaviour is expected; what happens now and what evidence shows it; when a decision is needed; and which constraints change the options. Asking these questions neither rejects the request nor promises a course before the cause of the gap is understood.
Distinguish three people who may or may not be the same: the sponsor who seeks and funds a result, the process owner who defines the performance standard and can change the workflow, and the learner who performs the task. A sales director might fund training while a contact-centre supervisor defines the complaint-handling steps and releases staff for practice. If the process owner is absent, excellent materials can be built for a standard the team cannot use. Record a named person or role rather than “management”; specify who can approve the diagnostic scope, performance standard, staff time, final material and release decision. An unresolved authority is an open decision to escalate before development effort is committed.
Next, state the difference between desired and observed performance in observable terms. “Improve communication” is too broad. “Verify the customer, restate the issue and check agreement before proposing a remedy on each eligible call” is behaviour that can be observed. A numerical indicator alone is not a substitute for behaviour. More complaints may come from long waits caused by staffing or a broken system rather than conversation skill. Record each metric with its denominator, period and source; then seek a work sample, short interview or observation to locate the failure. A manager's claim that “everyone needs training” is a hypothesis, not a baseline.
Use a one-page intake brief. At the top, state the next decision: conduct a bounded diagnosis before selecting an intervention, or develop an intervention because a reliable diagnosis already exists. Below it place the task and role, approximate audience, performance standard, available and missing evidence, constraints, decision date and approver. Also state what lies outside scope. In the contact-centre example, the specialist may review complaint coding and an authorised sample of calls. They cannot independently extract personal recordings or change system access. Employee and customer data access follows organisational policy and authorised decisions; it is not a detail to defer until launch.
Order the interview questions to make good use of colleagues' time. Begin with the claimed business result; ask for one specific successful and one unsuccessful task; compare them; then ask what has changed in tools, incentives and workload. Test whether the employee can perform the task under ordinary conditions. This is more useful than a general satisfaction survey. If the sponsor cannot define success, propose a provisional operational definition for the process owner to approve. Do not attribute that definition to the sponsor as an agreed fact. When sponsor and supervisor accounts conflict, record both and identify evidence that could distinguish them rather than choosing the more senior account.
There are three useful exit decisions. “Continue diagnosis” fits when a task, provisional standard and legitimate evidence route exist. “Pause for authority” fits when the standard owner, data access or realistic decision window is missing. “Route a likely non-learning cause” fits when initial evidence points to a process, tool or incentive beyond employee skill. Routing does not end the L&D role: the specialist can document a need for a job aid or coordination with operations, but should not label a system repair a training course. This early decision protects learner time and makes subsequent work reviewable.
Separate facts, assumptions and decisions in the brief. A fact might be “12 of 80 complaint records reviewed by a supervisor lacked a reason code during one specified week.” An assumption is “employees do not know the codes.” A decision is “inspect a sample of the input screen and interview two agents before selecting a response.” These numbers are hypothetical and do not establish a general quality rate for the centre. Twelve divided by 80 is 15% of that sample only; a sample selected from complex complaints cannot be generalised to all calls. The brief is good when it directs the next evidence and decision, not when it uses many words.
Lesson summary
Turn an initial training request into a bounded, testable L&D intake brief and identify who can authorise each next decision.