Cómo estructurar laboratorios de auditoría de llamadas de descubrimiento de clientes que capaciten a los representantes para evaluar los puntos débiles del cliente en lugar de simplemente presentarles funciones.

A rep opens a discovery call with a prepared demo. Ten minutes in, the buyer mentions, almost in passing, that any new tool has to be approved by their security team before a pilot can start. The rep says "great, let me show you the reporting dashboard" and carries on. The constraint never gets a follow-up question, and it quietly decides the deal three weeks later.

This is an illustrative scenario, not a real customer case. It is common enough to design training around: the signal was in the call, and the rep did not evaluate it. This guide is for Heads of Sales Enablement and sales managers who want reps to identify customer needs and ask useful follow-up questions, and who want a structure for doing it. The structure is the discovery call audit lab: a group reviews real or realistic call evidence, decides what mattered, and practises the next question.

What the role must be able to do

Define the task before you design anything. For discovery, a rep must be able to:

1. Identify what the customer needs, including constraints they mention only in passing.
2. Ask follow-up questions that test what the customer said, rather than moving to the next item in the pitch.
3. Recognise when the information or authority they have is not enough, and say so.

Notice what is not on the list: talking less, or running through a question checklist. Talk time is easy to measure and tempting to use, but a quiet rep can still miss the constraint, and a talkative one can ask exactly the right question. Treat the idea that talk ratio predicts discovery quality as a hypothesis to test with your own calls, not a rule.

Diagnose the gap before you prescribe a lab

Not every feature dump is a training problem. Before building anything, compare what the role requires with what reps currently practise and how they are assessed.

Carnegie Mellon's Eberly Center describes the principle well: assessments should reveal how well people have learned what you want them to learn, while instruction makes sure they learn it, and when objectives, activities and assessments are misaligned, learning suffers (Eberly Center, alignment guidance). It is design guidance, not evidence about any particular platform or sales outcome, but it gives you a useful audit question: does what we practise match what we assess, and does either match the job?

Apply it to discovery:

• Práctica actual: if onboarding is a product walkthrough and a recorded demo, reps practise presenting. Discovery needs analysing a customer's words.
• Evaluación actual: if reps pass a product knowledge quiz, you are assessing recall. That tells you they know the features, not that they can spot a constraint.
• Cognitive demand: the task needs reps to apply a questioning approach, analyse a call for what was said and left unexplored, and evaluate which follow-up is worth asking. Those map to the Apply, Analyze and Evaluate levels of Bloom's taxonomy, which is a way of naming the demand of a task, not a prescription for a delivery format.
• Otras causas: check the non-training explanations. Does the CRM template push reps to fill fields instead of listening? Do targets reward demos booked rather than qualified opportunities? Is there time between calls to prepare? If so, a lab will not fix the underlying issue, and you should say that to the people who own those things.

Treat your current method, and the ceiling of your current assessment, as hypotheses to check with the sales managers who listen to calls every week.

Design the lab: preparation, practice, feedback

A discovery call audit lab has three parts, matched to the task.

Preparation: check the prerequisites

Reps need a shared vocabulary before they can audit a call: what counts as a need versus a stated want, what a constraint is, what your approved qualification framework asks. A short self-paced module with a brief diagnostic covers this. Be clear about what the diagnostic shows: a recall check confirms the prerequisites are in place. It does not show that anyone can do the task.

Practice: audit a call and choose the next question

The core activity is a reasoned response to a case. Give reps a call transcript or short recording, ideally anonymised and used with permission, or a scripted call written by a sales manager. For each case, ask the learner to:

• explain what the customer actually needs and what the evidence is,
• mark the points where a follow-up question was available,
• write the question they would have asked, and
• say what further information or support they would need before moving forward.

Feedback should be tied to those decisions: "you noticed the approval step, and your question could have asked who is involved and how long it usually takes", rather than a general "good listening".

Review: when live input helps and when async is enough

Use live facilitation where the judgement is ambiguous: two reasonable reps disagree on which follow-up matters, and the discussion is the learning. A manager or experienced rep who can explain why a question works earns their time here. Asynchronous review is enough for clearer cases, where a rep submits their audit and receives written feedback against agreed criteria. Most programmes mix both, which is where tools that support live and self-paced delivery in the same flow help. In AhaSlides, for instance, the interactive slide types (polls, open-ended responses, quizzes) can be run in a live session or set for people to work through in their own time, so one case can be used either way. Check any specific authoring, scoring or reporting function against a demonstration before you rely on it.

Un ejemplo práctico

Here is a short illustrative case. A sales manager would need to validate it and its answer criteria before use.

El caso: a buyer in operations says, on a discovery call, "We looked at something like this last year. Our team needs to see it work with how we already log requests, and anything that touches customer data goes through our security review first."

Evidencia revisada: the transcript extract, plus the rep's next three lines, which move straight to a feature tour.

Learner choices. The rep is asked what the customer needs and which question to ask next.

Each verdict is judged against criteria agreed with the sales manager, not against the facilitator's taste.

Feedback and retry: the rep who picked the incomplete answer sees the buyer's second sentence highlighted, rewrites the follow-up, and tries a new short case with a different hidden constraint, such as a budget cycle or a decision-maker who is not on the call. The retry matters more than the first attempt, because it shows whether the feedback changed what the rep noticed.

A rubric you can adapt

Score each criterion from 0 to 2. This is a proposal to adapt with your own sales managers. It is not a validated readiness standard.

Anclas: 0 = absent or unsupported. 1 = partial, with relevant omissions. 2 = complete and justified against agreed criteria.

Define critical errors separately, for example making a commitment the rep has no authority to make. A total score must not hide a critical failure: a rep who scores well on questions but promises a security exemption has not passed.

Assess capability and plan transfer

Back to the two questions the lab has to answer.

What evidence would show that reps can identify customer needs and ask useful follow-up questions? A new case they have not seen, scored against the rubric, is the minimum: it shows the skill outside the lab case. The stronger evidence comes from real work. A sales manager reviews a sample of the rep's recorded or shadowed calls after some time in the field and scores them against the same rubric. The CDC's training evaluation guidance separates evaluating learning from evaluating learning transfer, which it describes as how well people apply what they learned back at work, and notes that a post-course evaluation alone cannot assess either objectively (CDC, evaluación de la capacitación). It also suggests a pretest and posttest for comparing results. It is evaluation guidance, not a forecast that your numbers will improve.

Which constraints could remain after the rep's performance improves? Plenty. The buyer may not reveal a constraint however good the question is. Territory design, deal-size targets, call volume and the CRM template can all limit what a rep can do. A rep may identify a constraint correctly and the company may still lack a security answer.

For measurement, keep it simple and honest:

• Medida: need identification and question quality, using the rubric.
• Comparar: a baseline task before the lab with an unseen follow-up task after it.
• Definir: who is in the group, the denominator and the timing.
• Check for other changes: a new pricing page, a different lead source or a manager change can all move the numbers you care about. Look for them before crediting the training. Do not promise a particular improvement in win rates or revenue.

Plan the follow-up in the workplace: after two or three weeks of live calls, run a short audit of one real call per rep, discuss one ambiguous choice in a team session, and decide what to change in the template, the coaching or the lab.

Por dónde empezar

Start with the rubric above. Pick one recent call, ask two managers to score it independently, and compare their scores before you build a single case. The disagreements show you where your criteria are vague, and that is what the lab needs to settle first.

Once the design is clear, you can build the case practice as an interactive session. Browse AhaSlides templates to see how polls, open-ended responses and quizzes could carry a case, or ask for a walkthrough of a live-and-self-paced workflow that fits your lab design.