Conversation Intelligence Coaching: A Manager’s Weekly Playbook
Use conversation intelligence for coaching by finding a relevant call, verifying the evidence, choosing one observable behavior, practicing a better response, and reviewing a later conversation. The manager’s job is to turn a finding into a useful action. A dashboard or automated score does not complete that loop.
This guide provides an adaptable weekly cadence, a coaching-session worksheet, and a worked example for sales managers. For program deployment and ownership, use the RevOps rollout checklist. For the broader first-quarter learning process, see the manager’s first 90 days with AI coaching data.
Choose a behavior, not a scoreboard position
Start with a relevant skill and comparable conversations. A low aggregate score may reflect an incomplete transcript, an inappropriate rubric, or a different call purpose. Inspect those factors before treating a ranking as a coaching priority.
Good targets describe something a reviewer can observe: a clarifying objection question, a buyer-confirmed consequence, or an agreed next step with an owner and timing. “Sound more confident” is harder to evaluate consistently and may not address the buyer’s concern.
Use question counts and talk ratios as diagnostic clues. A short, direct answer can be appropriate on one call and insufficient on another. A keyword mention is not the same as understanding the buyer’s situation.
An adaptable weekly conversation-coaching cadence
| Step | Manager action | Useful output |
|---|---|---|
| Team scan | Inspect comparable calls and repeated gaps. Check data quality and rubric consistency. | A small number of behaviors to investigate. |
| Evidence review | Verify the recording and context behind one finding. | A defensible example, with timestamp where available. |
| Coaching session | Invite the rep’s view, explain the behavior, and practice an alternative. | One agreed practice commitment. |
| Later-call check | Review a different conversation using the same criterion. | Observed change, unresolved gap, or insufficient evidence. |
| Program review | Review participation, workload, shared gaps, and rubric issues. | A decision to continue, adjust, or provide team training. |
The cadence can fit a short weekly review and a focused one-on-one, but there is no guaranteed time saving or required number of sessions. Reserve capacity for evidence review and follow-up, not only for the meeting itself.
Inspect the finding before showing it to the rep
- Confirm the call purpose. A qualification conversation may not establish every late-stage buying detail.
- Check the transcript and speakers. Ensure the cited statement was not a rep’s question misattributed to the buyer.
- Listen to surrounding context. A short clip can omit an answer that appears earlier or later.
- Review the criterion. Determine whether it assesses the seller’s behavior, the buyer’s knowledge, or deal readiness.
- Record uncertainty. Incomplete evidence is a reason to investigate, not to invent a confident assessment.
For discovery, use the evidence-based discovery scorecard. Keep a shared rubric version so managers are comparing the same standard.
A coaching-session worksheet
Copy these fields into your existing coaching notes. The worksheet is ungated and does not require a new form or system. Store customer material only through your approved process.
| Field | What to record |
|---|---|
| Conversation context | Role, stage, call purpose, and permitted recording reference. |
| Behavior | One observable action relevant to this call. |
| Evidence | Timestamp or short statement, speaker, and relevant context. |
| Rep perspective | What the rep intended and any missing information or disagreement. |
| Practice | The alternative question or response to rehearse. |
| Commitment | Where the rep will apply it and when the manager will review. |
| Follow-up result | Observed behavior, remaining gap, or insufficient evidence. |
Worked example: exploring an implementation objection
Fictional scenario: a buyer says, “We cannot disrupt the team this quarter.” The rep responds with a product feature list and moves on. The manager’s target is to clarify the concern before presenting an answer.
The manager plays the relevant context and asks, “What did you understand the buyer to mean?” Together they separate possibilities: limited admin capacity, timing, training burden, or a previous migration failure. The recording supports the objection, but does not establish which explanation is correct.
The rep practices a question such as, “Which part of a change would be most disruptive for your team?” In the next comparable call, the manager reviews whether the rep explored the issue and acknowledged the response. The criterion rewards relevant clarification, not a memorized phrase.
This exercise does not establish that the behavior will close the opportunity. It gives the team a specific, observable improvement to evaluate.
Use roleplay to practice the weak moment
Recreate the objection or discovery gap with a manager, peer, or approved AI simulation. Change the buyer’s answer on the next attempt so the rep practices listening rather than reciting. Review whether the response fits the buyer context.
Use the roleplay scenario library for examples and the roleplay software evaluation guide when choosing a platform. Practice feedback and facts about a real opportunity must remain separate.
Build a call library that teaches a skill
Label each approved clip with the behavior, call stage, buyer context, timestamp, and reason it is useful. Include a strong example and a near-miss, then explain the difference. A new rep should know what to listen for before playing the recording.
Verify sharing permissions and approved use. Review clips when the product, messaging, or sales process changes. Remove stale material from the training workflow rather than teaching an outdated promise. Use a maintained sales playbook to keep guidance consistent.
Measure behavior change without claiming causation
Track completed coaching actions, practice, review time, reviewer disagreement, and the target behavior in later comparable calls. Disclose the sample and assessment method. A changed rubric can produce a changed score even when execution is unchanged.
Inspect opportunity outcomes over an appropriate sales cycle. Compare similar stages, sources, segments, tenure, and deal sizes where possible. Include lost and stalled opportunities, not only wins. A correlation between a score and revenue may reflect buyer quality or rep experience rather than the coaching itself.
For program measurement, use the sales coaching measurement guide. Do not use an automated score as the sole basis for a consequential personnel decision.
What to demonstrate in a coaching software evaluation
Ask a manager to retrieve a relevant call, inspect a finding, discuss it with a rep, and record the next action. Test a disputed score and an incomplete transcript. Check that the normal role has the necessary access without broadening permissions unnecessarily.
Revenue.io Conversation AI supports conversation review. AI Generative Scorecards supports configured criteria, feedback, and trends for qualifying calls. Confirm licenses and settings. If the team needs guidance during a call, evaluate Moments separately.
Bring this worksheet and one representative coaching scenario to a manager-workflow demo. Compare the required plan with the manager effort and capabilities the team actually needs.
Frequently asked questions
Should I start with the lowest-scoring rep?
Not automatically. Check call comparability, evidence quality, rubric fit, and the rep’s context before prioritizing support.
How many calls should I review?
Enough varied, relevant examples to understand the behavior and check follow-through. A small sample can guide coaching but does not establish a reliable performance estimate.
Can coaching happen without AI scores?
Yes. Recording, context, a shared rubric, practice, and follow-up can support the same core process. Automation can help locate and assess evidence, subject to review.