Cold Call Calibration Session Guide

Purpose: make sure a "3" on the scorecard from one manager means the same thing as a "3" from another. Without calibration, coaching quality becomes a function of which manager a rep happens to report to.

Cadence

• Monthly for the first quarter after rolling out the scorecard/rubric
• Quarterly after that, or any time a new manager joins

Attendees

All managers/team leads who score calls, plus the enablement or RevOps owner facilitating. Cap at 6–8 people for real discussion — split into two sessions if the group is larger.

Prep (before the session)

Each manager brings 2 pre-selected calls:

1. One they scored strong (80+)
2. One they scored borderline (roughly 55–70, near the pass/fail line)

Scores are submitted to the facilitator in advance but not shared with the group until the reveal step.

---

Agenda (60 minutes)

0:00–5:00 — Frame the session

Facilitator: "We're not here to agree on this rep — we're here to agree on what 'good' sounds like, so every rep gets a fair, consistent bar no matter who's coaching them."

5:00–20:00 — Blind scoring, Call 1 (borderline)

Play the call. Every manager silently scores it against the Call Monitoring Scorecard, in isolation (written or in a shared form), before anyone says a number out loud.

20:00–30:00 — Reveal & discuss

Reveal all scores simultaneously. Where scores diverge by more than 1 point on any dimension, that manager explains their reasoning first — most-lenient scorer goes first, strictest goes last.

30:00–35:00 — Lock the anchor

Facilitator records the group's agreed anchor score plus the specific line or moment that justifies it. This becomes a saved reference clip.

35:00–50:00 — Repeat with Call 2 (strong call)

Usually faster — alignment builds from the first round.

50:00–58:00 — Capture rubric ambiguity

Any scorecard wording that caused disagreement is real signal the rubric itself is ambiguous, not just the managers. Write it down.

58:00–60:00 — Assign owners

Who updates the scorecard wording, who adds these 2 calls to the reference library, and the date of the next calibration.

---

Disagreement resolution protocol

1. State your score and the one specific moment that drove it — no vague "it just felt weak."
2. Listen to the other manager's moment before defending your own.
3. If still split after discussion, the facilitator makes the call for this instance — never resolve by vote — but flags the scorecard line for a wording review.
4. Never resolve a disagreement by averaging the scores. Averaging hides the real disagreement instead of fixing it.

Output: the Calibration Reference Library

A running library of 2–3 saved example calls per scorecard dimension, tagged with the agreed score and the specific reasoning behind it. Use this to onboard new managers into scoring without re-litigating every disagreement from scratch.