| name | call-scorecards |
| title | Call scorecards |
| description | Use this skill when sales call quality is inconsistent, coaching feedback is vague, reps and managers disagree on what a good call is, or the user asks to "build a call scorecard", "score this call", "review my rep's calls", or "make coaching more consistent". It builds behavior-based scorecards for each key conversation type (prospecting, discovery, demo, pricing, close) and uses them to score real calls, localize where reps get stuck, and run chunked practice. Also fires when the user pastes a call transcript and wants it graded. |
| category | Coaching |
Use this when nobody can say precisely what a good call looks like. A scorecard turns "what good looks like" into observable, gradeable behaviors per call type, so scoring, coaching, and practice all point at the same standard. The output is a scorecard per key conversation, a scoring cadence, and — when a transcript is supplied — a scored call with evidence per line item.
1. Map the key conversations
List the conversations where deals are won and lost in the user's motion: prospecting/meeting-setting calls, then the closing motion — discovery, demo, pricing and negotiation, close, and champion navigation. Some teams run these in one call; some across months. Each key conversation gets its own scorecard or its own section of one.
2. Build each card — aim small, miss small
Start with the 10–15 most important behaviors on that call type, not a fully baked card. Pull them from what the top reps actually do on real calls, and build with rep and manager input so the card is adopted, not imposed. Group items into the call's natural chunks (opener → body → wrap-up; discovery → pitch → close → navigation) so a low score localizes exactly where a rep is getting stuck. Each item is an observable behavior — "got them to agree they have a problem," not "good discovery." Score each item 1–5. Grow the card only after the scoring habit exists.
Card templates: for meeting-setting calls read references/prospecting-scorecard.md; for the closing motion read references/closing-scorecard.md.
3. Score on a cadence
Volume makes this work: an SDR having 2–3 conversations a day and an AE running 7–8 demos a month can score every one. Reps self-score, managers score a sample, and scores get tracked over time so trends are visible — where the team is improving, which chunk is dragging. Scoring is also proactive deal rescue: a scored miss ("never found out what matters to the decision-maker" = 1) tells the rep exactly what to go back and get before the next call.
4. Practice with the card
One role play per rep per week, minimum. Chunk it: don't role-play the whole call — drill only the section the rep scores lowest on, with different prompts and angles, until they leave the session on a five. Because the card defines the standard, reps can practice in pairs without the manager and the feedback stays aligned.
When a transcript is supplied
Score it against the matching card: one line per item with the 1–5 score and the quoted moment that earned it (or the absence that cost it). End with the lowest-scoring chunk and one specific drill for it.
What good looks like
A great scorecard program means everyone — rep, manager, leader — grades a call the same way, and a 90-day trend line shows the team improving chunk by chunk. Weak programs grade on vibes ("felt good energy"), roll out a 40-item card nobody uses, or score calls without ever practicing the misses. The scorecard is the bridge between call review and role play; if scores aren't changing what gets practiced, the loop is broken.
Rules
- MUST make every item an observable behavior with evidence. NEVER score on impressions.
- MUST start small (10–15 items) and build the habit before expanding the card.
- MUST chunk practice to the lowest-scoring section — never role-play the whole call to fix one piece.
- MUST keep cards aligned with the playbook as it evolves; a stale card grades the wrong game.
- When scoring a supplied call, NEVER invent moments — cite the transcript or mark the item unobserved.
