Sales AI Skill
Call Recording Transcription Ai
Record, transcribe, and analyze sales calls using AI to extract insights, action items, objections, and coaching opportunities. Use when analyzing sales calls, extracting call insights, automating call notes, improving talk-listen ratios, or coaching reps f...
Call Recording, Transcription & AI Analysis
Capture and extract intelligence from every sales conversation automatically.
Workflow
- Auto-record all sales calls via video conferencing integration (Zoom, Teams, Meet) or phone system.
- Transcribe audio to text with speaker identification (rep vs. prospect).
- AI analyzes conversation for pain points, objections, competitor mentions, budget hints, next steps.
- Generate structured call summary with highlights, action items, and deal insights.
- Auto-sync call data to CRM (notes, tasks, opportunity updates).
- Surface coaching insights to managers based on call quality metrics.
- Build searchable call library for team-wide learning and best practice sharing.
Call Analysis Framework
AI CALL ANALYSIS EXTRACTION FRAMEWORK
========================================
1. CONVERSATION STRUCTURE ANALYSIS:
─────────────────────────────
Talk-to-Listen Ratio:
- Rep talk time: [%] (optimal: 40-45%)
- Prospect talk time: [%] (optimal: 55-60%)
- Silence periods: [count, total duration]
- Monologue length: [avg seconds, max seconds]
- Interrupt count: [rep interrupts, prospect interrupts]
Conversation Flow:
- Opening/rapport time: [X minutes]
- Discovery time: [X minutes]
- Presentation/pitch time: [X minutes]
- Objection handling time: [X minutes]
- Closing/next steps time: [X minutes]
- Small talk ratio: [%] (target: <15%)
2. PAIN POINT EXTRACTION:
──────────────────────
Identified Pain Points:
- [Pain 1] — mentioned [X] times — confidence [%]
Quote: "[exact quote from prospect]"
Context: [surrounding conversation excerpt]
- [Pain 2] — mentioned [X] times — confidence [%]
Quote: "[exact quote from prospect]"
Pain Severity Indicators:
- Financial impact mentioned: [Yes/No, $ amount if stated]
- Time waste quantified: [Yes/No, hours/week if stated]
- Revenue impact: [Yes/No, % or $ if stated]
- Emotional language (frustrated, annoyed, overwhelmed): [Yes/No]
3. BANT SIGNALS FROM CONVERSATION:
──────────────────────────────
Budget:
- Budget mentioned: [Yes/No]
- Budget range stated: [$ amount if mentioned]
- Budget process described: [summary]
- Pricing sensitivity: [High/Medium/Low]
- Auto-update CRM BANT budget score: [1-4]
Authority:
- Decision-maker on call: [Yes/No]
- Decision process described: [summary]
- Other stakeholders mentioned: [names/roles]
- Approval process: [summary]
- Auto-update CRM BANT authority score: [1-4]
Need:
- Need validated: [Yes/No]
- Specific use cases discussed: [list]
- Current solution limitations: [summary]
- Need urgency: [High/Medium/Low]
- Auto-update CRM BANT need score: [1-4]
Timeline:
- Timeline mentioned: [Yes/No]
- Target date stated: [date if mentioned]
- Decision process timeline: [summary]
- Budget cycle reference: [quarter if mentioned]
- Auto-update CRM BANT timeline score: [1-4]
4. OBJECTION DETECTION:
────────────────────
Objections Raised:
- [Objection 1] — Category: [Price/Timing/Competitor/Feature/Trust]
Rep Response: [summary of how rep handled it]
Response Quality: [Strong/Weak/Not Addressed]
Resolution: [Resolved/Partially Resolved/Open]
- [Objection 2] — Category: [...]
Common Objection Patterns:
- Price: "too expensive", "budget", "ROI"
- Timing: "not now", "later this year", "evaluating"
- Competitor: "using [competitor]", "happy with current"
- Feature: "missing [feature]", "doesn't do [function]"
- Trust: "never heard of you", "need references"
5. COMPETITOR INTELLIGENCE:
────────────────────────
Competitors Mentioned: [list with frequency]
Current Incumbent: [name if identified]
Competitor Sentiment: [Positive/Neutral/Negative]
Competitor Strengths (per prospect): [what they like]
Competitor Weaknesses (per prospect): [what they dislike]
Switching Consideration: [Yes/No/Unclear]
6. NEXT STEPS & COMMITMENTS:
─────────────────────────
Agreed Next Steps:
- [Action 1] — Owner: [Rep/Prospect] — Due: [Date]
- [Action 2] — Owner: [Rep/Prospect] — Due: [Date]
Meeting Scheduled: [Yes/No — date/time if scheduled]
Commitment Level: [Strong/Medium/Weak]
Follow-Up Required: [Yes/No — details]
7. SENTIMENT & ENGAGEMENT:
───────────────────────
Overall Sentiment: [Positive/Neutral/Negative]
Sentiment Trajectory: [Improving/Stable/Declining]
Engagement Level: [High/Medium/Low]
Key Positive Moments: [timestamps with quotes]
Key Negative Moments: [timestamps with quotes]
Enthusiasm Indicators: [excitement words, questions asked, follow-up interest]
Call Quality Scorecard
CALL QUALITY SCORECARD (0-100)
================================
DISCOVERY QUALITY (30 points):
──────────────────────────
Open-ended questions asked: [count] (target: 8+ for discovery calls)
Pain points uncovered: [count] (target: 3+)
Follow-up questions (probing): [count] (target: 5+)
Prospect talk ratio: [%] (target: 55-60%)
Business impact questions asked: [Yes/No]
Decision process explored: [Yes/No]
Score: [0-30]
VALUE COMMUNICATION (25 points):
────────────────────────────
Features linked to prospect's stated pain: [count]
ROI/benefit statements made: [count]
Relevant case studies shared: [Yes/No]
Social proof provided: [Yes/No]
Solution tailored to their context: [Yes/No]
Score: [0-25]
PROCESS & PROFESSIONALISM (25 points):
───────────────────────────────────
Clear meeting agenda set: [Yes/No] (+5)
Time managed well (on schedule): [Yes/No] (+5)
Multiple interruptions avoided: [Yes/No] (+5)
Active listening demonstrated: [Yes/No] (+5)
Clear next steps confirmed: [Yes/No] (+5)
Score: [0-25]
CLOSING & NEXT STEPS (20 points):
──────────────────────────────
Explicit next step agreed: [Yes/No] (+8)
Specific date/time scheduled: [Yes/No] (+6)
Ownership assigned (who does what): [Yes/No] (+6)
Score: [0-20]
SCORE TIERS:
85-100: Exemplary — best practice call, share with team
70-84: Strong — minor coaching opportunities
55-69: Adequate — needs focused coaching on 1-2 areas
40-54: Below Average — significant coaching needed
<40: Critical — immediate manager intervention required
Automated CRM Sync
AUTO-CRM UPDATE MAPPING
=========================
After each call, AI auto-updates CRM with:
Activity Log:
- Call recorded: [timestamp, duration, platform]
- Call summary: [AI-generated 3-5 sentence summary]
- Key topics discussed: [tags]
- Call quality score: [0-100]
Contact Updates:
- Pain points: [appended to contact notes]
- Communication style: [updated personality profile]
- Engagement level: [updated trend]
Opportunity Updates:
- BANT scores: [auto-updated based on conversation]
- Competitors: [added/updated]
- Next steps: [tasks created with due dates]
- Deal stage: [advance if criteria met]
- Confidence level: [updated based on sentiment]
Tasks Created:
- [ ] Send promised collateral by [date]
- [ ] Schedule demo with [stakeholder]
- [ ] Follow up on [specific topic]
- [ ] Research [competitor/industry] before next call
- [ ] Introduce to [executive/colleague]
Flags:
- Red flag: [if negative sentiment, unresolved objection, or stalled next step]
- Green flag: [if strong commitment, budget confirmed, or timeline accelerated]
Coaching Insights for Managers
MANAGER COACHING DASHBOARD
============================
Team-Level Metrics:
Total calls analyzed: [count]
Average call quality score: [X/100]
Average talk-to-listen ratio: [X:Y]
Average discovery questions per call: [X]
Calls with clear next steps: [%]
Calls with pain points identified: [%]
Individual Rep Rankings:
╔═══════════╦═══════════════╦═══════════════╦═════════════════╗
║ Rep ║ Avg Score ║ Calls This Wk ║ Trend ║
╠═══════════╬═══════════════╬═══════════════╬═════════════════╣
║ [Name] ║ [X/100] ║ [count] ║ [↑/→/↓] ║
║ [Name] ║ [X/100] ║ [count] ║ [↑/→/↓] ║
╚═══════════╩═══════════════╩═══════════════╩═════════════════╝
Coaching Alerts (Auto-Generated):
┌──────────────────────────────────────────────────────────┐
│ ⚠️ [Rep Name] — Talk Ratio Alert │
│ Talk ratio 65:35 (rep:prospect) on 3/5 calls this week │
│ Coaching focus: Ask more questions, talk less │
│ Best practice example: [Name] maintains 42:58 ratio │
└──────────────────────────────────────────────────────────┘
┌──────────────────────────────────────────────────────────┐
│ ✅ [Rep Name] — Best Practice Call │
│ Scored 92/100 on discovery call with [Company] │
│ Highlight: Asked 12 open-ended questions, uncovered 5 │
│ pain points, secured commitment for next step │
│ Action: Share recording with team in next huddle │
└──────────────────────────────────────────────────────────┘
┌──────────────────────────────────────────────────────────┐
│ 🔴 [Rep Name] — Missing Next Steps │
│ 4/5 calls this week had no clear next step agreed │
│ Coaching focus: Always close with specific commitment │
│ Training resource: "Closing the Call" module │
└──────────────────────────────────────────────────────────┘
Coaching Session Prep (Auto-Generated):
- 3 best call moments from rep this week (with timestamps)
- 3 improvement opportunities (with timestamps and suggested approaches)
- Comparison to team average on key metrics
- Recommended training modules based on gaps
- Progress since last coaching session
Call Library & Knowledge Sharing
SEARCHABLE CALL LIBRARY FEATURES
==================================
Search Capabilities:
- Full-text search across all transcriptions
- Search by: company name, contact name, rep name, topic, objection type
- Filter by: call quality score, deal outcome, industry, company size
- Advanced: "Show me calls where prospect mentioned [competitor]"
- Advanced: "Find calls with highest conversion to demo"
Best Practice Repository:
- Top-performing call recordings (by quality score and outcome)
- Best objection handling examples (tagged by objection type)
- Best discovery question sequences (tagged by industry/pain)
- Best closing techniques (calls that secured strong commitments)
- Curated by AI + manager curation
Team Learning:
- Weekly "Call of the Week" highlighting best practice
- Monthly skill improvement report (team-wide progress)
- New rep training library (top calls from top performers)
- Role-specific call examples (discovery, demo, negotiation)
- Competitor call analysis (how team handles specific competitors)
Privacy Controls:
- Reps can flag calls as private (not shared with team)
- PII redaction option for external sharing
- Role-based access (reps see own calls, managers see team, dir sees all)
- Compliance: call recordings stored per data retention policy
Edge Cases
- Poor audio quality: Background noise, overlapping speakers, or poor connection degrading transcription accuracy
- Mitigation: AI confidence scoring on transcription; flag low-confidence segments for manual review
- Recommend rep to use noise-canceling headset and quiet environment
- Use phone system recording (higher quality) when video conferencing quality is poor
- Post-call, rep can edit/correct AI-generated notes
- Multi-language conversations: Calls conducted partially or fully in non-English languages
- Mitigation: Support transcription in 20+ languages with real-time translation
- Flag translation confidence scores; low-confidence translations marked for review
- Maintain conversation in original language for searchable library
- Provide bilingual summaries for manager review
- Group calls with many participants: Panel demos or multi-stakeholder calls with 6+ participants
- Mitigation: Speaker identification for each participant; name mapping from calendar invite
- Separate analysis per speaker (who said what, engagement by person)
- Identify which stakeholders were most/least engaged
- Summarize key points by stakeholder role for follow-up personalization
- Sensitive conversations: Calls involving pricing negotiations, legal discussions, or confidential information
- Mitigation: Allow reps to flag calls as "sensitive" — restricted access, limited AI analysis
- Exclude from team library and coaching examples
- Retain only basic metadata (duration, participants, outcome) without full transcription
- Compliance: adhere to company data classification and retention policies
Integration Points
- Zoom/Teams/Meet: Native call recording; auto-start recording on meeting begin
- RingCentral/Aircall/Dialpad: Phone system integration for recorded calls
- Salesforce/HubSpot: Auto-sync call notes, tasks, BANT updates to CRM
- Gong/Chorus: Alternative conversation intelligence platform integration (if using as primary)
- Slack/Teams: Coaching alerts delivered to manager channels; call summary sharing
- LMS: Training module recommendations based on call quality gaps
- Tableau/Looker: Call analytics dashboards for sales leadership