Analyze · Call-Transcript & Objection-Pattern Analysis
Identify Common Objection Patterns Across Calls
Cluster objections across a batch of agency calls to see which concerns are systemic vs. one-offs.
foundermanagerIntermediate⏱ 4-6 hours of manual call review
When to use
Run this monthly or quarterly across a batch of 8-30 transcripts to see which objections are actually killing deals vs. anecdotes you keep hearing in sales meetings. Use the output to prioritize sales enablement, case-study production, and pricing-page updates.
The prompt
You are a sales analyst trained to find objection patterns across agency sales calls. You only report patterns supported by quoted evidence. Agency: [AGENCY_NAME] — [SERVICES] Date range: [DATE_RANGE] Call mix: [CALL_MIX] Transcript batch (separated by ===CALL N===): [TRANSCRIPT_BATCH] Cluster objections that appear across multiple calls. Rank by frequency. For each pattern, show how often it appeared, 2-3 exact quotes from different calls, and a hypothesis for why this pattern exists at [AGENCY_NAME]. - Quote exact lines as evidence — no paraphrase as fact - Flag any pattern with sample Pattern table: Rank | Pattern Name | # of Calls | Sample Quotes (with call #) | Hypothesis | Suggested Org-Level Fix Then: "Top 3 things to fix this quarter" — ranked by ROI of fixing.
Variables
- [AGENCY_NAME] — Your agency name
- [SERVICES] — What you sell
- [DATE_RANGE] — Window the transcripts cover
- [CALL_MIX] — How many discovery vs pitch vs proposal calls in the batch
- [TRANSCRIPT_BATCH] — Multiple transcripts separated by ===CALL 1===, ===CALL 2===, etc.
Example input
Agency: NorthLoop Digital — SEO + paid search retainers Date range: Apr 1 – Apr 30 2026 Mix: 12 discovery, 6 proposal review (18 total) Batch: [18 transcripts pasted with ===CALL N=== separators]
Example output
| Rank | Pattern | # Calls | Sample Quotes | Hypothesis | Org Fix | |---|---|---|---|---|---| | 1 | "How do we know you're different from our last agency?" | 11/18 | Call 3: "the last guys promised the world and ghosted" • Call 9: "we've been burned twice" | Market is post-burn; you have no in-funnel proof of differentiation | Build a 'how we're different' 90-second video + 3 written case studies with departing-agency comparisons | | 2 | Pricing shock vs. cheaper local competitor | 8/18 | Call 5: "a local guy quoted $2,500" • Call 14: "that's 3x what we budgeted" | Discovery isn't anchoring value before price | Add a ROI-anchoring slide before pricing reveal; train reps on the 'cost of doing nothing' framing | | 3 | Commitment-length resistance (6mo minimum) | 6/18 | Call 2, 11, 17 quotes… | New buyers don't trust a 6mo lock without proof | Pilot a 90-day structured trial as the entry SKU | | 4 | "We need to talk to [partner/spouse]" | 2/18 | — | anecdotal — needs more data | — | Top 3 fixes this quarter: (1) differentiation proof asset, (2) value-anchoring in discovery, (3) 90-day pilot offer.
Pro tips
- Don't run this on fewer than 8 transcripts — patterns aren't real below that
- Tag each transcript with deal outcome (won/lost/open) in [CALL_MIX] so Claude can correlate objection patterns to losses
- Re-run quarterly with the SAME prompt — track whether your 'org fixes' actually reduced pattern frequency
Works with
ClaudeChatGPTGemini
Done with prompts? Time to install the system
Book a STAOS callRelated prompts