13 Gong call analysis prompts for product marketers, GTM, and CS teams

By Ahmet Nuri Ozcelik · Product Marketing Leader & GTM Engineer · 10 min read

Quick answer: Gong call analysis becomes useful when the same prompt runs across every call in a cohort — not when one call gets a per-call AI summary. Callmine reads every Gong call you have access to (sales, customer success, and renewal calls — not only sales) and applies a programmable prompt across the cohort, returning a structured pattern instead of a narrative summary. The 12 prompts below are the starting templates teams reuse week over week, plus one CS-only prompt for renewal-risk analysis.

You can run a programmable call analysis prompt against every Gong call in a period — sales, CS, and renewal — to surface patterns in win/loss, competitors, objections, discovery quality, customer language, deal risk, and renewal health. The trick is writing prompts that produce structured, comparable output across calls — not narrative summaries that only describe one conversation.

Generic call summaries are not enough for revenue and GTM analysis. A summary tells you what happened on call A. A programmable prompt, applied across calls A through Z, tells you what is happening across the pipeline.

The structural difference between per-call Gong AI and programmable cohort analysis:

How to use these prompts

Each prompt below is written to be run across a cohort, not a single call. Use them in Callmine or any programmable call analysis platform that supports custom prompts and CRM segmentation. Cohorts can be sales calls, customer success calls, renewal calls, or any mix — Callmine reads every call type in your Gong archive. The recipe is always the same:

1. 01Pick the cohort. Filter Gong calls by the right CRM attributes — call type (sales / CS / renewal), deal stage, segment, ACV band, rep or CSM, time window. The cohort defines the question.
2. 02Define the prompt once. Write it as if every call should be evaluated the same way. Use clear output structure (rank, score, list, theme + evidence).
3. 03Run it across the cohort. Each call gets the same prompt. Each per-call answer becomes input to an aggregate report.
4. 04Read the pattern. The output you care about is the cohort-level pattern, not the per-call answer.

The 13 prompts below are starting points. Adjust the wording, the output schema, and the segmentation to fit how your team thinks about its pipeline. For the deeper conceptual frame, see the post on programmable call analysis.

1. Win/loss analysis

For this call, identify the deal outcome (won, lost, no-decision, other). List the top three reasons the deal moved that direction, quoting the buyer or rep moments that support each reason. Note any competitor mentions, pricing pushback, discovery gaps, or stakeholder issues. If the call doesn't clearly indicate outcome, say so.

Run across closed-won and closed-lost calls in the period. The aggregate ranks the most common reasons by outcome.

2. Competitor mentions

Identify every competitor named on this call. For each mention, capture: competitor name, who brought it up (buyer, customer, rep, or CSM), the context (pricing, integration, feature, brand, renewal alternative), and how the rep or CSM responded (reframed, deferred, agreed, dodged).

Run across all sales and CS / renewal calls in the period — not only sales. Competitor mentions on renewal calls are a different signal than mentions in discovery: they're the live "should we switch?" conversation. The aggregate produces a competitor leaderboard split by call type (acquisition vs. renewal-threat) plus the typical objection patterns each competitor introduces at each stage.

3. Pricing objections

Did pricing come up on this call? If yes, describe: who raised it, the framing (sticker shock, budget timing, ROI uncertainty, comparison to alternative), the dollar context if mentioned, and how the rep responded. Include the moment as a quote.

Run across late-stage and closed-lost calls. The aggregate distinguishes "we're too expensive" from "they don't see the value yet" — different problems with different fixes.

4. Discovery quality scoring

Score the rep's discovery on this call from 1 to 5 across these dimensions: pain, urgency, decision process, stakeholders, business impact, and budget signal. For each dimension, quote the strongest moment and note any signal the rep missed or skipped past.

Run across discovery-stage calls per rep. The aggregate gives per-rep scorecards that are defensible to coach against.

5. Customer pain themes (voice of customer)

Extract every distinct pain point the buyer or existing customer mentioned on this call. For each, capture: their own words (verbatim quote), the underlying problem they're describing, the relationship stage (prospect, new customer, expansion, renewal), and the segment context (industry, company size, role) if known.

Run across sales calls AND customer success / renewal calls in a target segment — both populations are saying the same thing in slightly different vocabulary, and you want both. Prospects describe the pain they're hiring you to solve; existing customers describe the pain that's still unresolved (or newly emerged) after implementation. The aggregate becomes a voice-of-customer artifact in the customer's own language — not the marketing team's — and the prospect/customer split is where the most useful messaging insights live.

6. Messaging validation

The company's positioning is: [paste your one-liner]. For this call, identify any moment where the buyer's language matched, partially matched, or contradicted that positioning. Quote each moment. Note where the rep's pitch echoed the positioning and where they drifted.

Run across discovery and pitch calls. The aggregate tells you whether your positioning is landing in the room — and which segments resonate vs. which don't.

7. Feature requests

List every product capability the buyer asked for, mentioned wishing existed, or compared to a competitor's offering. For each, capture: the request, the use case the buyer described, and whether the rep committed, deflected, or noted it for product.

Run across all sales calls in the period. The aggregate is a triaged feature-request list anchored to real buyer language and deal context.

8. Status quo risk

Did the buyer mention any reason to stay with their current solution or do nothing? Capture each: the reason given (cost of change, satisfaction, low priority, internal politics, prior failed migration), and the rep's response.

Run across slow-moving and closed-lost deals. The aggregate explains why "no decision" is winning your deals — usually a more revealing pattern than competitor losses.

9. Stakeholder mapping

Identify every named stakeholder mentioned on this call (buyer-side or rep-side). For each, capture: name or role, function, decision authority signal (champion, blocker, user, economic buyer, unknown), and the moment that signal appeared.

Run across active opportunities. The aggregate builds a defensible stakeholder map per deal — and surfaces deals missing an economic buyer entirely.

10. Implementation concerns

Did the buyer raise any concerns about implementation, change management, integration, security, or rollout risk? Capture each concern with the buyer's exact words and the rep's response.

Run across mid- to late-stage calls. The aggregate is a list of the friction your sales motion needs to address — most of it solvable with a one-pager rather than a price discount.

11. Segment-specific objections

For this call, capture the buyer's segment (industry, company size, region) and every objection raised. Tag each objection with whether it was about price, fit, timing, integration, competitor, internal politics, or other.

Run across the full quarter, then segment the aggregate by industry or ACV band. The pattern almost always differs — enterprise objects to integration depth, mid-market objects to time-to-value.

12. Deal risk signals

Identify any signal on this call that could indicate the deal is at risk: stakeholder churn, missed economic buyer, vague next steps, low engagement, competitive heat, budget freeze mentions, or champion silence. Quote each signal.

Run weekly across active opportunities. The aggregate is an at-risk list anchored in actual call moments — not pipeline gut feel.

13. Renewal risk signals (CS calls)

For this customer success or renewal call, identify any signal that the account may be at risk of churn or downgrade. Capture each: signal type (champion departure, executive sponsor change, reduced usage, competitor evaluation, budget pressure, satisfaction complaint, integration breakage, unmet expectation), the customer's exact words, and the CSM's response (acknowledged, escalated, deflected, committed to follow-up). Score overall account risk from 1 (healthy) to 5 (likely to churn) with one sentence of justification.

Run weekly across all CS and renewal calls — not sales calls. The aggregate is a CS-side at-risk list for the upcoming renewal cohort, anchored in actual customer language, weeks or months before the renewal date. This is the same shape as Prompt 12 but for the install base, and it's the prompt CS leaders ask for first once they realize Callmine reads their CS calls and not just sales.

How to run this in Callmine

Each of the 13 prompts above is one Callmine run. Here's the end-to-end shape, using Prompt 2 (Competitor mentions) on the CS side as the worked example — because that's the prompt where teams most often realize their Gong archive is bigger than they thought.

1. Pick the cohort in Gong. Filter to: Call type = Customer Success OR Renewal, Date = last 30 days, Account segment = Mid-market and up. Callmine reads whichever Gong call set you select — it doesn't modify Gong or write anything back.

2. Run the saved "Competitor mentions — CS" template. The prompt:

Identify every competitor named on this call. For each mention, capture: competitor name, who brought it up (customer or CSM), the context (renewal alternative, price comparison, feature gap, frustration), and how the CSM responded. Flag any moment that suggests the customer is actively evaluating switching.

3. Output lands as a structured table in your Callmine workspace — one row per competitor mention with account, call date, who raised it, context, CSM response, and switching-signal flag. Export to Slack, Sheets, or Notion. Pipe the switching-signal flagged rows into your CS team's weekly account review.

The same three-step shape works for all 13 prompts; the only thing that changes is the Gong filter and the prompt body. Because the prompt is held constant week over week, the output is comparable across periods — which is the whole point of programmable Gong call analysis.

FAQ

What is a Gong call analysis prompt?

A Gong call analysis prompt is a structured question or instruction applied to a call's transcript and metadata. The same prompt run across many calls — sales, CS, or renewal — produces an aggregate view: patterns across the cohort, not the anecdote from one call.

How do you run the same prompt across many Gong calls?

Use a programmable call analysis platform like Callmine: connect Gong, filter the calls you want (by call type, deal stage, segment, owner, time window), define the prompt once, and run it across the entire selection. The platform applies the prompt call by call and aggregates the per-call outputs into a single report.

Are these prompts a substitute for the prompts inside Gong?

No. Gong's built-in AI focuses on per-call summaries and rep coaching. These prompts are designed for cross-call analysis — pattern-finding across hundreds or thousands of calls, segmented by CRM data, including non-sales call types Gong's built-in tools don't analyze cohort-wide.

Can I run these prompts on a small sample?

Yes, but the value compounds at scale. Ten calls may surface anecdotes; a hundred starts to surface patterns. Aim for the cohort size that matches the question — quarterly closed-lost might be 30 calls; competitor mining across the year might be 500; renewal-risk analysis across the install base might be every CS call in the quarter.

Run these prompts on your own Gong calls

Each of these prompts is one Callmine run away. Connect Gong, filter the cohort, paste the prompt, hit run. The free trial is 100 calls, 30 days, no credit card — enough to actually pick three of these prompts and prove out the workflow on your own pipeline.

For the longer-form playbook on running a defensible win/loss program, read win/loss analysis from Gong calls. For the conceptual frame on the analysis layer above conversation intelligence, read programmable call analysis.