AI agents
How to use AI Builder: Prompt templates for creating targeted plays
Sunny Huang
Director of Strategy at Gong
Published on: December 22, 2025
GTM strategies can only be successful if reps use plays consistently and correctly. But static playbooks and disconnected sequencing solutions often add more hurdles than they remove.
That’s where AI Builder, a new Gong Agent, can help. It turns what your top performers are already doing in the field into repeatable content the rest of your org can use — battle cards, case studies, objection handling guides, engagement flows, and more — all built directly from your Gong data and revenue graph.
Below, we’ll walk you through exactly how to get the most value from AI Builder. Let’s begin by setting up AI Builder for success.
The golden rule: Filter first and prompt second
To achieve the best results, it’s essential to understand how the AI works. It doesn’t "know" metadata like dates, stages, or deal amounts until you tell it where to look. This two-step workflow will ensure you get the best outputs from AI Builder at the start:
1. The setup (UI filters)
Use Gong’s filters, folders, or streams to define the context (who, when, and where), allowing AI Builder to isolate information in the dataset.
Examples:
- Date: "Last 30 days"
- Stage: "Discovery" or "Negotiation"
- Outcome: "Closed won" or "closed lost"
- Industry: "Financial services" or "manufacturing"
- Team: "Enterprise SDRs" or "northeast sales team"
2. The prompt (the action)
Write your instructions and specific keywords into the AI Builder prompt window.
Your instructions will define the content of the call (what was said). Once submitted, the AI scans your isolated dataset to find the best calls based on the density of the concepts in your prompt.
Be sure to include words that are likely to be spoken in the transcript or instructions on formatting:
Don't write: "Find discovery calls..."
Why: The AI cannot see the stage.
Do write: "Summarize/analyze the questions asked about pricing..."
Why: The AI can find questions and topics related to "pricing."
Best practices to get the most out of your AI Builder prompts
Format creativity
You are not limited to text summaries. You can instruct the AI to format the output however you like:
- "Create a 3-column table..."
- "Write a multiple-choice quiz..."
- "Draft an email in the voice of..."
- "Write a poem about..." (Yes, it works!)
Keyword density is key
Because the AI searches for the best subset of calls based on your prompt, be sure to use concepts that would naturally be expressed in a real conversation:
Bad prompt: "Tell me about the top reps’ negotiation calls."
Why: AI has no idea who your top reps are Add this to the filters instead.
Good prompt: "Tell me about pricing/discounts/budget objections."
Why: Strong keywords to find relevant calls.
Prompts you can use to accelerate every GTM role
Using our two-step workflow above, we’ve created five groups of role-based prompts that you can enter into AI Builder. Each prompt is customizable and designed to find fast solutions to the specific challenges these roles encounter most often.
Prompts for sales enablement and training
For creating dynamic content, gauging methodology adherence, and knowledge transfer:
| Challenge | Step 1: The filter (context) | Step 2: The prompt (action) |
|---|---|---|
| Dynamic playbooks (always-fresh) | Topic: [Competitor/product] Stage during call: Closed won | "Refresh our playbook for [topic]. Find 5 examples where a rep successfully handled an objection related to this topic. Script the exact response used." |
| Certification and quizzes | Stage: Negotiation Owner: [Top reps] | "Create a multiple-choice certification quiz for new hires. Scenario: The prospect asks for a discount. Provide four potential answers, where only one is the 'correct' answer based on the successful tactics used in these calls." |
| Methodology adoption check | Date: Last 30 Days Team: [Target team] | "Scan these calls for specific mentions of our methodology terms (e.g., 'decision criteria,' 'paper process,' 'economic buyer'). Generate a frequency table showing which terms are being used consistently versus which are being ignored." |
| Harvesting tribal knowledge | Industry: [Target vertical] Stage: Discovery | "Create a 'cheat sheet' for a new rep selling into this industry. List the top five acronyms or industry-specific terms used by these prospects, and provide a simple definition for each based on the context." |
| Onboarding cheat sheet | Industry: [Target vertical] Stage: Discovery | "Create a 'cheat sheet' for a new rep selling into this industry. List the top five acronyms or industry-specific terms used by these prospects, and provide a simple definition for each based on the context." |
| Persona-based elevator pitch | Owner: [Top SDRs/Reps] Outcome: Meeting Booked / Closed Won | "I am pasting our current cold call script here: [Paste Script]. Revamp this script to be persona-specific. Instead of a generic pitch, I want to ask the prospect which of 3 problems is top of mind. Create 3 distinct elevator pitches for a [Target Persona, e.g., VP Sales] tailored to the specific problem they choose (e.g., win rates, sales cycle, ramp time), based on how we successfully solved that problem in these calls." |
| Data-driven onboarding guide | Owner: [Top Performing SDRs] Outcome: Meeting Booked | "Based on these successful calls, create a 'New Hire Guide.' What are the core email principles (subject lines, CTAs) and call structures that actually work here? Ignore general industry advice; focus only on what is working in this dataset." |
| Messaging certification scorecard | Stream: Certification/Role Play Calls | "Analyze these pitches against our new messaging framework. Did the rep mention the 'Trigger Event'? Did they articulate the 'Common Pain'? Did they anchor the 'Value Pitch' back to the pain? Grade the call on a scale of 1-5 based on adherence to this specific talk track." |
| Pipeline "power-up" cards | Stage: Discovery Industry: [Target ICP] | "I want to map our new features to customer problems. Scan these calls for mentions of [Specific Pain Point]. Extract the exact language the customer uses to describe this struggle so I can build a 'Flashcard' for reps on how to position our new [Feature Name] as the solution." |
| Persona-based elevator pitch | Owner: [Top SDRs/Reps] Outcome: Meeting Booked / Closed Won | "I am pasting our current cold call script here: [Paste Script]. Revamp this script to be persona-specific. Instead of a generic pitch, I want to ask the prospect which of 3 problems is top of mind. Create 3 distinct elevator pitches for a [Target Persona, e.g., VP Sales] tailored to the specific problem they choose (e.g., win rates, sales cycle, ramp time), based on how we successfully solved that problem in these calls." |
Prompts for AEs and SDRs
For understanding deal velocity, competitive maneuvering, and prospecting:
| Challenge | Step 1: The filter (context) | Step 2: The prompt (action) |
|---|---|---|
| Fresh competitive intel | Tracker: [Competitor name] Date: Since [Merger/news date] | "What are prospects saying about the recent news regarding [competitor]? Summarize their fears, uncertainties, and specific questions based on these conversations." |
| Pricing battlecard | Stage during call: Proposal Stage now: Closed Won | "Create a 'pricing battlecard.' List the top 3 pushbacks prospects give when we present the enterprise tier, and script the most successful responses used in these calls." |
| Cold call opener script | Call Type: Outbound / Cold call Outcome: Meeting booked | "Analyze the first 30 seconds of these successful calls. Write a script for a cold call opener based on the patterns that successfully kept the prospect on the line." |
| Industry jargon guide | Industry: Financial Services / [Other industry] Stage during call: Discovery | "What are the acronyms and jargon used by financial services prospects in these calls? Create a 'glossary' table with the term and the context in which it was used." |
| Demo framing playbook | Industry: Financial Services / [Other Industry] Stage during call: Discovery | "Analyze how the rep introduces [feature]. What analogy, story, or problem do they describe before they talk about the feature?” |
| Mutual action plan (MAP) generator | Account: [Customer name] | "Build a mutual action plan to close this expansion. Extract specific next steps, dates, and milestones discussed regarding technical evaluation, business case, and implementation. List the key stakeholders mentioned for each step." |
| The "Holiday push" objection | Date: Last Q4 Outcome: Meeting Booked | "Find calls where the prospect said 'call me back after the holidays' or 'reach out next year,' but the rep successfully booked a meeting in December or January. Script the exact argument the rep used to create urgency." |
| The "golden" cold call script | Stream: SDR Meetings Booked Date: Last 90 Days | "Analyze successful cold calls that led to a meeting. Identify the patterns in the first 2 minutes. Write a 'Best Practice Script' that includes the opener, the value prop, and the transition question used most frequently by top performers." |
Prompts for customer success
For retention and identifying risk:
| Challenge | Step 1: The filter (context) | Step 2: The prompt (action) |
|---|---|---|
| QBR prep (the win sheet) | Account: [Customer name] Date: Last 90 days | "Create a summary of all 'wins' or positive sentiments expressed by this customer. List the specific goals they achieved." |
| Churn forensics | Stage Now: Churned / Closed lost Date: Last quarter | "Create a win / loss report on why these customers churned." |
Prompts for product marketing
For understanding the voice of the customer, refining messaging, and more:
| Challenge | Step 1: The filter (context) | Step 2: The prompt (action) |
|---|---|---|
| Detailed battlecards | Competitor: [Competitor Name] Outcome: Closed Won | "Create a table that summarizes the prospect’s stated pain with [Competitor], the feature of ours they valued most, and a direct quote from the customer validating our win." |
| Launch messaging/naming | Topic: [Problem Area] Date: Last 30 Days | "We are naming a new feature regarding [Function]. List the verbs and nouns customers use when complaining about this specific problem.” |
| Customer case studies | NPS: Promoter OR Sentiment: High | "Find 5 calls where the customer used the words 'love,' 'game changer,' or 'ROI.' Extract the specific quote and the business metric they associated with it." |
| Ad copy generation | Stage: Discovery Industry: [Target Industry] | "Based on the pain points described by these prospects, write 5 headlines for a LinkedIn ad campaign that uses their exact vocabulary to describe the problem." |
| Objection handling guide | Topic: [New Feature] | "Create a table of objections regarding [New Feature] that summarizes, the Objection, the Rep's Response" |
AI Builder deploys your proven revenue strategies, automatically and at scale
As part of the Gong Revenue AI Operating System, AI Builder turns repeatable GTM execution into winning plays that every team can ship quickly, confidently, and grounded in the data that drives revenue.
From enablement leaders creating consistent messaging to RevOps teams operationalizing plays with clear scorecards to sales reps asking, “Show me what actually wins against this competitor,” every role can use AI Builder to uncover real customer insights and replicate what works.
To get started with AI Builder, take a demo of the Gong Revenue AI OS to see how you can start building better GTM plays.
State of Revenue 2026
Loading form...
Discover more from Gong
Check out the latest product information, executive insights, and selling tips and tricks, all on the Gong blog.
