Quick Note:
Thank you to everyone who joined today's AI Agent Best Practice Meetup. I will be honest, I had no slides, no seed content, and no idea how many people would show up. You showed up anyway and you brought the goods.
27 Visioneers. 10 real workflows. 45 minutes of the most practical AI conversation I have hosted in five years at Gong.
Here is what we covered:
- Using Gong AI Agents natively:
@Jamie Wolfson shared how his team at Coherent uses AI Builder to generate business cases from months of call recordings in 30 minutes -- and saw a measurable lift in deal wins as a result. @Andrea Ihara shared three workflows she runs every week: a deal close query, a weekly Forecast Edge digest for sales leaders, and an outbound script generator built from closed-won data. @Itai Sagi at Palo Alto Networks is using AI Builder to find product gaps across all accounts and monetize them with pipeline impact. @Sydney Tress built a system that surfaces customer story leads automatically using AI Trackers -- so the evidence comes to her instead of her hunting for it. - Connecting Gong to external AI:
@Tasha Hendershot exports Gong call data to Claude Cowork to build MEDDIC heat maps per rep -- showing exactly where pipeline is stalling for each person. @Kyleigh Oosthuyse at Invoca is connecting Gong to Gumloop to build full account research briefs automatically across her entire book of business.
And a few things that just work: Jamie and Itai both independently confirmed they are using the AI Data Extractor to auto-populate MEDDIC fields in their CRM -- reps just validate instead of typing.
This is so exciting!
-Nisha
I have written all 10 workflows up as simple step-by-step how-to guides.
Organized into three buckets:
1. Gong AI Agents — workflows using native Gong AI features
2. AI in General — workflows connecting Gong data to external AI tools
3. Everything Else — workflows using Gong features without AI as the primary driver
BUCKET 1: GONG AI AGENTS
Workflows powered by AI Builder, AI Trackers, and Ask Anything
WORKFLOW 1: AI-generated business case from call history
Shared by: Jamie Wolfson, Coherent
Gong feature: AI Builder
What it does: Turns 9-12 months of recorded calls into a ready-to-send business case in 30 minutes.
Step by step:
1. Go to AI Builder in Gong
2. Select the account you want to build the case for
3. Filter to the calls you want to include
4. Enter this prompt: Develop a business case. Include: executive summary, key challenges, proposed solution, ROI. Keep it concise.
5. Run the builder and review the output
6. Copy into a Word or Google Doc in your company branding
7. Sales rep spends 30 minutes formatting — not writing from scratch
8. Attach to your proposal before sending
The outcome: Coherent saw a measurable improvement in deal wins. Reps who previously skipped writing exec summaries are now consistently sending them.
Tip: Works best for deals with 3+ months of recorded calls. The more Gong data on the account, the stronger the output.
WORKFLOW 2: Top 3 actions to close this deal on time
Shared by: Andrea Ihara, Algosec
Gong feature: AI Ask Anything (inside a deal)
What it does: Gets you a prioritized action list for any deal in 30 seconds — no call re-listening required.
Step by step:
1. Open the deal in Gong
2. Navigate to the AI Ask Anything field
3. Type: What are the top three things we need to do to close this deal on time?
4. Gong pulls from all calls, emails, and connected Salesforce records
5. Use the output in your next forecast call or send directly to the AE
The outcome: Sales leaders get deal-level intelligence they do not have time to gather themselves.
Tip: Andrea uses this as the backbone of her weekly Forecast Edge digest — see Workflow 3.
WORKFLOW 3: The Forecast Edge — weekly pipeline digest
Shared by: Andrea Ihara, Algosec
Gong feature: AI Trackers + AI Ask Anything
What it does: A short weekly email to sales leaders connecting one tracker insight to real pipeline revenue.
Step by step:
1. Every Friday morning, filter your AI trackers to deals at 80%+ probability closing this quarter
2. Pick one tracker theme for the week — competitive mentions, multi-threading, negative sentiment
3. For each flagged deal, run the close-time query from Workflow 2
4. Write a short 3-5 paragraph email called the Forecast Edge
5. Connect the tracker to a revenue number: this signal is showing up in 35% of our 80% deals worth $X
6. Send to sales leaders and optionally the broader sales team
7. Rotate the tracker theme each week
The outcome: Sales leaders start paying attention to tracker data because it is tied to real pipeline numbers every week.
Tip: Data without revenue context gets ignored. Always attach a dollar amount.
WORKFLOW 4: Customer evidence and story signal tracker
Shared by: Sydney Tress
Gong feature: AI Trackers + Smart Trackers
What it does: Builds a pipeline of customer story leads automatically — signals come to you instead of you hunting for them.
Step by step:
1. Identify the keywords that signal a success story in your industry
2. Create Smart Trackers or AI Trackers for: cannot live without this, our team uses this every day, we saw a decrease in, ROI was
3. Set up a Gong stream or Slack notification for calls that trigger those trackers
4. Review flagged calls weekly
5. For each signal note: what team is using the product, how they are using it, what was the measurable outcome
6. If the call does not answer those questions, ask the AE for context before going back to the customer
7. Qualify into your evidence program and initiate the story process
The outcome: A scalable pipeline of story leads without cold outreach.
Tip: Not every tracker hit is a story. You still need to qualify. But having the list saves hours of manual call review.
WORKFLOW 5: Product gap finder across all accounts
Shared by: Itai Sagi, Palo Alto Networks
Gong feature: AI Builder + AI Trackers
What it does: Synthesizes what customers are saying about product gaps across all your accounts — with frequency and pipeline impact.
Step by step:
1. In AI Builder, filter calls using your product-specific AI Tracker
2. Select the relevant date range — last 90 days or last planning cycle
3. Enter this prompt: What are customers identifying as product gaps or missing features? Group by theme and show which deal types are most affected.
4. Identify the top 5-10 gaps by frequency
5. For each gap, note the pipeline value of deals where it appeared
6. Format into a one-page report: gap name, frequency, pipeline impact, representative quotes
7. Share with your product team and roadmap owners
The outcome: Product team gets customer-validated gap analysis grounded in real call data. Pipeline impact helps prioritize what matters most to revenue.
Tip: Same workflow works for competitive intelligence — just swap the product tracker for a competitor mention tracker.
WORKFLOW 6: Outbound call script from closed-won deals
Shared by: Andrea Ihara, Algosec
Gong feature: AI Builder
What it does: Generates an outbound call script from your own closed-won call data — built on what actually works.
Step by step:
1. Go to AI Builder
2. Filter calls to your inside sales or SDR team
3. Further filter to closed-won deals
4. Set your date range to the last 12 months
5. Enter this prompt: What keywords, pain points, and topics appear most frequently in these calls? Help me create an outbound call script based on these patterns.
6. Review and edit into your company script template
7. Test with your inside sales team and track conversion against the old script
The outcome: A script built from real customer language and real winning patterns — not guesswork.
Tip: If the output is too long follow up with: Shorten this to the top 5 talking points and make it conversational.
BUCKET 2: AI IN GENERAL
Workflows connecting Gong data to external AI tools
WORKFLOW 7: MEDDIC heat map for rep coaching
Shared by: Tasha Hendershot (company withheld)
Tool: Gong export + Claude Cowork (external AI)
What it does: Shows exactly where each rep is strong and weak on MEDDIC qualification — across their entire book of deals.
Step by step:
1. In Gong, filter your call library to calls where MEDDIC-related trackers were triggered
2. Export the call data as a CSV or transcript export
3. Open Claude Cowork or your preferred AI tool
4. Upload the exported data
5. Enter this prompt: Produce a MEDDIC heat map per rep showing which qualification questions were asked well and which were missed
6. Review the output for each rep
7. Use the heat map in your next coaching session — start with reps where pipeline is stalling
The outcome: Identified exactly where pipeline was stalling for each rep. Helped the team ask the right questions in the right stages.
Tip: Tasha is testing whether this can be done entirely inside Gong using AI Builder to skip the export step. Worth trying.
WORKFLOW 8: Account research workflow using Gumloop
Shared by: Kyleigh Oosthuyse, Invoca
Tool: Gong + Gumloop (external workflow platform)
What it does: Builds a full account research brief automatically by combining Gong call data with external data sources.
Step by step:
1. Sign up for Gumloop (gumloop.com) — a workflow platform similar to Zapier but built for AI
2. Connect your Gong account to Gumloop using the integration
3. Build a workflow that: pulls conversation history from Gong, combines it with external data, runs an AI analysis
4. Configure your output format — brief, summary, or talking points
5. Run across your full book of business or trigger on a specific account
6. Validate key outputs in Gong AI while you are testing
The outcome: Account managers get a full picture of each account before renewals or expansion calls — without hours of manual research.
Tip: Gumloop supports Claude, web scraping, and custom code. Good for account managers and CSMs with large books of business.
BUCKET 3: EVERYTHING ELSE
Workflows using Gong features without AI as the primary driver
WORKFLOW 9: Auto-populate MEDDIC fields in your CRM
Shared by: Jamie Wolfson, Coherent — also confirmed by Itai Sagi, Palo Alto Networks
Gong feature: AI Data Extractor + HubSpot or Salesforce
What it does: Gong infers MEDDIC field answers from call recordings and pushes them to your CRM automatically — reps just review and validate.
Step by step:
1. Work with your Gong admin to enable the AI Data Extractor
2. Map your MEDDIC fields in HubSpot or Salesforce to Data Extractor outputs
3. Configure the extractor to infer answers for each MEDDIC category from call recordings
4. Set up the CRM push so fields update automatically after each call
5. Train reps to review and validate pre-populated fields instead of typing
6. Run a data quality check after the first month
7. Iterate on field mappings where the extractor is missing context
The outcome: Reps stopped leaving MEDDIC fields blank. Data quality improved dramatically. Both Coherent and Palo Alto Networks confirmed the same result independently.
Tip: Jamie moved fast on this one and the data quality improvement made the case for itself.
WORKFLOW 10: Pipeline insights pushed to Slack
Shared by: Jarrad Walter, Honorlock
Gong feature: AI Trackers + Gong Streams + Slack integration
What it does: Sends real-time Slack alerts when important signals appear in calls — leadership gets intelligence without reviewing calls.
Step by step:
1. Identify which tracker signals matter most to leadership — competitive mentions, negative sentiment, deal risk
2. Create or verify AI Trackers for each signal
3. Set up Gong Streams subscribed to those trackers — one stream per audience
4. Connect your Streams to Slack using the Gong Slack integration
5. Configure the right channel for each stream — #exec-deal-alerts, #sales-managers, etc.
6. Test with a real call and confirm the alert fires correctly
7. Tune tracker sensitivity after the first two weeks
The outcome: Leaders get deal intelligence delivered to them instead of having to find it. Signals that used to be missed are now visible in real time.
Tip: Keep Slack channels focused. Too many notifications and people stop reading them. Start with one high-signal tracker per channel.