How to Create an AI Sales Coach Using CustomerIQ and Zapier

Sales leaders know that coaching wins deals—but reviewing every sales meeting manually just doesn’t scale. That’s where an AI Sales Call Analysis Agent comes in.

In this guide, we’ll show you how to build an AI-powered agent using CustomerIQ and Zapier that automatically analyzes your Zoom, Google Meet, or Microsoft Teams calls, identifies key insights, and shares them with your team in real time.

What You’ll Build

You’ll set up an automation that:

1. Captures and transcribes a virtual sales meeting (Zoom, Meet, or Teams)
2. Sends the transcript to Zapier via webhook
3. Uses OpenAI (ChatGPT) to identify key insights and areas of improvement
4. Send the analysis to tools like Slack, HubSpot, or email

What You’ll Need

• A CustomerIQ account with meeting assistant enabled
• Access to the CustomerIQ Webhooks feature (Admin access)
• A Zapier account
• An OpenAI platform account (can also swap out for Anthropic's Claude or other LLMs)
• This AI call coaching prompt

Step-by-Step Setup

Step 1: Enable Meeting Capture in CustomerIQ

First, make sure CustomerIQ is set up to record your virtual meetings.

• Connect your calendar (Google or Outlook)
• Make sure your assistant is set to join your next meeting.

Once a meeting is processed, CustomerIQ emits a webhook event:
meeting.transcript.completed

‍

Step 2: Create a Webhook in CustomerIQ

To push the transcript to Zapier:

1. Navigate to Settings > Apps & Integrations > Webhooks
2. Go to“Add New Webhook”
3. Set the Webhook URL to your Zapier webhook endpoint (you’ll generate this in the next step)
4. Choose the event type: meeting.transcript.completed

‍

Step 3: Set Up a Zapier Webhook to Receive the Data

1. In Zapier, create a new Zap
2. Set the Trigger to: Webhooks by Zapier → Catch Hook
3. Copy the unique webhook URL Zapier provides
4. Paste this URL into your CustomerIQ webhook setup
5. Run a test meeting to trigger the webhook

CustomerIQ will send a JSON payload like this:

{  "title": "Sales Call",  "transcript": "[00:00:02] Sarah (Acme): Thanks for jumping on today...",  "submission_url": "https://app.getcustomeriq.com/folders/.../submissions/...",  "account_name": "Acme Inc",  "contact_email": "rep@acme.com"}

‍

Step 4: Send a test payload to Zapier

From your new webhook:

1. Click on "Actions"
2. Send a test webhook to Zapier

This will not only test your connection, it will also provide you with fields to map the rest of the workflow

‍

Step 5: Analyze the Transcript with OpenAI

From Zapier, add a step after recieving the webhook to analyze the call

1. Add the Zap: ChatGPT (OpenAI) > Conversation
2. Connect your account
3. Select "Show advanced options"
4. In the assistant message paste the following prompt
5. In the user message, map the transcript from the CustomerIQ webhook

Assistant prompt:

You are “CallCoach‑GPT”, trained on all sales best‑practice research for discovery and demo calls.

OBJECTIVE  
Evaluate the following transcript and produce a concise coaching report that:

1. **Benchmarks Key Metrics**  
   • Talk‑time (% rep vs. buyer)  
   • # of questions (total & open‑ended)  
   • # of distinct pains explored  
   • Speaker switches per minute  
   • % of time spent on Next Steps / Commitment  

2. **Assesses Quality Against Best Practice**  
   • Agenda & expectation‑setting  
   • Depth of pain & impact discovery (SPICED elements)  
   • Alignment between discovery findings and demo content (if demo)  
   • Use of buyer‑centric storytelling vs. feature dump  
   • Clarity and strength of agreed Next Steps  

3. **Highlights Strengths** (bullet list)  
4. **Identifies Improvement Opportunities** (bullet list, ordered by impact)  
5. **Prescribes Actionable Coaching Tips**  
   • Give specific wording or behavioral tactics for next call  
   • Reference exact time‑stamped examples (e.g., “12:45 – missed follow‑up question on budget impact.”)

FORMAT  
Return Slack's lightweight markup language called mrkdwn with the following sections:

Executive Snapshot  
(2‑3 sentences)

Metric Benchmarks  
...

Strengths  
…

Improvement Opportunities  
…

Coaching Recommendations  
…

GUIDELINES  
- If data (e.g., talk time) cannot be reliably extracted, estimate and note as “Approx.”  
- Use clear, direct language; no fluff.  
- Anchor every critique or praise in observable call moments.
- Respond in Slack's lightweight markup language called mrkdwn

‍

Step 6: Route the Insights to the Right Tools

You can customize the Zap to deliver insights to:

Slack

• Action: Slack → Send Channel Message
• Content: “New Sales Meeting Summary for {{account_name}}: {{formatted summary}}”

HubSpot or Salesforce

• Action: CRM → Create Note or Task
• Use submission_url to link back to the full call inside CustomerIQ

Email

• Action: Gmail/Outlook → Send Summary to the rep or their manager

You Now Have an AI Sales Call Analysis Agent

With just a few tools—CustomerIQ, Zapier, and OpenAI—you’ve built an AI sales call agent that:

• Captures and analyzes every virtual sales meeting
• Surfaces key insights automatically
• Enables real-time coaching without manual review

Expanding the Workflow

Once the core loop is running, you can expand it to:

• Score reps using frameworks like BANT, MEDDIC, or Challenger
• Filter to only analyze trascripts from certain accounts
• Track competitor mentions across calls
• Route summaries to different managers based on territory or rep
• Aggregate call themes in CustomerIQ for reporting

Final Thoughts

AI is no longer just for transcribing sales calls—it can coach your team in real time. With CustomerIQ and Zapier, you're not just collecting data, you’re turning every meeting into a moment for improvement. This is how sales teams win more deals, faster.