Auto-aggregate Slack, Zoom, Jira product signals into Google Docs PRD updates with n8n

Multi-source product signals → standardization → AI analysis → structured PRD output to Google Doc

Who

Product managers, product ops, and teams needing continuous multi-source feedback tracking

You get

Automatically aggregate Slack, Zoom, Jira, Zendesk, Figma, customer forms, and platform analytics into structured PRD updates appended to Google Doc

You need

n8n · google-docs · slack · zoom · jira · zendesk · figma

Steps

1. Configure third-party credentials

   Create and connect the following credentials in n8n: 1. Slack API: Create an App at api.slack.com, enable app_mentions:read, channels:history, chat:write, install to workspace, copy Bot OAuth Token. 2. Zoom OAuth2: Create an OAuth App in Zoom Marketplace, add n8n's Redirect URL, copy Client ID and Secret. 3. Google Docs OAuth2: Create a Google Cloud Project, enable Google Docs API. 4. Jira Software Cloud: Generate API token from Atlassian, prepare email and domain. 5. Zendesk: Generate API token, prepare subdomain and email. 6. Figma: Generate Personal Access Token, prepare team ID. 7. OpenAI API: Replace the template's free credits credential with your own API key. Note: All sensitive information should be configured in n8n's Credentials management UI, never in node parameters or prompts.

2. Connect product signal sources

   The workflow captures signals through (enable as needed): - Slack Trigger: listens for message and app_mention events in a specified channel - Customer Form Trigger: built-in form collecting customer feedback (Name, Email, Feature Request, Priority, Additional Comments) - Schedule Trigger → Get Zoom Recordings: periodically fetches Zoom cloud recordings, transcribed via OpenAI Whisper - Schedule Trigger → Get Jira Comments: periodically pulls Jira comments - Schedule Trigger → Get Zendesk Tickets: periodically pulls open tickets tagged customer_complaints - Figma Comment Trigger: listens for comments on file updates - Webhook → Fetch Platform Analytics: receives external platform analytics data (replace <__PLACEHOLDER_VALUE__your_analytics_api_endpoint__>) Check trigger node parameters in workflow.json to ensure channelId, teamId, query conditions match your environment.

3. Standardize data formats

   Each signal source feeds into an Edit Fields (Set) node mapping raw data to unified fields: - source: signal source identifier (e.g., 'Slack', 'Zoom Recording', 'Jira Comments') - content fields: message_text, transcript, comment_text, ticket_subject, feature_request, etc. - metadata fields: user, channel, meeting_topic, issue_key, priority, status, etc. - timestamp: normalized to {{ $now }} or original time Check assignments parameters in each Format node in workflow.json. Note: workflow.json shows empty connections arrays for Format Jira Data, Format Figma Data, and Format Zendesk Data—these branches may need manual wiring to the Merge node after import.

4. Merge all signal streams

   Use the Merge node (numberInputs=4) to combine multiple standardized signal streams into one. Currently Slack, Form, Zoom, and Platform Analytics are wired to Merge; Jira, Zendesk, and Figma branches need connection verification. Merge operates in default append mode, ensuring all signals reach downstream in arrival order.

5. AI analysis generates structured PRD

   The PRD Analysis Agent node uses a LangChain Agent configured as: - Language Model: OpenAI Chat Model (gpt-4o-mini) - Output Parser: Structured PRD Output (JSON Schema constraint) - Prompt Type: define (custom prompt) The node contains two prompts: 1. User prompt (runtime expression injecting merged signal data): prefixed with = to indicate expression, passes {{ JSON.stringify($json, null, 2) }} as input, requesting feature request extraction, scope change detection, priority evaluation, target user identification, and structured PRD update proposal generation. 2. System message: defines the PRD Analysis Agent role, specifying output must include purpose, user_problem, user_value, target_users, user_feedback[], assumptions[], out_of_scope[], acceptance_criteria[], next_steps[] fields, and guides analysis of feature request priorities, scope changes, target user segment shifts, technical constraints and risks from multi-source signals.

   Execution prompt · Source text。PRD Analysis Agent node text parameter (expression mode, runtime-injected merged signals)

   Copy prompt

   =Analyze the following aggregated product signals from multiple sources:
   
   {{ JSON.stringify($json, null, 2) }}
   
   Extract feature requests, detect scope changes, evaluate priority signals, identify target user segments, and generate a structured PRD update proposal based on all the signals above.

   Execution prompt · Source text。PRD Analysis Agent node options.systemMessage parameter

   Copy prompt

   You are a PRD Analysis Agent. Analyze the aggregated product signals from multiple sources (Slack, Zoom, Jira, Figma, Salesforce, HubSpot, Zendesk, customer forms, and analytics). Extract feature requests, detect scope changes, target user changes, identify constraints and risks, evaluate priority signals, and generate structured PRD update proposals.
   
   Provide your analysis in the following structured format:
   - purpose: Clear statement of the feature/change purpose
   - user_problem: The specific problem users are facing
   - user_value: The value this provides to users
   - target_users: Who will benefit from this
   - user_feedback: Array of relevant user feedback quotes
   - assumptions: Key assumptions being made
   - out_of_scope: What is explicitly not included
   - acceptance_criteria: Array of criteria for completion
   - next_steps: Array of recommended next actions
   
   Analyze all signals for:
   - Feature requests and their priority
   - Scope changes or expansions
   - Target user segment changes
   - Technical constraints and risks
   - Priority signals from multiple sources

6. Write to Google Doc

   The Update PRD Document node uses Google Docs update → insert action to render structured output as Markdown-formatted text appended to the specified document. The expression-built text includes: PRD UPDATE timestamp, Purpose, User Problem, User Value, Target Users, User Feedback list, Assumptions list, Out of Scope list, Acceptance Criteria list, Next Steps list. You must replace the example documentURL in workflow.json (https://docs.google.com/document/d/1m-u2wUpA-J3RKir5ceeIqYnMfidbba1_dH8htNAw4QM/edit?usp=sharing) with your own PRD document URL.

You are an n8n workflow setup assistant. The user wants to replicate the "auto-aggregate multi-source product signals into PRD updates" template. First confirm they have: n8n instance access, 7 third-party accounts (Slack, Zoom, Google Cloud, Jira, Zendesk, Figma, OpenAI), and edit access to a target Google Doc. Then guide them through importing workflow.json, configuring credentials, replacing placeholders, checking node connections, and activating the workflow. If the user lacks certain signal sources, suggest disabling those trigger branches while keeping the core flow.