n8n HubSpot Closed-Lost → OpenAI Blameless Retro Prompts to TeamRetro

Turn HubSpot “Closed-Lost” reasons into blameless retro prompts—automatically

This n8n workflow detects when a HubSpot deal moves to your Closed-Lost stage, uses OpenAI (GPT-5 Mini) to convert the loss details into neutral, blameless TeamRetro prompts, deduplicates them against recent items, and then parks the prompts for your team to review—while logging a “Queued” activity back on the deal.

What this workflow does

• Triggers in HubSpot whenever a deal’s dealstage property changes.
• Checks whether the new stage matches your configured Closed-Lost stage ID, then fetches the full deal record.
• Extracts deal fields including name, amount, loss reason, notes, and the deal URL.
• Skips deals below a configured minimum amount.
• Sends the loss reason and notes to OpenAI (GPT-5 Mini) to generate up to two blameless retro discussion prompts plus a parkable flag.
• If parkable, retrieves the last 30 days of TeamRetro parked items and deduplicates new prompts by token overlap.
• Creates new parked items in TeamRetro for non-duplicates.
• Writes a completed “Queued for win/loss retro” task to the HubSpot deal timeline including the generated prompts.

Use cases

• Automate consistent win/loss retros by turning every meaningful HubSpot loss into a discussion starter.
• Prevent retro spam using TeamRetro deduplication against recent parked items.
• Ensure blame-free retros by generating neutral, blameless prompts directly from HubSpot deal notes.

Technical details

• n8n workflow with nodes: HubSpot trigger, if, set, code, HubSpot, Sticky Note.
• Uses OpenAI (GPT-5 Mini) to generate up to two prompts and a parkable flag.
• Requires HubSpot credentials (deal read/write + engagement write) and OpenAI credentials.
• Setup includes replacing REPLACE_WITH_TEAM_ID with your TeamRetro team ID and confirming parked-items support.