Qualify inbound demo requests with AI and create HubSpot deals only for in target leads

Step by step

1. 1

   Create a Flow triggered by a new demo form submission

   Flow

   In Customerly, go to AI Studio → Flows → New Flow. Name it Demo qualification. Set the trigger to Customer Activity on Page and select the client event your site fires when the demo form is submitted (for example demo_requested). If demo requests come in through the Messenger instead, use New Conversation Started and narrow the audience to visitors on your demo page.

   The Flow runs the second the lead submits. There's no SDR triage in between.

   textFlow name

   Copy

   Demo qualification
   

2. 2

   Add an AI Mission that runs the qualifying conversation with the lead

   AI Mission

   Add an AI Mission step. Unlike a static "fill this form" qualifier, the Mission talks to the lead in chat, asks the right questions in order, handles follow ups, and produces a structured qualification verdict. This is the step BRUM used to convert 18.5% of qualified chats into paying customers.

   Paste the prompt below. Aura will run the conversation and emit a JSON verdict the rest of the Flow branches on.

   promptAI Mission prompt

   Copy

   You are an SDR for Customerly, an AI customer service platform for SaaS teams.
   
   Your job is to qualify this inbound demo request through a short conversation. Be warm, conversational, and brief. Don't sound like a form.
   
   Ask, in this order, only the questions you don't already know the answer to from the form submission:
   
   1. What is the size of your customer support team today?
   2. What tool are you using for customer support right now?
   3. What is the main thing you want AI to help with (resolution, qualification, routing, knowledge base, other)?
   4. What is the timeline (this quarter, next quarter, just exploring)?
   
   Stop asking the moment you have a complete picture. Don't push past four questions.
   
   Then decide if the lead is IN TARGET for Customerly. The ICP is:
   
   - SaaS company (B2B or B2C SaaS)
   - 5+ person support, sales, or success team OR clear plans to scale support within 6 months
   - Currently using Intercom, Zendesk, Help Scout, Front, HubSpot Service, or has no tool yet
   - Has a real timeline (this quarter or next quarter)
   
   NOT in target:
   
   - Students, individual operators with no team
   - Direct competitors (Intercom, Zendesk, Decagon, Ada, Forethought employees)
   - Agencies pitching their services
   - "Just exploring" with no timeline AND no team size
   
   Output exactly this JSON and nothing else:
   
   {
     "verdict": "in_target | not_in_target",
     "team_size": "<answer to question 1>",
     "current_tool": "<answer to question 2>",
     "primary_use_case": "<answer to question 3>",
     "timeline": "<answer to question 4>",
     "summary": "<two sentence brief for the AE: who they are, what they need, why now>",
     "reason_not_in_target": "<one sentence, only if verdict is not_in_target>"
   }
   

3. 3

   Branch on the verdict and stop wrong fit leads here

   Flow

   Add a Conditional step right after the Mission. Read verdict:

   • in_target → continue to the HubSpot + AE notification steps below.
   • not_in_target → skip to the polite close step further down. No deal created. No human time burned.
   textConditional rule

   Copy

   {{mission.output.verdict}} == "in_target"
   

4. 4

   Create a HubSpot deal pre populated with the qualification context

   REST API block

   Add a REST API step on the in_target branch. Create a HubSpot deal so the AE walks into the meeting with the qualification summary already in the CRM, not living in a separate chat log.

   You'll need:

   • A HubSpot private app token with crm.objects.deals.write and crm.objects.contacts.write scopes
   • The owner ID of the AE assigned to inbound demos (or use a round robin Mission to pick the right one)
   • The pipeline ID and dealstage ID for your inbound new business pipeline
   httpHTTP method + endpoint

   Copy

   POST https://api.hubapi.com/crm/v3/objects/deals
   

   httpRequest headers

   Copy

   Authorization: Bearer pat-eu1-XXXXXXXXX
   Content-Type: application/json
   

   jsonRequest body

   Copy

   {
     "properties": {
       "dealname": "Demo: {{company.name}} ({{contact.email}})",
       "pipeline": "default",
       "dealstage": "presentationscheduled",
       "hubspot_owner_id": "12345678",
       "amount": "0",
       "description": "{{mission.output.summary}}\n\nTeam size: {{mission.output.team_size}}\nCurrent tool: {{mission.output.current_tool}}\nUse case: {{mission.output.primary_use_case}}\nTimeline: {{mission.output.timeline}}\n\nConversation: {{conversation._url}}",
       "lead_source": "Customerly demo form"
     },
     "associations": [
       {
         "to": { "id": "{{contact.hubspot_contact_id}}" },
         "types": [{ "associationCategory": "HUBSPOT_DEFINED", "associationTypeId": 3 }]
       }
     ]
   }
   

5. 5

   Ping the AE in Slack with the qualification summary

   REST API block

   Add a second REST API step on the in_target branch. Slack alert to the AE with the conversation summary and a direct link to take over the chat. The AE jumps in with the lead's context already in hand.

   httpHTTP method + endpoint

   Copy

   POST https://hooks.slack.com/services/T0000/B0000/XXXXXXXXX
   

   jsonRequest body

   Copy

   {
     "text": "Qualified demo: {{company.name}}",
     "blocks": [
       { "type": "header", "text": { "type": "plain_text", "text": "Qualified inbound: {{company.name}}" } },
       { "type": "section", "text": { "type": "mrkdwn", "text": "{{mission.output.summary}}" } },
       { "type": "section", "fields": [
           { "type": "mrkdwn", "text": "*Team size*\n{{mission.output.team_size}}" },
           { "type": "mrkdwn", "text": "*Current tool*\n{{mission.output.current_tool}}" },
           { "type": "mrkdwn", "text": "*Use case*\n{{mission.output.primary_use_case}}" },
           { "type": "mrkdwn", "text": "*Timeline*\n{{mission.output.timeline}}" }
         ]
       },
       { "type": "actions", "elements": [
           { "type": "button", "text": { "type": "plain_text", "text": "Take over the chat" }, "url": "{{conversation._url}}" }
         ]
       }
     ]
   }
   

6. 6

   Politely close the conversation for not in target leads

   Aura

   On the not_in_target branch, add an Aura reply step (or a static Send Message step if you want exact wording). Aura politely tells the lead this isn't the right fit and offers a self serve path. No SDR time spent. No lead ghosted.

   textSend Message template (not_in_target branch)

   Copy

   Thanks so much for the interest in Customerly! Based on what you shared, we don't think we're the right fit for your team right now ({{mission.output.reason_not_in_target}}).
   
   If you'd like to explore on your own, our free trial is open and self serve here: https://app.customerly.io/signup
   
   And we'll be here whenever the timing is right.
   

More AI Recipes