Clay + GPT inbound enrichment (L3)
Every inbound demo request is enriched with funding stage, hiring signals, tech stack, and routed with a one-page brief to the AE. This is where most teams should aim first.
The steps
- 01
Connect form to Clay via Webhook
The process begins by capturing inbound lead data from your marketing automation platform (HubSpot, Marketo, or Calendly) and pushing it to a Clay workspace. • In Clay, create a new table from a "Webhook." This generates a unique URL. • In HubSpot, go to 'Workflows' and create a 'Contact-based' workflow triggered by your demo request form submission. • Add an 'Output Webhook' action (requires Operations Hub Professional) and paste the Clay URL. Choose 'POST' and select properties like Email, Name, Website, and Company. • If you lack Hubspot Pro, use Zapier or Make.com as a bridge: Trigger (HubSpot New Form Submission) -> Action (Webhook Post to Clay). • Owner: Marketing Ops or RevOps. Time: 45 minutes. • Prerequisite: A live demo form on your site. • Pitfall: Sending 'Company Name' only; always send 'Website/Domain' as it serves as the primary key for enrichment. • Definition of Done: Submitting a test form results in a new row appearing in the Clay table within 30 seconds.
- 02
Enrich with 5 high-intent signals
Once the lead lands in Clay, use built-in providers to build a 360-degree view of the lead's company. • Click 'Add Enrichment' and search for 'Find Company' using the domain. • Add the 'Cruchbase' enrichment to pull 'Last Funding Type' and 'Total Funding Amount'. • Use 'LinkedIn' or 'Claygent' to calculate 'Headcount Growth (90 days)'. Use a formula like: (Current - 90d ago) / 90d ago. • Use 'BuiltWith' or 'PredictLeads' to find the Tech Stack. Specifically, look for competitors or complementary tools (e.g., 'Has Salesforce' or 'Uses Snowflake'). • Use 'Claygent' (AI Web Scraper) with the prompt: 'Visit the careers page of [Company Website] and identify if they are hiring for Data, AI, or Sales roles. Return 'Yes' and the job titles if found.' • Owner: Sales Ops. Time: 1.5 hours. • Pitfall: High API costs. Set 'Run Settings' to only execute when 'Email' is not empty to avoid wasting credits on junk entries. • Definition of Done: Five distinct columns populated with fresh, accurate company signals for every new row.
- 03
Generate the Sales Brief via LLM
Transform raw data into a readable executive summary using an LLM connector like GPT-4o or Claude 3.5 Sonnet. • Add a column 'Cloud Intelligence' -> 'ChatGPT'. • Use this detailed prompt: 'You are a world-class SDR. Based on these signals: Funding: [Funding], Tech: [Stack], Hiring: [Roles], News: [Recent News], write a 3-paragraph brief. Paragraph 1: "Why Now" (trigger event). Paragraph 2: Pain Hypothesis based on their tech stack. Paragraph 3: Three customized discovery questions. Keep it under 250 words.' • Map the variables from your enrichment columns into the prompt brackets. • Set the 'System Message' to: 'Adopt a professional, concise, and analytical tone. Avoid marketing fluff like "delighted to assist."' • Owner: RevOps. Time: 60 minutes. • Pitfall: Generic output. If the AI is too vague, give it specific "if/then" logic in the prompt (e.g., 'If they use Salesforce but not a dialer, focus on manual entry pain'). • Definition of Done: A 'Brief' column in Clay containing a high-quality, actionable summary for each lead.
- 04
Automate Routing to Slack and CRM
Deliver the insights to the AE immediately so they can prep before the lead gets cold. • In Clay, click 'Export/Integrate' and select 'Slack'. • Map the destination to a specific channel (e.g., #inbound-alerts) or use a lookup to send a Direct Message to the Lead Owner. • Construction: '🚨 New Demo Request: [Name] from [Company]. [Brief Column Content].' • Next, add a 'HubSpot' or 'Salesforce' action to 'Update Company/Opportunity'. Map the 'LLM Brief' column to a custom long-text field called 'AI_Sales_Brief__c'. • This ensures the AE sees the data both in the tool they live in (Slack) and the system of record (CRM). • Owner: RevOps. Time: 45 minutes. • Pitfall: Routing to a generic channel where it gets ignored. Use the 'Lookup AE' feature in Clay to route specifically to the person assigned to the meeting. • Definition of Done: A Slack notification is triggered automatically with the brief, and the lead record in the CRM shows the same brief.
- 05
Measure conversion and refine prompt
Validation is critical to prove the ROI of the AI enrichment stack. • Establish a 'Control' period or group. Use the 30 days prior to implementation as your baseline for 'Inbound-to-Meeting' conversion rate and 'Discovery-to-Opportunity' rate. • In your CRM, create a report filtering for leads where 'AI_Sales_Brief__c' is 'Is Not Null'. • Compare the 'Time to First Touch' (Time from Webhook to First AE Email/Call) for enriched leads vs. non-enriched. • Track ACV (Average Contract Value). Hypothetically, AEs with better intel should be able to hunt larger deal sizes within the same account. • Review the 'LLM Briefs' weekly with the AEs to refine the prompt. If they find the 'Discovery Questions' useless, update the prompt in Step 3. • Owner: Sales Leadership / RevOps. Time: Ongoing (Review every 30 days). • Pitfall: Forgetting to track 'No-Show' rates. Sometimes better prep leads to more aggressive outreach that may increase or decrease attendance. • Definition of Done: A dashboard showing a clear delta in conversion rates between enriched leads and the historical baseline.
Next playbooks
Unfamiliar terms are defined in the AI and Revenue Dictionary. Related frameworks live in the framework library.