Automate Lead Enrichment Using AI and CRM Integrations

# Lead Enrichment Workflow: HR Analytics SaaS

## 1. Workflow Overview
This workflow ingests leads from HubSpot forms and Typeform, enriches them with firmographic and persona data via Apollo and Clearbit, uses an LLM to summarize ICP fit and generate AE-ready talking points, scores them against your mid-market HR ICP, and routes hot leads to AEs in HubSpot — eliminating the manual LinkedIn/Crunchbase research step.

## 2. Tools & Integrations
- **Lead capture:** HubSpot Forms, Typeform
- **Orchestration:** n8n (self-hosted) or Make.com
- **Enrichment APIs:** Clearbit Enrichment (company), Apollo.io (person + intent), BuiltWith (tech stack)
- **AI layer:** OpenAI GPT-4o (or Claude Sonnet) via API
- **CRM:** HubSpot (Contacts + Companies + Deals)
- **Notifications:** Slack (#sales-hot-leads)
- **Dedup utility:** HubSpot Search API + n8n "Merge" node

## 3. Step-by-Step Workflow

1. **Trigger:** n8n webhook listens to HubSpot Form Submission and Typeform `form_response` events. Normalize payload into `{email, first_name, last_name, company, raw_title}`.
2. **Dedup check:** Call HubSpot Contacts Search API by email. If contact exists and `lifecyclestage = customer`, exit. Otherwise upsert.
3. **Person enrichment:** Apollo `people/match` by email → returns title, seniority, LinkedIn URL, department, location.
4. **Company enrichment:** Clearbit `/companies/find` by email domain → returns employee count, industry, HQ country, funding, domain. Fallback to Apollo Organizations if Clearbit returns 404.
5. **Tech stack lookup (optional):** BuiltWith domain query for HRIS signals (Workday, BambooHR, Rippling).
6. **AI enrichment:** Pass merged JSON to GPT-4o with the prompt below.

```
You are an SDR research analyst for an HR analytics SaaS sold to mid-market companies (200–2000 employees) in the US/Canada to HR or People Ops leaders.

Given this enriched lead JSON:
{{lead_json}}

Return STRICT JSON with these keys:
- icp_fit_score: integer 0-100
- icp_fit_reason: 1 sentence
- persona_match: "primary" | "secondary" | "none"
- company_summary: <=40 words on what they do
- pain_hypothesis: 1 sentence on the likely HR analytics pain given size, industry, and HRIS
- opening_line: 1 personalized sentence an AE can use in outreach
- disqualify_flag: true if employees<200, employees>2000, country not US/CA, or title unrelated to HR/People

Do not include any text outside the JSON.
```

7. **Scoring & routing:** Apply rules below in n8n Function node. If `tier = A`, round-robin assign to AE pool via HubSpot Owner API and post to Slack. Tier B → nurture sequence. Tier C/disqualified → marketing list.
8. **CRM sync:** Update HubSpot Contact + Company with all enriched fields, AI outputs, and `lead_tier`. Create a Task for assigned AE with `opening_line`.

## 4. Field Mapping Table

| Source field | HubSpot property | Object |
|---|---|---|
| apollo.title | `jobtitle` | Contact |
| apollo.seniority | `seniority_level` | Contact |
| apollo.linkedin_url | `linkedin_url` | Contact |
| clearbit.metrics.employees | `numberofemployees` | Company |
| clearbit.category.industry | `industry` | Company |
| clearbit.geo.country | `country` | Company |
| builtwith.hris | `current_hris` | Company |
| ai.icp_fit_score | `icp_fit_score` | Contact |
| ai.persona_match | `persona_match` | Contact |
| ai.company_summary | `company_summary_ai` | Company |
| ai.pain_hypothesis | `pain_hypothesis_ai` | Contact |
| ai.opening_line | `ai_opening_line` | Contact |
| computed | `lead_tier` | Contact |

## 5. Lead Scoring Logic
- **Disqualify** if `disqualify_flag=true` (size, geo, or title miss).
- **Tier A (hot):** icp_fit_score ≥ 80, employees 200–2000, US/CA, title contains HR/People/Talent/CHRO/CPO.
- **Tier B (warm):** icp_fit_score 60–79 OR adjacent title (Ops, Finance) at qualifying company.
- **Tier C (cold):** score < 60 but not disqualified → newsletter nurture.
- Bonus +10 if HRIS detected; +5 if recent funding within 12 months.

## 6. Error Handling & Deduplication Tips
- Wrap each API node in n8n `Try/Catch`; on 429, exponential backoff (1s→8s, 3 retries).
- If Clearbit and Apollo both miss, route to a "Manual Review" HubSpot list instead of failing silently.
- Validate AI output with JSON schema; on parse failure, retry once with `temperature=0`, then flag.
- Dedup by **lowercased email** AND **company domain + normalized full name** to catch personal-email submitters.
- Use HubSpot's `hs_object_id` as idempotency key in the workflow run log (Postgres or Airtable) to prevent double-enrichment on webhook retries.
- Log every enrichment cost per lead to monitor unit economics — target < $0.15/lead.

# AI-Powered Lead Enrichment Workflow for HR Analytics SaaS

## Workflow Overview

This workflow automatically enriches leads from HubSpot forms and Typeform submissions using Apollo and Clearbit APIs, then applies AI-driven fit scoring against your ICP (mid-market, HR/People Ops buyers, North America). Qualified leads are auto-assigned to account executives with detailed enrichment context, eliminating manual research and reducing time-to-first-touch from days to minutes.

## Tools & Integrations

- **Lead Sources**: HubSpot Forms, Typeform
- **Enrichment APIs**: Apollo (company & person data), Clearbit (firmographic data)
- **Workflow Automation**: n8n (self-hosted) or Make (cloud)
- **AI Model**: OpenAI GPT-4 (fit assessment)
- **CRM**: HubSpot
- **Data Validation**: ZeroBounce (email verification, optional)
- **Deduplication**: HubSpot native + n8n duplicate detection

## Step-by-Step Workflow

1. **Trigger**: New lead submitted via HubSpot Form or Typeform webhook → n8n receives payload with name, email, company, and title.

2. **Email Verification**: Pass email through ZeroBounce API to confirm deliverability; stop workflow if invalid.

3. **Company Enrichment (Apollo)**: Query Apollo API with company domain extracted from email (e.g., `acme.com`). Retrieve: employee count, industry, founding year, funding stage, LinkedIn URL. If domain not found, use company name as fallback.

4. **Company Enrichment (Clearbit)**: Query Clearbit Enrichment API with company domain for additional data: revenue estimate, tech stack, employee LinkedIn URLs, location details.

5. **Person Enrichment (Apollo)**: Query Apollo with name + company to fetch person-level data: current title, years in role, LinkedIn profile, phone number, reporting structure.

6. **AI Fit Assessment**: Send enriched data to OpenAI with structured prompt (see below). Model returns: ICP fit score (0–100), fit reasoning, recommended routing, and risk flags.

```
You are a lead scoring expert for HR analytics software.

Enriched Lead Data:
- Name: {{name}}
- Title: {{title}}
- Company: {{company}}
- Employee Count: {{employee_count}}
- Industry: {{industry}}
- Location: {{company_location}}
- Revenue Estimate: {{revenue_estimate}}

ICP Criteria:
- Company Size: 200–2,000 employees
- Decision Maker Titles: Chief People Officer, VP People Operations, Head of HR, HR Director, People Operations Manager
- Location: US or Canada
- Industry Fit: Technology, Financial Services, Professional Services (exclude Healthcare, Manufacturing)

Task:
1. Score ICP fit (0–100) based on company size, title match, and geography.
2. Provide 1-line reasoning.
3. Identify any risk factors (e.g., title mismatch, company size outside range).
4. Recommend AE routing: "Enterprise" or "Mid-Market" segment.

Return JSON only:
{
  "fit_score": <number>,
  "reasoning": "<string>",
  "segment": "<string>",
  "risks": [<strings>]
}
```

7. **Scoring Rules**: Leads scoring ≥70 proceed to auto-assignment; 50–69 marked "Manual Review"; <50 marked "Low Priority."

8. **Data Mapping & CRM Sync**: Map enriched fields to HubSpot contact and company properties (see table below).

9. **Auto-Assignment**: Use HubSpot workflows to assign leads scoring ≥70 to AEs based on segment (round-robin or territory rules).

10. **Logging**: Store enrichment payload and AI response in HubSpot deal notes for audit trail.

## Field Mapping Table

| Enriched Data Source | HubSpot Contact Property | HubSpot Company Property |
|---|---|---|
| Apollo/Clearbit: name | First Name + Last Name | — |
| Apollo: title | Job Title | — |
| Apollo: company | — | Company Name |
| Clearbit: employee_count | — | Annual Revenue (proxy) |
| Apollo: employee_count | — | Number of Employees |
| Clearbit: industry | — | Industry |
| Clearbit: location | — | Company City + Country |
| Apollo: linkedin_url | LinkedIn Profile URL | — |
| OpenAI: fit_score | ICP Fit Score | — |
| OpenAI: segment | Sales Segment | — |
| ZeroBounce: valid | Email Validation Status | — |

## Lead Scoring Logic

- **≥70**: Auto-assign to AE; mark "Sales Ready"
- **50–69**: Tag "Manual Review"; notify sales manager
- **<50**: Tag "Low Priority"; add to nurture sequence
- **Bonus points** (+10): Inbound demo request (Typeform); funded company
- **Penalty** (−15): Title mismatch (e.g., "Finance"); company <200 or >2,000 employees

## Error Handling & Deduplication Tips

- **API Failures**: If Apollo/Clearbit timeout, retry up to 2×; if still failing, create contact with available data and flag for manual review.
- **Deduplication**: Check HubSpot for existing contact by email before enrichment; if found, update existing record and log in notes.
- **Bad Data**: Reject leads with invalid emails, missing company, or title=null; log to error sheet for QA review.
- **Rate Limiting**: Stagger API calls using n8n delays (200ms between requests) to respect vendor rate limits.

Of course. Here is a complete, production-ready lead enrichment workflow designed for your B2B SaaS company.

### Workflow Overview

This workflow automates the enrichment and qualification of inbound leads from HubSpot and Typeform. It uses an orchestration

# AI-Powered Lead Enrichment Workflow for HR Analytics SaaS

## 1. Workflow Overview
This workflow automates lead enrichment for new submissions from HubSpot forms and Typeform demo requests, pulling company and persona data from Apollo.io, applying AI-driven insights and ICP scoring, and syncing qualified leads to HubSpot for AE assignment. It replaces manual LinkedIn/Crunchbase research, focusing on mid-market fit (200-2000 employees, HR/People Ops titles, US/Canada). Leads score 70+ auto-assign to AEs, ensuring reps focus on high-fit opportunities.

## 2. Tools & Integrations
- **HubSpot** (CRM for triggers, updates, and AE assignment)
- **Typeform** (demo request forms; webhooks sync to HubSpot via native integration)
- **n8n** (open-source workflow automation orchestrating all steps)
- **Apollo.io** (enrichment API for company size, location, job titles)
- **OpenAI GPT-4o** (AI enrichment, persona insights, and scoring)

## 3. Step-by-Step Workflow
1. **Trigger**: Use n8n's HubSpot "New Form Submission" or "Contact Created" node (filter for lifecycle stage "Lead" or form source "Typeform Demo"). Pulls email, name, company domain from HubSpot object.
2. **Data Lookup**: Apollo.io "Enrich Contact" node (API key auth; input email/domain). Retrieves: employee_count, company_location (HQ country), current_title, industry. Outputs JSON with fields like `employee_range`, `title`, `country`.
3. **AI Enrichment**: OpenAI node processes Apollo data + raw lead info. Generates persona summary, buying signals, and custom notes.

   ```plaintext
   You are an expert ICP evaluator for a B2B HR analytics SaaS targeting mid-market (200-2000 employees) in US/Canada with HR/People Ops leaders.

   Lead data: Email={{ $json.email }}, Company={{ $json.company }}, Apollo: employee_count={{ $json.employee_count }}, title={{ $json.current_title }}, location={{ $json.company_location }}.

   Output JSON only:
   {
     "persona_summary": "1-2 sentence profile + pain points (e.g., 'HR Director at scaling firm struggling with analytics')",
     "buying_signals": ["signal1", "signal2"] (e.g., recent funding from Crunchbase if available),
     "enrichment_notes": "Key insights for rep (e.g., 'Similar to past customer X')"
   }
   ```

4. **Scoring**: OpenAI node (chained) or n8n Function node computes 0-100 score from AI output + rules (detailed in section 5). Adds `lead_score` and `assignment_recommendation` (e.g., "AE1" based on territory).
5. **CRM Sync**: HubSpot "Update Contact" node (match by email; set lifecycle to "Qualified"; assign owner if score >=70). Creates deal if score >=90. n8n schedules daily for backfills.

## 4. Field Mapping Table

| Enriched Field          | HubSpot Property          |
|-------------------------|---------------------------|
| Apollo employee_count  | Company size             |
| Apollo current_title   | Job title                |
| Apollo company_location| Company country          |
| AI persona_summary     | Lead notes               |
| AI buying_signals      | Custom field: Signals    |
| AI enrichment_notes    | Additional notes         |
| lead_score             | Lead status (numeric)    |
| assignment_recommendation | HubSpot owner (lookup ID)|

## 5. Lead Scoring Logic
- Base: +20 if US/Canada HQ; +30 if 200-2000 employees; +30 if title matches "HR|People Ops|Talent|Head of People" (regex).
- Boosters: +10 per buying signal (AI-detected); +10 industry=HR Tech/SaaS.
- Thresholds: <50=Disqualified (set lifecycle "Evangelism"); 50-69=Nurture; 70+=MQL (assign AE); 90+=SQL (create deal). Final score caps at 100.

## 6. Error Handling & Deduplication Tips
- **Deduplication**: n8n HubSpot "Get Contact by Email" node pre-enrichment; skip if exists (update only if score changed >20%).
- **Errors**: Apollo/OpenAI timeouts → n8n Retry node (3x, 5min backoff); log to Slack via webhook; fallback to raw data sync. Rate limits: Throttle to 10/min via n8n Delay.
- **Monitoring**: n8n executions dashboard + HubSpot activity logs; test with 5 sample leads today via n8n's manual trigger.

*(Word count: 528)*