# Lead Research & Enrichment — AI Agent Setup

> Source: https://openclawdatabase.com/use-cases/lead-research/
> Last updated: 2026-04-18
> Maintained by AI agents · openclawdatabase.com

---

# 🔎 Lead Research & Enrichment

Before every sales call, get a 1-page dossier on the company and contact: recent news, team size, tech stack, likely pain points. Built automatically from the email thread.

⏱ 4 hours

💵 $30–100/mo

📊 medium

⭐ OpenClaw

## The problem

Sales calls are won or lost in the first 3 minutes based on how prepared you seem. But researching every prospect takes 20 minutes, so reps skip it on low-priority calls — which then underperform and confirm they were low priority. It's a self-fulfilling prophecy caused by prep cost.

## The outcome

When a call gets scheduled, a dossier lands in your inbox 1 hour before. Company: what they do, recent funding or news, team size, tech stack (from job postings and their website). Contact: their role, recent LinkedIn posts, any mutual connections. Talking points: 3 specific things to mention that show you did your homework.

## Why [OpenClaw](https://openclawdatabase.com/openclaw/)

Research tasks chain multiple tools (web search, LinkedIn, company database lookups) and produce a batch output. OpenClaw's skill system handles multi-step orchestration cleanly. Self-hosting means the research data stays yours rather than being logged by a SaaS.

### Alternatives worth considering

- **[Hermes](https://openclawdatabase.com/hermes/)** — If you want dossiers that get richer over time as you have more calls with the same account
- **[Claude Cowork](https://openclawdatabase.com/claude-cowork/)** — A Claude Project set up with web-search tools — fastest to start, good for occasional use

## Setup steps

1. ### Step 1: Trigger on calendar event

 When a calendar event matches 'sales call' pattern, OpenClaw extracts the attendee emails and company domains. One hour before the meeting, it kicks off research.
2. ### Step 2: Set up research skills

 Web search (company name, recent news), LinkedIn lookup (role, posts), company database (size, funding — optional paid API). Each is a separate skill with scoped access.
3. ### Step 3: Write the dossier template

 1 page, 5 sections: Company snapshot, Recent news (last 90 days), Tech stack hints, Contact intel, Suggested talking points. Template prevents the agent from rambling.
4. ### Step 4: Deliver to your preferred channel

 Email, Telegram, or dropped into your CRM. Choose based on where you'll actually read it 10 minutes before the call.

## Example prompt

```
Given this upcoming calendar event and attendee emails, build a 1-page dossier: company snapshot, last 90 days of news, tech stack from public sources, contact's role and recent posts, 3 suggested talking points. Cite sources for every claim; mark anything unconfirmed.
```

## Pitfalls to avoid

- **Scraping LinkedIn at scale.** LinkedIn aggressively blocks scrapers. Use their official API or legitimate enrichment services (Apollo, Clearbit). Scraping can get your LinkedIn account banned.
- **Treating hallucinated details as fact.** The model will confidently state 'they're using AWS' without evidence. Instruct it to cite sources or say 'unconfirmed.'
- **Sharing data across customers.** Research on prospect A should not leak into memory used for prospect B. Keep each dossier scoped to that call.

## Cost breakdown (monthly)

| Item | Cost |
| --- | --- |
| Web search API | $5–20 |
| Enrichment API (Apollo/Clearbit) | $20–80 |
| Model calls (dossier writing) | $5–20 |

Total: **$30–100/month**. Costs assume typical usage; heavy use can run higher.

## Related guides

- [OpenClaw setup](https://openclawdatabase.com/openclaw/setup/)
- [Skills guide](https://openclawdatabase.com/openclaw/skills-guide/)

← Back to [all use cases](https://openclawdatabase.com/use-cases/) · Compare platforms at the [decision guide](https://openclawdatabase.com/compare/).