TL;DR: Clearbit charges $1.50–$3 per enrichment lookup, which at 5,000 leads/month is $7,500–$15,000. A weekend-built workflow on free or low-cost data (Apollo free tier, company website scrape, LinkedIn lookup) plus an LLM costs roughly $50–$200/month at the same volume. The architecture: form fill triggers a workflow, fetches company and person signals from 3–4 sources, an LLM normalizes and scores against your ICP, output writes to CRM. Stack: n8n or a small Next.js worker, Postgres, OpenAI or Claude, one paid data source for fallback. Build time: 1–2 days for v1.
Clearbit charges $1.50–$3 per enrichment lookup. At 5,000 leads a month, that's $7,500–$15,000. A custom enrichment workflow built on free or low-cost data sources plus an LLM costs roughly $50–$200/month at the same volume.
The build is a weekend's worth of work. The savings compound monthly. Here's the architecture, the stack, and the cost trade-off.
What "enrichment" actually means
A lead comes in with limited info — usually name, email, maybe company. Enrichment fills in:
- Company name, size, industry, funding stage.
- Person's role, seniority, LinkedIn profile.
- Domain info: site age, tech stack, employee count.
- Recent activity: hiring signals, news, social posts.
Enrichment quality decides whether your lead routing and scoring works. Bad enrichment = bad routing = wasted rep time.
The buy vs build math
| Approach | Per-lead cost (5k/mo) | Setup cost | Maintenance |
|---|---|---|---|
| Clearbit Reveal | $1.50–$3.00 | $0 | None |
| Apollo Pro | ~$0.50 | $0 | Light |
| Hunter + LinkedIn enrichment via third party | $0.20 | A few hours | Light |
| Custom build on free APIs + LLM | $0.01–$0.05 | 1–2 weekends | Light to moderate |
The "build" approach trades up-front time for compounding monthly savings.
The stack
A custom enrichment workflow needs four pieces.
1. Domain and company info. Free options:
- Hunter.io free tier: domain + employee count.
- Apollo free tier: 60 lookups/month.
- BuiltWith free API: tech stack.
- Direct fetch of the company's homepage (free) for context.
2. Person info. Trickier without paying.
- LinkedIn enrichment requires a paid third party (PhantomBuster, RocketReach) or risky scraping.
- For B2B, the email format reveals a lot.
firstname.lastname@domain.complus a name = high probability profile. - Email validators (Hunter Email Verifier) confirm the address is real.
3. LLM for synthesis. This is the magic.
- Pull all the raw info above.
- Feed it to GPT-4o or Claude Sonnet 4 with a structured prompt.
- Get back: company description, ICP fit score, suggested talking points, pre-call brief.
- LLM cost: ~$0.01–$0.03 per lead.
4. CRM write-out. Map the enriched fields to your HubSpot, Pipedrive, or Notion record.
The architecture
A simple webhook-driven workflow:
Form submission
↓
Webhook to your API
↓
Pull domain info (Hunter, BuiltWith)
↓
Pull person info (LinkedIn lookup or email pattern)
↓
Fetch company homepage (cheerio or simple fetch)
↓
LLM call with all the context
↓
Structured JSON response
↓
Write to CRM, post to Slack
This runs on a $20/mo Vercel deployment or a $5 VPS with n8n.
A working code example
Here's the LLM step in TypeScript, the part that does the actual synthesis:
import Anthropic from "@anthropic-ai/sdk";
const anthropic = new Anthropic();
async function enrichLead(rawData: {
name: string;
email: string;
company: string;
companyInfo: string;
homepageContent: string;
}) {
const response = await anthropic.messages.create({
model: "claude-sonnet-4-6",
max_tokens: 1024,
messages: [
{
role: "user",
content: `Enrich this lead and return strict JSON.
Lead:
Name: ${rawData.name}
Email: ${rawData.email}
Company: ${rawData.company}
Company info: ${rawData.companyInfo}
Homepage extract: ${rawData.homepageContent}
Return JSON with these fields:
{
"company_summary": "1 sentence",
"industry": "specific category",
"company_size_estimate": "string",
"icp_fit_score": 0-10,
"icp_reasoning": "1-2 sentences",
"talking_points": ["string", "string", "string"],
"next_action": "string"
}`,
},
],
});
const text =
response.content[0].type === "text" ? response.content[0].text : "";
return JSON.parse(text);
}
That's the core function. The rest is plumbing — webhook handling, error retry, CRM writes — and an afternoon of work each.
What to budget for
Time:
- Discovery and ICP definition: 2 hours. Most of the build's quality depends on a clear ICP scoring rubric.
- API integrations and code: 4–6 hours.
- Prompt iteration: 2–4 hours. The first prompt is never the right one.
- CRM integration: 2–4 hours, depending on the CRM.
- Testing on 50–100 sample leads: 2 hours.
- Documentation and handoff: 1 hour.
Total: 12–18 hours. A weekend or a long week of evenings.
Money:
- Hosting: $20/mo Vercel or $5 VPS.
- Anthropic or OpenAI API: $20–$100/mo at typical service-business volume.
- Hunter.io free tier: $0–$49/mo if you upgrade.
- Total: $40–$170/mo for 5,000 leads.
Compared to Clearbit at $7,500/mo for the same volume.
When to skip the build
Three cases where it's not worth it.
1. Volume under 500 leads/month. Clearbit at $1.50 × 500 = $750. Apollo at $99/mo. Either is cheaper than building, deploying, and maintaining your own. Stay on the off-the-shelf option until volume justifies the build.
2. You don't have a developer. This needs someone who can write code, debug API responses, and tune a prompt. Without that, the off-the-shelf vendors are correct.
3. You need 100% data accuracy on B2B persona info. Clearbit's verified data is better than anything you'll synthesize from public sources. For specific compliance or sales use cases, the paid option is the right answer.
For the rest of the cases — which is most service businesses and most early-stage SaaS — the build wins on cost, speed, and customizability.
What we ship for clients
A typical Webdimonia enrichment build:
- Architecture review and ICP scoring rubric: included.
- Webhook + enrichment pipeline: $2k–$5k.
- LLM prompt engineering and iteration: $1k–$2k.
- CRM and Slack integration: $500–$1.5k.
- Internal dashboard: $500–$1.5k (optional).
- Documentation, Loom handoff, 30 days post-launch: included.
Total: $3.5k–$8k for a working enrichment system that costs $50–$200/mo to run instead of $5k–$15k.
Three questions to decide this week
- What's your monthly lead volume, and how much is rep time on each lead? If the answer is "more than 50 hours of rep time," enrichment automation pays back.
- What's your ICP definition? If you don't have one, fix that first. Enrichment without ICP is data without judgment.
- Do you have a developer (or budget for one)? If yes, build. If no, Apollo or Clearbit at small scale.
If you want a quote on a custom enrichment build tuned to your ICP and CRM, send us your form fields, your CRM, and a rough lead-volume number. We send a tiered proposal back within two days.