How to Build a Lead Gen Campaign With Natural Language
May 19, 2026
•Airtop Team
How Do You Build a Full Outbound Campaign Without a GTM Engineer?
You describe who you want to reach. An AI agent finds the leads, enriches them, and drafts personalized outreach. That's how Airtop Mark turns one conversation into a deployed campaign.
What you say to Mark
Open Mark and type this:
"Find 50 Director+ marketing and sales leaders at B2B SaaS companies with 50-500 employees who started their role in the last 6 months. Enrich each lead with firmographic data and LinkedIn profile. Draft a personalized cold email for each one. Save everything to a Google Sheet."
Mark builds an agent that runs that workflow end to end.
What a lead gen campaign actually requires
A working outbound campaign has three layers: targeting, enrichment, and personalization. Most teams stitch these together with Apollo or ZoomInfo for data, Clay for enrichment, Instantly for sending, and a copywriter for the emails.
Each tool has its own pricing, learning curve, and API limits.
The manual workflow
| Step | Tool | Cost/month |
|---|---|---|
| Find matching leads | Apollo, Sales Navigator, ZoomInfo | $200-$1,000 |
| Enrich with firmographic data | Clay, Clearbit | $300-$800 |
| Research for personalization | Manual LinkedIn review | Time cost |
| Write personalized emails | Copywriter or AI tool | $200-$500 |
| Load into sequence tool | Instantly, Outreach | $100-$300 |
| Total | 5-7 tools | $800-$2,600/mo |
What Mark builds from that one message
- Searches for leads matching your criteria - industry, company size, seniority, role tenure
- Enriches each lead with company data and LinkedIn profile
- Scores against your ICP so you prioritize best-fit prospects
- Drafts personalized outreach using each lead's career context
- Outputs to Google Sheet with all data in structured columns
- Runs again on demand - same agent, new targets
Other things you can say to Mark
"Build a campaign targeting VP Engineering at fintech companies that use React and raised Series A or B"
"Find 100 heads of growth at e-commerce companies and draft LinkedIn connection messages"
"Create a weekly agent that finds 25 new prospects every Monday and adds them to HubSpot"
"Run a campaign for CFOs at healthcare companies that appeared in recent funding news"
How Airtop agents compare to traditional lead gen stacks
| Traditional stack | LLM-based agents | Airtop compiled agents | |
|---|---|---|---|
| Setup | Days to weeks | Hours | Minutes |
| Per-lead cost | $5-$15 | $1-$3 | ~$0.01-$0.05 |
| Personalization | Template merge tags | AI but slow | AI, compiled for speed |
| Reliability | Breaks on API changes | Varies per run | Deterministic |
| Scaling | Linear cost increase | Token costs spike | Near-flat marginal cost |
Airtop agents compile your workflow into deterministic code. The AI reasons once to build the logic. Then the agent executes like software.
Common mistakes when building lead gen campaigns
Targeting too broadly kills response rates. The best campaigns have three to five specific criteria. "VP Marketing at B2B SaaS with 50-200 employees who changed roles in the last 90 days" outperforms "marketing leaders at tech companies" every time.
Skipping enrichment makes personalization hollow. Real personalization requires company context and person context. A name and email is not enough.
Sending volume before testing wastes leads. Start with 10-20 leads. Test the messaging. Then scale.
Not iterating on the ICP. Your first targeting criteria are a hypothesis. The people who reply tell you more about your market than any report.
Cost and scaling
A traditional stack costs $800-$2,600/month in tools. At 50 leads per campaign, that is $16-$52 per lead.
Airtop agents run ~10x faster and ~100x cheaper. Processing 1,000 leads costs roughly $10-$20. The agent compiles once and executes like software.
FAQ
How many leads can an Airtop lead gen agent process in one run?
No hard limit. Agents routinely process hundreds or thousands of leads. Throughput depends on enrichment sources and rate limits on data providers.
What data sources does Airtop use for lead generation?
Airtop aggregates data from multiple B2B databases and public web sources including LinkedIn profiles, company websites, and funding databases. No separate subscriptions required.
Can an Airtop lead gen agent write results to HubSpot or Salesforce?
Yes. Specify the destination in your prompt. Agents write to HubSpot, Salesforce, Google Sheets, and other connected tools.
How does Airtop personalize outreach at scale?
The agent researches each lead individually - LinkedIn profile, career history, company news - then drafts outreach based on that context. Each email is unique.
Can I reuse a lead gen agent with different targeting criteria?
Yes. Re-run with different criteria or duplicate and adjust for a new ICP segment. The compiled workflow persists.
How does Airtop compare to building campaigns in Clay with Apollo?
Clay requires configuring waterfall enrichment and connecting Apollo or ZoomInfo step by step. Airtop compiles the entire pipeline from a natural language description with data sources built in.
Related Agents
