---
title: Case study: a lead engine with 50,000 qualified leads
description: A web3 growth firm needed opinion leaders, traders and ambassadors. The lead engine we built found 50,000 qualified leads in eight weeks.
url: https://www.streamlight.ai/case-studies/web3-lead-engine/
updated: 2026-10-23
---

# A lead engine that found 50,000 qualified leads in eight weeks.

A growth firm needed the right people in an anonymous market: key opinion leaders, traders, power users and ambassadors. We built them a lead engine that finds, checks and matches those people across X, Telegram and web3 protocols, with a person approving every contact.

**50,000** qualified leads

**8 weeks** from start to 50,000

**4 weeks** to a working first version

How it works

A source Runs on its own A person decides The work, moving Bots and fakes, removed A message, drafted

**What it learns about each person** What they post Who they follow Their projects Their ties Signals: raising, promoting

We gather from X, Telegram and web3 protocols; enrich every profile (what they post, who they follow, their projects, their ties, signals such as raising or promoting) and connect them; check engagement and legitimacy, where bots and fakes drop out; match each person to an ICP; draft a message for each; and a person approves before anything is sent.

Business

Growth firm working with web3 projects

What we built

A custom lead engine

Connected to

X, Telegram and web3 protocols

Timeline

First version in 4 weeks, full analysis in 8

Status

Running, and still being extended

[Get your free prototype ›](https://www.streamlight.ai/book/) [or see it run in the console ›](https://www.streamlight.ai/#tour=consult)

The problem

Web3 is anonymous by design. The people a growth firm needs to reach aren’t in any lead database. They’re spread across X, Telegram and the protocols themselves, under handles and wallet addresses rather than names. Each of the firm’s clients needs different people, and telling them apart means knowing what they post about, who they’re connected to, and whether they’re real. No off-the-shelf tool does that, and doing it by hand doesn’t scale.

What we built

1.  01 **Gathering by client profile**

    It searches X, Telegram and web3 protocols for the people each client profile describes, and keeps going on its own.

2.  02 **Enrichment**

    For each person: what they post about, which projects they follow and promote, and the wallets linked to them.

3.  03 **Connection mapping**

    Who knows whom, and which projects each person is tied to.

4.  04 **Influence and legitimacy checks**

    Whether their reach is real, and whether the account is who it says it is.

5.  05 **Matching**

    AI scores every person against every client profile.

6.  06 **A person approves**

    Nobody is contacted until someone on the team says yes.

How long it took

A working first version in four weeks: gathering, enrichment and matching for the first client profiles. Connection mapping and the influence and legitimacy checks followed in the next four. We’re still extending it, with richer data views and deeper analysis.

What changed

50,000 qualified leads in about eight weeks, in a niche no data provider covers. Each campaign now starts from a list of the right people, with what they care about already known.

> Streamlight helped us rethink how the work that brings in revenue actually runs, and showed us opportunities we hadn’t seen. Within a few weeks we had automated lead generation and outreach running, at a very low cost to run. We’re already on to the next piece.

**Founder** Growth agency, 8 people

[Next case study Product usage in the CRM, so every customer gets the right campaign. Software company, 20 people ›](https://www.streamlight.ai/case-studies/web3-lead-engine/#crm)

Described by category at the client’s request; figures as reported by the client.