Where our data comes from
Every provider OmniLead reads from, what each one supplies, how the crawler behaves, and the sources OmniLead will never use.
OmniLead only uses public, verifiable sources. Every value it shows you, whether an email, a phone number, a fund size or a grant deadline, keeps a record of the source it came from and the date it was found.
This page lists every source, what it's used for, and the sources OmniLead refuses to use.
Free public sources
These work in every workspace with no setup. They're official, public datasets and APIs.
The Leads sources for local businesses and licensed professionals have their own rules, such as which people are never collected. See Free public sources for local businesses.
How Form D is used
Form D is a notice companies and funds file when they raise money under Regulation D. OmniLead reads it this way:
- Investors are funds. Only issuers whose Item 4 industry group is Pooled Investment Fund are listed as investors, typed by the fund type in the filing: venture capital, private equity, hedge fund or other. A company raising money files a Form D too, and is never shown as an investor.
- People are related persons. The executive officers, directors and promoters a fund's filing names become person records with their role, linked to the fund. The filing's addresses for them aren't stored, and their contact details come only from the fund's own website.
- Sector is a name keyword. Form D doesn't state what a fund invests in, so a sector filter matches words in the fund's name, and is labeled that way.
- Fund size is the offering amount. For a fund, the total offering amount is the size of the fund being raised. Form D has no stage and no check size.
- Activity counts are what companies report. The investors-by-state pages count new Form D offerings by companies in a sector's industry groups (funds excluded) and add up the number of investors those filings say have already invested (Item 14) and the amount sold (Item 13). Form D gives that number, never the investors' names, so an investor in two offerings counts twice. In busy states the page reads the most recent filings in full and says how many; the offering count comes from EDGAR full-text search, and the page links to the same search on sec.gov.
- Fund lists are funds only. The "funds based in" list on those pages shows venture capital and private equity funds by the state of their business address, with the fund type, size and filing link from the filing. No personal names or contact details are published there.
Company websites
For the Leads and Investors lenses, OmniLead reads organizations' own websites with a crawler built to be a good guest.
- It respects
robots.txt. If a site disallows crawling, OmniLead doesn't crawl it. - It identifies itself honestly. The user-agent names OmniLead and links to a page explaining what it does.
- It's rate-limited per domain, so it never puts noticeable load on a site.
- It works within a page budget. It reads the pages where organizations publish contact information: home, about, team, contact, impressum and the footer. It doesn't spider a whole site.
- It never logs in, never solves CAPTCHAs and never rotates proxies to get around a block.
From those pages it extracts published emails (including common obfuscations such as "name [at] company [dot] com"), phone numbers, social profile links, and team members' names and titles. Each value is stored with the exact page URL it came from.
Email patterns
When a person's email isn't published but other addresses on the same domain are, OmniLead learns the pattern the organization uses, such as first.last@ or f.last@. It generates the likely addresses for that person, ranks them by how well they match the real pattern, and verifies the best candidate. The provenance says Ranked from domain pattern and lists the real addresses the pattern came from.
Sources that need an API key
These providers are paid or licensed. They work when an admin adds a key in Settings → Workspace.
Without a key, these providers show Connect provider and return nothing. OmniLead never substitutes sample or made-up data.
How results are combined
All providers share one interface. When you search:
- Each relevant provider returns candidates.
- OmniLead normalizes them to one format (for example, phone numbers to E.164 and names to a consistent case).
- Duplicates are merged. The same fund from Form D and from its website becomes one profile, with each field keeping its own source.
- Results are cached, so repeat lookups are fast and don't hit the source again unnecessarily.
Sources we never use
- No language models for data. OmniLead never asks an AI model to guess an email, invent a person or describe an investor. The optional AI drafting feature only helps write email copy.
- No bought lists of unknown origin.
- No personal or special-category data: no health, religion, ethnicity, political opinion or similar data, and no home addresses shown as contact details.
- No logged-in content: nothing behind a sign-in, a paywall or a CAPTCHA.
People we don't list
OmniLead doesn't collect or show records about individuals located in California, Vermont, Texas or Oregon. Those states require companies that compile data about their residents to register as data brokers, and OmniLead has chosen not to hold that data instead. A person counts as located there when their record, or their organization's record when theirs has no state, places them in one of those states.
Businesses and organizations in those states are listed as usual, with their published business contacts. Contacts you import yourself are your own data and aren't affected.
Removing data
Anyone can remove their own data from OmniLead through the public opt-out portal at /privacy/opt-out, after verifying their email. The removal applies across every workspace, and the address is added to a global suppression list so it isn't collected again. The portal also removes a single phone number the same way.