Home/Methodology
Methodology

How our b2b lead generation service builds a list

Seven steps, in this order, all working from open-source data and publicly available sources. Most list problems come from skipping the first one or the last two.

  1. Define ICPwho fits
  2. Sourceopen sources
  3. Matchpeople to companies
  4. Dedupeone row per entity
  5. Verifyroles, domains, emails
  6. Sampleyou approve a slice
  7. Deliverfiles + QA note
7steps from brief to delivery
70+fields per record in the final file
1sample you approve before the full build

What happens at each step of a b2b lead list build

Define the ICP

We write down who you want to reach: industry, company size, geography, roles, and who to exclude. If it is vague here, the list will be vague everywhere.

Source

We gather candidates from open-source data and publicly available sources, using OSINT methods. Several source families feed each list, so no single one decides who is on it. See data sourcing.

Match

We connect people to the right companies and roles, and companies to their real web domains.

Dedupe

Same company under two names, same person twice, the same domain in three rows. We collapse them to one clean record each.

Verify

We check roles for staleness, domains for a match, and emails for deliverability. Details on the data quality page.

Sample

Before the full build, you get a small slice to review. You tell us what is off, we adjust the rules, then we run the rest.

Deliver

CSV or spreadsheet in your column layout, plus a short note on what we checked and what we dropped. Upload-ready audience files for ad platforms on request.

Two steps in motion

Dedupe and verify do the quiet work that decides whether a list is usable. Here is what each one does to a handful of made-up rows.

Three spellings, one company. We keep one record and note that it came from more than one source.

Each address gets a status. Rows we drop are counted in the QA note, not deleted in silence. More on data quality.

What we hold to

  • Open-source data and publicly available sources only, drawn from several families.
  • Each list is built for your brief, not pulled from a shared file.
  • You approve a sample before we build the full list.
  • Dropped rows are reported, not silently deleted.

Cost depends on list size, how narrow the ICP is, and how many fields you want. See pricing or the how it works overview.

WebsitesRegistriesPressEvents Match + dedupe Verify Your list

Questions about how we build a list

Do I approve anything before the full list is built?

Yes. You approve the ICP before we source anything, and you review a small sample before the full build. See how it works for the full engagement.

What if the sample is not right?

You tell us what is off, we adjust the rules and send a new sample. Fixing the rules at this stage is cheaper than fixing a finished list.

Which sources feed a list?

Open-source data and publicly available sources, gathered with OSINT methods. Several source families feed each list. The details are on data sourcing page.

Want your list built this way?

Send us your ICP, even a rough one. We reply with a scope and a sample plan.