A sales leads list is a set of named contacts at companies that match the customer profile, with a working channel to each one. A generator is any process, manual or tooled, that turns criteria into that list. The step that separates a usable list from a spreadsheet is verification: in the BounceZero corpus, April to October 2026, only 32.6% of non-freemail (business and ISP) addresses on lead lists could be positively confirmed, 14.5% were invalid and 27% unknown. This guide builds a list in seven steps, prices each step, and shows how to segment by what the verifier returns.
Each step has an input, an output and a cost. Skipping the fourth step is how a list of 2,000 names becomes a bounced sequence and a throttled domain.
Industry codes, headcount band, geography, technologies, funding stage, and the two or three job titles that buy. Write it as filters a database can run, not as a persona. Output: a filter set. Cost: an hour.
Run the filters in a database (Apollo, ZoomInfo, Cognism, Crunchbase), a free source (Companies House, OpenCorporates, job boards, directories) or a scraper of a vertical listing. Output: 500 to 5,000 companies with domains. Cost: free to a few cents per company.
For each company, the one to three titles from step one. Database exports give names and addresses; pattern finders generate them from name plus domain; LinkedIn Sales Navigator gives the names and a finder gives the addresses. Output: a contact row per person. Cost: 5 to 50 cents per find on subscription. Mechanics in the email finder guide.
Bulk verification returns valid, invalid, catch-all, unknown and role for every row. Expect 10 to 25% invalid on finder output and more on older lists. Output: the same list with a result column. Cost: 0.3 cents per address on the Starter pack, less in bulk.
Valid to the cold sequence. Catch-all and unknown to LinkedIn, phone or a low-volume warm sequence from a secondary domain. Invalid dropped and claimed back from the finder if it charged. Then split valid by vertical and seniority for messaging. Output: three to six sequences. Cost: an hour.
Add headcount, funding, tech stack and triggers to the valid segment, not the whole list. Enrichment costs per row and there is no point paying for a dead address. Output: scored rows. Cost: 2 to 20 cents per row.
B2B addresses decay fast, with only 19.2% of addresses still valid 90 days after verification in the BounceZero decay study. Re-verify any segment older than ninety days before it goes into a new sequence. Output: a calendar entry. Cost: the same 0.3 cents per row, before every send.
What each source returns and what the verifier finds when the output is checked. Invalid rates for named sources are the on-site figures; corpus rates are from the lead list quality benchmarks.
| Source | Gives | Invalid on verification | Cost | Notes |
|---|---|---|---|---|
| Your own inbound and trial signups | Names, addresses, intent | 5 to 10% | Free | Typos and throwaways; validate at the form |
| Database export (Apollo, ZoomInfo, Cognism) | Names, titles, addresses, phones | 5.8% invalid plus 18.4% catch-all | Subscription | Database age; see the Apollo study |
| Pattern finder (Hunter, Snov, FindyMail) | Addresses from name plus domain | 15 to 28% | Per find | Catch-all domains inflate find rate |
| Sales Navigator plus finder | Current names and roles, then addresses | 15 to 25% | Subscription plus finds | Freshest names; addresses still inferred |
| Public registries and directories | Companies and generic addresses | High share of role addresses | Free | Good for step two, weak for step three |
| Bought lists | Everything, of unknown age | Corpus: about 1 in 7 invalid; .au 36.5% | Per thousand | Sample-verify before paying |
| Scraped lists | Whatever was on the page | Highest role and catch-all share | Low | Expect to discard a third |
Corpus: 6,514,884 non-freemail (business and ISP) addresses, April to October 2026, 14.5% invalid, 2.3% catch-all, 27.0% unknown, 32.6% confirmed valid. Full tables by country and provider on the benchmarks page.
Teams price lists per row and then wonder why the sequence cost more than planned. Price per reachable row instead. A database export at 20 cents a contact with 20% invalid costs 25 cents per deliverable address before any enrichment. A finder at 10 cents a find with 25% invalid costs 13 cents. A bought list at 5 cents a row that comes back 36% invalid, which is what .au lists ran in the corpus, costs 8 cents per reachable row and arrives with the oldest data of the three.
Verification is the cheapest line in the stack. At 0.3 cents per address on the Starter pack and 0.22 cents on the Scale pack, checking 5,000 rows costs less than a single finder subscription month, and it is the step that turns the other costs into reachable contacts rather than bounces.
For a first list of a few hundred contacts, no subscription is needed. These four routes cover most verticals.
Companies House, OpenCorporates, state registries and sector regulators list every company with an address and often a director. Filter by SIC code and incorporation date to find companies at the stage you sell to.
A company hiring for a role your product supports is a trigger. Search the boards by title, collect the companies, then find the hiring manager. Fresh by definition.
Agency directories, app marketplaces, industry awards, conference sponsor pages. Curated lists of companies that already spend in the category.
With the company domain and two addresses from its website, infer the pattern, generate the candidate and check it. The free tier covers 100 checks a month, documented on the free email verifier page. Method ranking in how to find business email addresses.
The result column is the first segmentation key, before vertical or seniority. Each result gets a channel and a sending domain.
| Result | Meaning | Channel | Domain | Share on non-freemail (business and ISP) lists |
|---|---|---|---|---|
| Valid | Mailbox confirmed | Cold email sequence | Primary outreach domain | 32.6% (higher after dropping unknowns) |
| Catch-all | Domain accepts any address; mailbox unconfirmed | LinkedIn first, then low-volume email | Secondary domain | 2.3% |
| Unknown | Server would not answer | LinkedIn, phone, re-check in 7 days | Secondary domain if emailed | 27.0% |
| Role | info@, sales@, hr@ | Skip for personal outreach | None | Varies by source |
| Invalid | Mailbox does not exist | Drop; claim refund from finder | None | 14.5% |
Unknown is high on non-freemail (business and ISP) lists because ISP and corporate gateways and some ISPs refuse probes. A vendor that reports no unknowns is guessing on that share. Policy: an address that cannot be confirmed is never sold as valid.
Products sold as "lead list generators" are usually a database with filters and an export button, sometimes with a finder and a basic validity check attached. They do steps two and three. None of them do step one, and the built-in check at step four is an indexing-time check, not a send-time one, which is why their exports still carry 5.8% invalid and 18.4% catch-all addresses.
The process above runs with any of them or with none. What matters is that verification sits between the export and the sequencer, and that the result column drives segmentation. The list software comparison covers the tools; this page covers the process they slot into.
A 2,000-row list is not 2,000 leads. On business lists it is roughly 650 confirmed mailboxes, 540 unconfirmed, 290 dead and the rest role or catch-all. Price, segment and sequence the 650.
Lead types, sources and qualification
Sources ranked by data quality, build versus buy
The tools compared
Building lists for marketing rather than outbound
Segmentation beyond the verifier result
Why the ninety-day re-check exists
Start with public registries, job boards and vertical directories to build the company list, then find two or three contacts per company by inferring the address pattern from the company website and verifying the candidates. BounceZero gives every account 100 free checks a month, which covers a first list of a few hundred contacts. Paid finders and databases become worth it above a few hundred companies a month.
It depends on the step. Apollo, ZoomInfo and Cognism are strongest at pulling companies and contacts by filter. Hunter, Snov and FindyMail are strongest at turning a name and a domain into an address. None of them verify at send time, so the stack needs a verifier between export and sequencer regardless of which tool generated the list.
Enough for the sequence capacity you have, not more. One rep running personalised outreach reaches 150 to 250 new contacts a week, so a monthly list of 600 to 1,000 verified contacts is plenty. Lists larger than the team can work in ninety days decay before they are used.
Because the address was guessed from a pattern, or collected months ago, or sits on a catch-all domain that accepts anything. In corpus data from April to October 2026, 14.5% of non-freemail (business and ISP) addresses on lead lists were invalid and 27% could not be confirmed. Verifying before the first send removes the invalid share and routes the unconfirmed share to other channels.
Every ninety days for any segment that will be emailed again, and before any list older than that goes into a new sequence. Business addresses decay fast, with only 19.2% of addresses still valid 90 days after verification in the BounceZero decay study as people move roles, so a list verified in January is about 5% dead by April even if nothing else changed.
Upload the export, get valid, invalid, catch-all and unknown per row, and build sequences from the confirmed segment. 100 free checks a month, bulk jobs to a million rows.
Ayoub built BounceZero's 5-stage validation pipeline, its dedicated BGP-announced IP infrastructure, and the Patroni HA PostgreSQL cluster behind every verification. Previously built high-volume email delivery infrastructure. Trained at 1337 Benguerir (École 42 network, 2019). Open-source: bgp_analyzer.
Scoring, qualification, list building and the outbound and inbound stacks
SaaS lead scoring guide 2026
Twelve sales prospecting techniques compared on reply rate, effort and data quality
Lead qualification frameworks compared for 2026
The operating playbook for an outbound lead generation agency in 2026
Inbound leads software compared for 2026 by layer
Cold email lead generation agency guide
Continue through related topics