Published by the SELLIGNITE team. A practical operating model with examples and checks that show whether it is working. Examples are illustrative unless stated otherwise.
To build a B2B lead list, start from your ideal customer profile, not from a database export. Pull matching accounts from two or three sources, cut the ones that fail your exclusion rules, find one or two people per account who own the problem, enrich only the fields your message needs, then verify every email address and drop anything risky. A small, verified list of a few hundred contacts beats a large unverified one every time.
- Build the account list first and the contact list second; a wrong account cannot be saved by a right title.
- Use two or three sources and cross-check them, because every B2B database has gaps and stale records.
- Enrich only the fields your message and routing actually use, such as role, trigger and verified email.
- Verify every email address before the first send and remove invalid, risky and catch-all results you cannot confirm.
- Re-verify any list older than about 30 days, because people change jobs and inboxes close.
- 01Define the filter
- 02Pull accounts
- 03Find contacts
- 04Enrich the gaps
- 05Verify and load
What makes a good B2B lead list?
A good B2B lead list is a short list of the right people at the right companies, with contact data you have checked. It is not the biggest export your database allows. Every row should answer three questions: does this account fit, does this person own the problem, and will this email address actually receive the message?
Most outbound problems that look like copy problems start in the list. A strong email sent to the wrong role gets ignored, and a strong email sent to a dead address bounces and hurts the sending domain for everyone else on the list. That is why we treat list building as its own stage in a repeatable outbound system, with its own owner and its own quality checks.
A useful list has a few fields that matter and many that do not. The ones that matter are company name and domain, the reason the account fits, the contact's name, role and verified email, the LinkedIn profile URL, the source of each record and the date it was verified. If a field does not change who you contact, what you say or how you route the reply, you do not need it yet.
Step 1: Turn your ICP into filters a database understands
Start from your ideal customer profile and translate it into filters that a tool can apply. A description such as "growing B2B service firms with messy sales handoffs" is useful for people and useless for a database. A database needs industry codes, headcount ranges, countries, technologies used and keywords in the company description.
Write two lists: inclusion filters and exclusion rules. Exclusions do more work than most teams expect. Typical ones are current customers, open opportunities, competitors, partners, companies below a minimum headcount and regions you cannot serve. Our ICP guide explains how to set fit, need and timing criteria before you touch a database.
Then add a timing layer if you have one. Accounts with a recent trigger, such as a new head of sales, a funding round or a new office, deserve to be first in line. The buying signals guide lists the triggers worth tracking and how to read them without guessing.
Step 2: Choose your lead list sources and cross-check them
Use two or three sources rather than one, because no B2B database is complete or fully current. Each source is strong at something different, so the combination fills gaps that any single one leaves.
| Source type | Examples | Good for | Watch out for |
|---|---|---|---|
| Professional network | LinkedIn Sales Navigator | Current roles and job changes | No email addresses; manual export limits |
| Contact databases | Apollo, ZoomInfo, Lusha | Volume, firmographics, emails | Stale records and guessed emails |
| Enrichment platforms | Clay | Combining several providers in one table | Cost and setup time per field |
| Your own data | CRM, past deals, event lists, website visitors | Context and relationship history | Duplicates and missing consent records |
| Public signals | Job posts, company news, registries | Timing and the reason to write | Manual effort to collect |
These are examples, not recommendations; plans and data coverage change, so test any provider against a sample of accounts you already know well. If a source gets your own customers wrong, it will get prospects wrong too.
Record the source on every row. When a segment bounces or replies badly, the source column tells you which provider to question.
Step 3: Find the right contacts at each account
Pick one or two contacts per account who own the problem you solve, and add a third only for larger companies. More contacts per account rarely means more meetings; it often means several colleagues comparing the same cold email in a team chat.
Search by function and seniority, not by exact title. Titles vary widely between companies: the person who owns new business might be a head of sales, a commercial director, a managing partner or the founder. Write down three to five acceptable titles per persona and a short list of titles to skip, such as interns, assistants and roles in unrelated departments.
Check the person still works there. Job changes are the biggest single source of decay in a lead list, and a LinkedIn profile updated last month is better evidence than a database record updated last year. A recent job change can also be a reason to write, which ties list building to the message itself.
Step 4: Enrich only what your message needs
Lead list enrichment means adding missing data to each row, such as email address, company size, technologies or a recent trigger. The rule is simple: enrich a field only if it changes who you contact, what you say or how you route the reply. Everything else adds cost and clutter.
Use a waterfall for emails. Waterfall enrichment means asking a first provider for the email, and only if it has no confident answer, asking a second and then a third. Tools such as Clay make this easy to set up, and the same logic works with a spreadsheet and two providers. A waterfall usually finds more addresses than any single source, and it lets you prefer the provider that proved most accurate in your tests.
Add one research field per account: the reason you are writing. It can be a trigger, a hiring pattern or a line from the company website. This is the field that turns a list into a campaign. An AI agent can draft it from public pages, as described in our guide to using AI to research accounts before outreach, with a person checking each note before it is used.
Step 5: Verify every email in your B2B lead list before you send
Verify every email address before the first send, every time. An email verification tool, such as NeverBounce, ZeroBounce or the checker built into many databases, tests whether the domain accepts mail and whether the mailbox appears to exist, without sending a message. Results usually come back in four groups:
- Valid: the mailbox appears to exist. Keep it.
- Invalid: the mailbox or domain does not exist. Remove it.
- Catch-all (accept-all): the domain accepts every address, so the check cannot confirm the person. Keep only if a second source confirms the same address, and send to these in a separate, smaller batch.
- Unknown or risky: the server did not answer clearly, or the address is a role inbox such as info@ or sales@. Remove it from cold sequences.
Bounces matter because mailbox providers use them to judge whether a sender is careful. Our own rule is to pause a campaign and check the list if hard bounces go above 2%. Verification works together with the domain setup in our email deliverability guide; neither one replaces the other.
Finally, deduplicate against your CRM before loading, so that customers, open deals and people who asked you to stop never receive a cold email. In the EU and UK, make sure you have a documented basis for B2B outreach, such as legitimate interest, and an easy way to opt out; check the details with your own legal adviser.
Example: building a 300-contact list in one afternoon
Imagine a firm that sets up outbound systems for B2B software companies with 20 to 150 employees in the UK and the Netherlands. The example is illustrative; the steps are the ones we use.
- Filter. Software, 20-150 employees, UK or NL, at least one open sales job posted in the last 60 days. Exclusions: current customers, open deals, agencies and companies with no sales team listed.
- Pull accounts. Sales Navigator returns the account list; a contact database adds domains and headcount. After exclusions and a manual skim, 180 accounts remain.
- Find contacts. Personas: founder or CEO for companies under 50 people, head of sales or VP sales above that. One contact per account, two for the 40 largest accounts: 220 contacts.
- Enrich. A two-provider email waterfall finds addresses for most contacts. An AI agent drafts one "reason to write" line per account from the job post; an SDR reviews and edits each line.
- Verify and load. Verification removes invalid and risky addresses, and catch-all addresses go into a separate small batch. The CRM check removes a handful of existing contacts. The rest load into the sequence with source and verification date.
The list is smaller than the first export, and that is the point. Every row now has a fit reason, an owner of the problem, a reason to write and an address that should receive the email. Repeat the same process every week with a new trigger, rather than buying one giant list once.
How to keep a lead list from going stale
Treat a lead list as perishable. People change jobs, companies reorganize and inboxes close, so a list that was clean in spring is not clean in autumn. Re-verify any list older than about 30 days before you send to it, and refresh contacts on accounts you plan to keep working.
Close the loop with results. Track bounces, replies and meetings by source and by segment. When one source keeps producing bounces or wrong-person replies, lower its priority in the waterfall. When one segment produces meetings, build more lists like it. For well-targeted outbound we plan for an 8-12% reply rate and about one meeting for every three replies, as targets rather than guarantees, and the list is usually the first lever to check when results fall short; see the outbound metrics that matter.
This is also where AI agents save the most time. An agent can run the waterfall, flag job changes, draft research notes and prepare clean rows for the CRM, while a person approves the filters and spot-checks the output. We build that workflow as part of our outbound sales service, usually within the first two weeks of an engagement, and you can start with a free AI Audit call.
Your implementation checklist
- Write inclusion filters and exclusion rules from your ICP
- Use two or three sources and record the source on every row
- Pick one or two contacts per account who own the problem
- Run an email waterfall and add one reason-to-write field per account
- Verify every email and remove invalid, risky and role addresses
- Deduplicate against the CRM before loading the sequence
- Re-verify any list older than about 30 days
What to measure: Hard bounce rate, reply rate and meetings booked per 100 contacts, tracked weekly by list source and segment.
Frequently asked questions
How many contacts should a B2B lead list have?
Start with a few hundred contacts per segment, enough for one or two weeks of sending, rather than thousands at once. Smaller lists are easier to verify, personalize and learn from, and you can build the next batch with what the first one taught you.
What is waterfall enrichment?
Waterfall enrichment means asking several data providers in order for the same field, usually an email address, and moving to the next provider only when the previous one has no confident answer. It typically finds more valid addresses than one provider alone and lets you rank providers by accuracy.
Should you send cold emails to catch-all addresses?
Send to catch-all addresses only when a second source confirms the same address, and keep them in a separate, smaller batch so any bounces are easy to spot. A catch-all domain accepts every address during verification, so the check alone cannot confirm that the person exists.
How often should you re-verify a lead list?
Re-verify any list that is older than about 30 days before you send to it. People change jobs and companies close or rename inboxes, so contact data decays steadily even when it was clean on the day you built it.
Is it better to buy a lead list or build one?
Building your own list from your ICP is usually better, because you control the filters, the contacts and the verification. A bought list can be a starting source, but it should go through the same exclusion, enrichment and verification steps as any other data.