How to Build a B2B Prospect List, Step by Step

How do you build a B2B prospect list that actually sends?

You build it in six steps: scope the ICP, source contacts, enrich them, verify the emails, segment, then hand off to outbound. The verify step is the one most teams skip. Google now caps spam complaints at 0.30% for bulk senders (Google, Email sender guidelines, 2024), so an unverified list can sink every campaign behind it.

This guide covers the build mechanics, not ICP theory. If you need to define your ideal customer first, start with the GTM glossary and our ICP field notes, then come back here.

Key takeaways

List building is a recurring hygiene cadence, not a one-time scrape. B2B databases decay roughly 22.5% per year (HubSpot, citing MarketingSherpa, 2024).
Verification is now a deliverability requirement. Keep total bounces under 2% to protect sender reputation.
Test any data provider on a 1,000-contact sample before you commit budget. Tested accuracy lags marketing-page claims.
A smaller, verified, ICP-tight list beats a big unverified one. Precision drives reply rates, not volume.
Own the data. Treat the list as infrastructure you refresh, not a file you buy once.

Step 1: How tightly should you scope the ICP before sourcing?

Scope it tightly, then stop. Poor data quality already costs organisations an average of $12.9 million per year (Gartner, 2020), and a loose ICP multiplies that waste at the sourcing stage. You only need four filters to start: industry, headcount band, geography, and a job-title set tied to the buying decision.

Resist the urge to write a 20-line persona doc here. The build stage rewards filters you can actually query in a data tool, not adjectives. "Series A UK SaaS, 11 to 40 staff, Head of Sales or RevOps" is buildable. "Ambitious scale-ups who value efficiency" is not.

Scope creep at this step is the quiet tax on every later step. A vague ICP produces a list that enrichment cannot sharpen and segmentation cannot rescue. Define four hard filters, source against them, and refine the persona in the glossary separately. The list is for matching, not for storytelling.

Step 2: Where should you source contacts from?

Source from three layers and combine them: a paid data provider for coverage, LinkedIn for verification and gaps, and intent signals for timing. Independent benchmarks suggest no single provider exceeds roughly 90% email accuracy on real lists (Cleanlist, 2026), so leaning on one source guarantees blind spots.

Data providers

Pick a provider on regional coverage, not record count. For UK and EU lists, phone-verified EMEA data and GDPR compliance matter more than a headline contact total. The Cleanlist test is a directional benchmark from a vendor-adjacent source with partly disclosed methodology, so treat the numbers as a starting point, not gospel.

Provider (independent test)	Email accuracy (directional)	Notes
Cognism	~90%	Stronger EMEA coverage, phone verification
ZoomInfo	~85%	Historically thinner on EU mobile
Apollo	~80%	Broad, budget-friendly, more cleanup

Source: Cleanlist 2026 1,000-lead test. Directional only; methodology partly disclosed.

LinkedIn and intent signals

Use LinkedIn to confirm titles and catch role changes the database missed. Layer intent signals (hiring spikes, funding, tech adoption) to prioritise who gets contacted first. Signals do not build the list; they order it. Our data research work treats signal as a ranking layer on top of owned data.

Step 3: What does enrichment actually add?

Enrichment fills the gaps that sourcing leaves: missing job titles, company size, direct dials, and the data points your segmentation depends on. It matters because 76% of organisations say less than half of their CRM data is accurate and complete (Validity, State of CRM Data Management, 2025). Thin records are unusable records.

Enrich against your ICP filters, not against everything available. You need the fields that drive segmentation and personalisation, and nothing else. Extra columns you never query are just more data to decay.

Validity surveyed 602 CRM users across the US, UK and Australia and found 37% of organisations lose revenue directly because of poor data quality, with teams losing an average of 16 deals per quarter to bad data. Enrichment is the step where you decide whether a record is rich enough to act on, or noise you carry forward.

Step 4: Why is email verification non-negotiable now?

Verification protects deliverability, and deliverability is now governed by hard thresholds. Google requires bulk senders to keep spam complaints under 0.10% and never reach 0.30%, with SPF, DKIM and DMARC mandatory above 5,000 daily messages (Google, 2024). One unverified list can breach the bounce limit and damage every send after it.

The mechanism is simple. High bounce rates tell mailbox providers your data is dirty, and dirty data signals a spammer. Keep total bounces under 2%, and hard bounces under 1% if you want top-performer numbers (MailReach, 2025). Verification is the primary lever to stay under that line.

What verification actually catches

A good verifier removes invalid syntax, dead domains, and known-bad mailboxes before you send. Independent testing from EmailToolTester puts the better tools around 95% or higher real-world accuracy, with budget tools nearer 87 to 95% (EmailToolTester, 2025). Note the gap: vendors advertise 99%, independent tests land lower. Use the tested figure.

Watch catch-all domains too. Verification vendor ZeroBounce reported that over 9% of the emails it processed were catch-all (risky) addresses (ZeroBounce, Email List Decay Report, 2025). Decide your tolerance for catch-alls per campaign rather than blanket-sending to them. We keep the verification step inside our outbound build so nothing reaches a mailbox unchecked.

Step 5: How should you segment the verified list?

Segment by who the message is for, not by what you have. Segmentation is where precision starts paying off. Cognism's State of Outbound 2026, drawn from 451,895 logged calls, found verified data let cold SDR outreach reach a 13.3% answered rate. That is close to the 14.4% AEs hit on warm leads (Cognism, 2026). Tight targeting closes the gap.

Build segments around the angle, not the database field. Group by trigger (recent funding), by role (RevOps versus founder), and by pain (stale CRM versus weak deliverability). Each segment should map to one message you can actually write. If two segments would get the same email, merge them.

Cognism's data is from its internal teams, so it may not generalise, but the direction is consistent with what operators see: a smaller, verified, well-segmented list outperforms a large unverified one. The spray-and-pray approach pays a tax in bounces, complaints, and reputation that the volume never recovers.

Step 6: How do you hand the list to outbound?

Hand off with the verification status, the segment, and the sourcing date attached to every record. The date matters because the list is already decaying: ZeroBounce found at least 23% of an email list degrades per year (ZeroBounce, 2025), independently echoing the ~22.5% benchmark. A record's age is a sending risk, not a footnote.

Set a refresh cadence at handoff, not later. A 10,000-contact list built in January is roughly 2,000 records wrong by December. Treat re-verification as a recurring job, monthly for active segments, before each major send for the rest. This is why we treat the list as owned infrastructure you maintain, not a file you buy once.

Then book the campaign against a list you trust. If you want the build and the sending run as one owned system, that is what we do: book a call.

Frequently asked questions

How often should I rebuild a B2B prospect list?

Re-verify continuously rather than rebuild from scratch. B2B databases decay roughly 22.5% per year, about 2% per month (HubSpot, citing MarketingSherpa, 2024). Run monthly verification on active segments and a full refresh before any large send. Treating it as ongoing hygiene beats periodic mass scrapes.

Is it better to buy data or build the list manually?

Combine both. Buy for coverage, then verify and enrich against your own ICP filters, so you own the result. Test any provider on a 1,000-contact sample first, because independent benchmarks show real-world accuracy near 80 to 90%, below marketing-page claims (Cleanlist, 2026).

What email verification accuracy is realistic?

Expect around 95% or higher from the better independent-tested tools, and 87 to 95% from budget options (EmailToolTester, 2025). Vendor pages advertise 99%, but independent tests land lower, so plan around the tested figure. Verification removes the bulk of invalid addresses, not every single one.

What bounce rate is safe for cold email?

Keep total bounces under 2%, and hard bounces under 1% if you want top-performer numbers (MailReach, 2025). Google caps spam complaints at 0.30% for bulk senders (Google, 2024). Verifying before send is the main lever to stay under both thresholds and protect sender reputation.

Does a bad prospect list really cost revenue?

Yes, and the cost is measurable. Gartner puts the average annual cost of poor data quality at $12.9 million per organisation (Gartner, 2020). Validity found 37% of organisations lose revenue directly to bad data, with teams losing about 16 deals per quarter (Validity, 2025).

About the author

Written by Hugo Dupont, founder of Empra Labs, where the team has sent 1.6M+ emails at a measured 7.4% reply rate and built owned pipeline infrastructure for 40+ B2B teams, generating £60M+ in pipeline. All proof numbers are Empra's own measured results; client outcomes vary, and any AI-assisted drafting is reviewed by a human before it ships.

Frequently asked questions

How often should I rebuild a B2B prospect list?

Re-verify continuously rather than rebuild from scratch. B2B databases decay roughly 22.5% per year, about 2% per month (HubSpot citing MarketingSherpa, 2024). Run monthly verification on active segments and a full refresh before any large send. Treating it as ongoing hygiene beats periodic mass scrapes every time.

Is it better to buy data or build the list manually?

What email verification accuracy is realistic?

What bounce rate is safe for cold email?

Keep total bounces under 2%, and hard bounces under 1% for top-performer numbers (MailReach, 2025). Google caps spam complaints at 0.30% for bulk senders (Google, 2024). Verifying before send is the main lever to stay under both thresholds and protect sender reputation across every campaign.