How to Build a Cold Email List with 97%+ Deliverability (2026)

Every cold email campaign that dies, dies for one of two reasons: the list was wrong (right people, wrong data — bounces, spam traps, dead inboxes) or the list was lazy (right data, wrong people — technically deliverable emails to prospects who were never going to buy). Copy gets the blame, but the list decides the outcome before the first email sends.

Having built lead lists behind campaigns targeting tens of thousands of prospects, here is the exact process that keeps deliverability at 97–98% and reply rates worth waking up for.

Start with the ICP, Not the Database

Before touching any data tool, write down your ideal customer profile in filterable terms: job titles (the 3–5 titles that actually sign off on your offer), company size band, industry, geography, and any qualifying signal (tech stack, funding stage, hiring activity). The discipline matters because every filter you skip now becomes manual list cleaning later — and because “everyone who could theoretically buy” is how you end up emailing 50,000 people to book four calls.

A useful test: can you write one email that feels personally relevant to everyone on the list? If not, the list needs to be split into smaller, tighter segments before you build it.

Choose Your Data Sources

Three sources cover almost every B2B use case, and they stack:

Apollo.io is the workhorse — the broadest filter set and the fastest path from ICP to raw list. Best for standard title/industry/size targeting. LinkedIn Sales Navigator wins when accuracy of current role matters most, since LinkedIn data is self-reported and fresh; the catch is that Sales Nav has no native export, so extraction requires tooling or a service. Crunchbase is the layer for funding-based targeting — companies by round, date and investor, which no general contact database filters as well.

For hard-to-reach niches, waterfall enrichment runs your list through 8+ data providers in sequence, each filling gaps the previous one missed. Typical recovery: 85–95% of contacts get a verified email or direct dial, versus 50–70% from any single source.

Verify Everything — and Handle Catch-Alls Properly

Raw exports are not send-ready. Email data decays 2–3% per month, so even a “verified” flag from the source platform needs re-checking. Run the full list through a dedicated verifier (MillionVerifier is our default) and act on the three outcomes:

Valid emails go into your campaign. Invalid emails get deleted, not “saved for later” — there is no later for a dead inbox. Catch-all results are the nuanced case: the domain accepts mail for any address, so the verifier can’t confirm the specific mailbox. Catch-alls are often perfectly good emails (many companies configure mail this way deliberately), but they carry higher bounce risk. The professional move is segmentation: send to catch-alls from separate inboxes and separate domains at reduced volume, so if a batch goes bad, your primary infrastructure never feels it.

The benchmark to hold: 97–98% deliverability on the valid segment. Below 95%, mailbox providers start reading your bounces as spammer behavior, and recovery takes weeks.

Segment by ESP Before You Send

An underused technique that meaningfully lifts inbox placement: detect which email provider each prospect’s company uses (Google Workspace, Microsoft 365, or a secure email gateway like Proofpoint or Mimecast) and route accordingly. Google-hosted recipients respond best to mail sent from Google-hosted inboxes; Microsoft-to-Microsoft behaves the same way. Prospects behind strict gateways deserve their own segment with more conservative sending and sometimes a different channel entirely. MX-record lookups reveal the ESP for every domain on your list — a detail most senders skip and most deliverability problems trace back to.

Format for Your Sequencer

The last mile: deduplicate on email address (not name — two John Smiths are two prospects, one email twice is a spam signal), normalize the columns your templates use (first names in proper case, company names stripped of legal suffixes), and add custom variable columns for personalization. Import into Instantly or Smartlead, map fields, and send a test to yourself before launching — one broken {{firstName}} merge tag costs more replies than a mediocre subject line ever will.

Build vs. Buy: The Honest Math

Building this pipeline yourself requires an Apollo subscription, a verification tool, ESP-detection tooling and several hours per list. That’s a fine investment if list building is your core skill. If it isn’t, the economics favor delegation: a done-for-you verified list — exported from your Apollo or Sales Nav search, verified, catch-all-segmented and formatted for your sequencer — starts at $4.99 per 1,000 leads with 24–48 hour delivery. You spend your hours on the copy and the replies, which are the parts that actually need you.

Frequently Asked Questions

How big should a cold email list be?
Smaller than you think. 1,000–3,000 tightly-matched prospects per campaign segment outperforms 20,000 loose ones, and lets you personalize copy per segment.

Is it legal to build cold email lists?
B2B cold email is legal in most jurisdictions with conditions — CAN-SPAM in the US requires accurate headers and an unsubscribe mechanism; GDPR in the EU requires legitimate-interest grounds and easy opt-out. The list itself isn’t the compliance risk; how you send is. This is general information, not legal advice — check the rules for your target countries.

How often should a list be refreshed?
Treat any list older than 90 days as unverified. Re-verify before reusing, and expect to lose 5–10% of it.

Want your next list built, verified and segmented for you? Order a verified lead list — delivered ready for Instantly or Smartlead within 48 hours.