The 2026 cold email deliverability checklist

A clever opener cannot compensate for a domain that fails authentication. Configure SPF to authorize the systems that send for your domain, sign mail with DKIM, and publish a DMARC policy that aligns the visible From domain with SPF or DKIM. Also verify forward and reverse DNS where your sending setup controls the underlying IP.

• Keep one valid SPF record and include every legitimate sender.
• Use a DKIM key and selector supported by your mailbox provider.
• Begin DMARC monitoring safely, then strengthen policy as reports become clean.
• Send over TLS and keep message formatting standards-compliant.

Do not mix product alerts, password resets, support messages, and outbound prospecting on the same sending identity. Separating message types limits the damage if one stream receives complaints and makes reputation problems easier to diagnose.

• Keep customer-critical transactional mail away from outbound campaigns.
• Use consistent sender identities instead of frequently changing names and domains.
• Monitor each domain and mailbox independently.

Deliverability starts before the send. Validate addresses, remove obvious role accounts when they are not relevant, avoid stale scraped lists, and segment by a real reason the recipient could care. A smaller list with a defensible fit is safer than a broad list with superficial personalization.

• Suppress addresses that hard-bounce or explicitly unsubscribe.
• Review the source and age of every imported list.
• Pause segments with unusual bounce or negative-reply patterns.
• Never retry a permanent delivery failure as if it were a temporary delay.

Increase sending gradually and hold variables steady long enough to understand the result. Sudden spikes, inconsistent daily patterns, and many new inboxes activating at once make reputation harder to interpret. Provider limits are ceilings, not targets.

• Start below the mailbox's configured safety limit.
• Use predictable sending windows and avoid abrupt spikes.
• Change one major variable at a time when diagnosing performance.
• Slow down when replies, bounces, or account health deteriorate.

Open tracking is increasingly noisy and can add another remote asset to the message. Use delivered volume, hard bounces, spam complaints, total replies, positive replies, and meetings as the core operational funnel. Review these by domain, mailbox, audience segment, and campaign rather than only as blended account totals.

A weekly review should identify what changed before performance drops become account-wide. Check authentication reports, provider dashboards, sending volume, rejection codes, bounce reasons, complaint signals, and the quality of replies. Document the change you make and the date so the next review has context.

• Authentication pass rate and DMARC reports
• Hard and soft bounce patterns
• Spam complaint rate and provider warnings
• Reply quality by segment and sender
• Unexpected volume or configuration changes