Improve Email Deliverability: 10 Fixes Ranked by Impact

The fastest way to improve email deliverability is to fix issues in order of impact: authentication and alignment first, then complaint and list-quality problems, then infrastructure and content, because a failed DKIM record costs you more placement than any subject-line tweak ever will. Below are the 10 fixes that move the needle, ranked, each with an effort estimate and a concrete way to verify it worked, because a fix you have not verified is a guess.

The ranked list at a glance

Fix 1: pass SPF and DKIM on every message you send

Effort: low to medium. Since February 2024 for Gmail and Yahoo, and May 2025 for Outlook.com, bulk senders must pass both SPF and DKIM, and since November 2025 Gmail rejects non-compliant bulk mail outright at SMTP time. Nothing else on this list matters if authentication fails. Audit every service that sends as your domain (ESP, transactional provider, CRM, billing system, support desk) and make sure each is covered by your SPF record and signs DKIM with your domain. Common breakage: SPF's 10-DNS-lookup limit exceeded by stacking includes, and DKIM signing under the ESP's default domain.

Verify it worked: run an SPF check and DKIM check, then open the Authentication-Results header on a real delivered message from each sending system and confirm spf=pass and dkim=pass.

Fix 2: publish DMARC and fix alignment

Effort: medium. Publish at least v=DMARC1; p=none; rua=mailto:[email protected] on your From domain, required for bulk senders since February 2024. Then fix alignment: the domain that passed SPF or DKIM must match your visible From domain, and mail through ESPs frequently passes both while aligning neither. Read your aggregate (RUA) reports for two to four weeks, align every legitimate source, then move toward p=quarantine and eventually p=reject so lookalike abuse of your domain stops feeding your reputation problems.

Verify it worked: a DMARC check shows a valid record; delivered headers show dmarc=pass; RUA reports show your real senders at or near 100% aligned before you tighten the policy.

Fix 3: get off blacklists, and fix what put you there

Effort: medium. Impact: high, but only if you are actually listed. Check every sending IP, your From domain, and your click-tracking domain against the lists that matter (Spamhaus, SpamCop, Barracuda, SURBL/URIBL). A Spamhaus DBL listing on your domain will suppress delivery no matter which ESP you use. Crucially, delisting is step two; step one is diagnosing the cause (spam trap hit, compromised account, list quality), because operators re-list unfixed senders within days. The full process per list is in our blacklist guide.

Verify it worked: a clean blacklist check across all sending assets, and no recurrence over the following 30 days of monitoring.

Fix 4: add one-click unsubscribe and honor it within 2 days

Effort: low. Add the List-Unsubscribe header (HTTPS URI) plus List-Unsubscribe-Post: List-Unsubscribe=One-Click per RFC 8058 to all marketing and subscription mail. Required for bulk senders since 2024, enforced since June 2024, and genuinely good for you: every easy unsubscribe is a spam complaint that did not happen. Opt-outs must take effect within 2 days; instant is better.

Verify it worked: send to a Gmail seed account and confirm the unsubscribe button appears beside the sender name; click it and confirm suppression in your ESP within 2 days.

Fix 5: get spam complaints under 0.10%

Effort: medium. Google's thresholds are explicit: stay under 0.10% user-reported spam sustained, never touch 0.30%. Complaints come from mailing people who forgot you, mailing too often, or mismatched expectations at signup. The levers: confirm consent at collection, send a real welcome message immediately so recipients recognize you later, state cadence up front, and stop mailing chronically unengaged segments (see fix 7).

Verify it worked: Google Postmaster Tools spam-rate graph trending under 0.10% for four consecutive weeks; complaint data from Yahoo's feedback loop and Microsoft JMRP showing the same direction.

Fix 6: clean your list of bounces, traps, and dead weight

Effort: medium. Hard bounces above roughly 2% mark you as a list-quality problem; recycled spam traps live precisely in the addresses that stopped engaging years ago. Suppress hard bounces immediately and automatically, run new imports through address validation, never add purchased or scraped addresses, and sunset recipients with no opens or clicks in 6 to 12 months (send a final "still want these?" message, then stop).

Verify it worked: bounce rate under 2% on the next three sends, no new trap-driven blacklist listings, and a stable or rising unique-open trend even if list size dropped. Shrinking a list usually grows its output.

Fix 7: segment by engagement and mail your engaged readers first

Effort: medium. Mailbox providers score you on how recipients react. Sending every campaign to everyone means your least interested subscribers dilute the signal your best ones generate. Split by recency of engagement (for example 0-30, 31-90, 91-180 days), send new campaigns to the hottest segment first, and touch colder segments less often with your strongest content. This is also the honest core of ramping a new domain, as opposed to the bot-driven shortcuts covered in our warmup post.

Verify it worked: Postmaster Tools domain reputation moving from medium toward high over 4 to 8 weeks, and seed tests showing improved placement on campaigns sent engaged-first versus full-list blasts.

Fix 8: ramp volume gradually and keep a consistent cadence

Effort: low. Reputation systems distrust volatility. A domain that sends 500 messages daily then blasts 80,000 on Friday looks compromised. New domains and IPs should start in the low hundreds per day and roughly double every 2 to 3 days while metrics stay clean; established senders should keep week-to-week volume within a sane band and split giant sends over several days.

Verify it worked: no bounce-code spikes (Gmail 421 rate-limit deferrals, in particular) during ramp, and steady delivered rates as volume steps up.

Fix 9: fix the template flags: text ratio, link count, message weight

Effort: low. Content matters less than reputation, but templates still trip filters: image-only emails with no real text, dozens of tracking links, link shorteners (bit.ly and friends are heavily abused and heavily scored), mismatched link text versus destination, and messages over Gmail's 102 KB clipping limit. Keep a sensible text-to-image balance, link to your own domain, and always include a plain-text part.

Verify it worked: run the template through a spam checker for rule-level flags before sending, and compare seed placement for the revised template against the old one.

Fix 10: verify infrastructure: reverse DNS, TLS, tracking domain

Effort: low. Every sending IP needs forward-confirmed reverse DNS (PTR resolves to a hostname that resolves back to the IP), transmission must use TLS, and your click-tracking domain should be a branded subdomain (like link.yourdomain.com) rather than an ESP-shared default that pools reputation with strangers. If you use an ESP, most of this is theirs to run but yours to check.

Verify it worked: dig -x on each sending IP round-trips correctly, delivered headers show TLS in the Received chain, and your tracking domain comes back clean on domain blacklists.

Monitor it, because deliverability is not set-and-forget

Every fix above decays. A colleague edits DNS and breaks the SPF record in March. A blacklist adds your IP overnight in June. A form gets abused, complaints creep from 0.06% to 0.2%, and nobody notices until revenue does. Deliverability failures are quiet by design: mail does not bounce back from the spam folder, dashboards keep showing "sent," and the first human signal is often a sales team asking why reply rates halved six weeks ago. The senders who stay in the inbox are not the ones who fixed things once; they are the ones who notice regressions in days instead of quarters. In practice that means a recurring placement test on your main streams (monthly at minimum, weekly at meaningful volume), continuous blacklist monitoring across every sending asset, DMARC aggregate report review so new or broken sources surface fast, a weekly glance at Postmaster Tools, and alerts wired to a person who owns the response.

And one honest note before you work the list: nobody can guarantee inbox placement, including us. Seed tests are directional estimates, mailbox providers personalize filtering per recipient, and reputation moves on their schedule, not yours. These ten fixes remove the known, measurable negatives, which is everything a legitimate sender can actually control. Anyone promising more than that is selling certainty they do not own, and it is fair to distrust the pitch.

If you want to know which of the ten applies to you, measure before you fix: run a placement test with Inboxes and you get your landing position across Gmail, Outlook, Yahoo, iCloud, GMX, and Zoho, plus authentication, blacklist, and template checks in the same report, ranked by impact. Start at the top of your list, not ours.