Trust beats tactics™

Results · names removed, mechanisms named

What we find in lists other tools call clean.

Real numbers from real clients, anonymized but specific.

Every one of these lists had been validated before we touched them. Four situations, four industries, four different findings. Start with the one that sounds like yours.

01Before you upload anything

Three checks you can run yourself in 20 minutes

If your authentication is broken, no list cleaning will save you. Start here.

  1. Count short freemail addresses.

    Filter your list for gmail, yahoo, or hotmail addresses with 6 characters or less before the @. Now look at where those signups came from. Referrals and direct organic signups get benefit of the doubt: pull their engagement (opens, clicks) for the last 6 months and demote silent ones to a monitor cohort. Form-fill signups with no rate-limiting or paid ad traffic get a harder look: silent non-openers in this cohort are more likely typos or bot submissions.

  2. The alphabetical eyeball test.

    Export your list to Excel or Google Sheets, sort A to Z. Scroll to the top and bottom. Anything look wrong? Random letter strings, weird digit patterns, gibberish, sequential prefixes like abc123 at the same domain? Those are the ones your validator called valid. They didn’t bounce. They also aren’t opening.

  3. Pull every negative signal from the last 18 months.

    Not just hard bounces. Soft-bounces (any address that soft-bounced even once), complaints (spam-button hits), and addresses that stopped opening or clicking 3+ months ago. Some are suppress candidates, some are demote-to-monitor candidates. Either way they’re the cohort dragging your sender reputation.

02The case library

Four lists. Four industries. Four different findings.

Same intent (clean your list), but the same list in different industries gets different results. Each card names what the client believed, what we found, the receiver-side mechanism behind the miss, and what happened after suppression.

Case 01Client A · E-commerce 127,000 addresses

What they believed

They had cleaned their list with a well-known validation tool three months earlier. 96.4 percent valid. Green light across the board.

What the review found

111,442 keep11,340 monitor4,218 suppress
  • 1,600 addresses with address-shape signals validators can’t see: short freemail local-parts (typo signups), digit-heavy random strings (form-fill spam), sequential-signup patterns.
  • 1,500 had soft-bounced three or more times in the last 18 months without ever hard-bouncing, so validators still marked them valid.
  • 1,100 sat on domains where the MX lookup now returns nothing. The sending path is dead but the address still exists in your ESP.
  • 8,400 in the monitor cohort had no engagement signals for over 12 months.

What changed after suppression

Inbox placement 74% 91% +17 pts
Complaint rate cut by more than half

Within two send cycles of suppressing the flagged addresses. Measured via seed-list panels at Gmail, Outlook, and Yahoo.

A green light from a validator is a statement about mailboxes, not about what your next send does to your reputation.

Case 02Client B · B2B SaaS 34,000 addresses

What they believed

Small, curated list. Mostly inbound signups from their website. They had never cleaned it because these people opted in.

What the review found

26,569 keep4,560 monitor2,871 suppress
  • Over 1,400 addresses used company domains that no longer resolve. The MX lookup returns NXDOMAIN, but most validators flag the address as risky without surfacing the domain failure, so senders ignore the warning.
  • 980 were role addresses (info@, sales@, support@) that had accumulated over three years of form submissions.
  • 490 sat on domains showing high-risk reputation signals: old subdomains repurposed for outbound spam, domains freshly transferred with no send history, freemail-adjacent throwaways your validator’s library hadn’t seen yet.
  • The bounce code waiting for them would have been a 5.7.1 policy reject, not the 5.1.1 mailbox-doesn’t-exist that validators predict.

What changed after suppression

Share of the list that could not receive email, or should never have been mailed 22%
Open rates declining quarter over quarter back to their level of 18 months earlier

They had been wondering why their open rates kept declining quarter over quarter. This was why.

Case 03Client C · Marketing agency 410,000 addresses · 12 client lists

What they believed

They ran every client list through validation before onboarding. Thought they were covered. One client got blocklisted anyway and blamed them.

What the review found

328,100 keep53,200 monitor28,700 suppress
  • The blocklisted client’s list had 340 addresses our address-pattern flags called high-risk. Not because the mailbox failed, but because the shape of the address and the age of the receiving domain matched patterns we’ve seen tank sender reputation in the same window.
  • The list was also single opt-in only, which meant it was full of hidden issues from signup: bot-form submissions, typo-catches, and people who never actually confirmed they wanted to subscribe. Their validation tool said valid for every single one.
  • Reputation at major receivers is scored on the per-IP-and-domain pair, not the IP alone, so even a clean sending IP carried the listed domain straight into the spam folder.

What changed after suppression

Delisting, after removing the flagged addresses within 1 week
Clients blocklisted since 0
Client lists now reviewed before the first send 0 of 12 12 of 12

The agency now runs every client list through Review My Emails before the first send.

Case 04Client D · Newsletter publisher 89,000 subscribers

What they believed

Organic growth over four years. No purchased lists. Double opt-in. Our list is clean because we built it right.

What the review found

66,080 keep19,800 monitor3,120 suppress
  • Even clean lists decay. 14,300 addresses had not opened or clicked in over 18 months. Not invalid, just gone. People who signed up, read for a while, then stopped. Their mailbox still exists but nobody is home.
  • 3,100 sat on domains that had changed MX providers, with the old provider still accepting mail into a black hole.
  • The suppression group included 1,800 addresses at companies that had been acquired and migrated to new domains.

What changed after suppression

Active subscribers won back via re-engagement about 4,000
Sender reputation at Google Postmaster Tools medium high 6 weeks
Gmail inbox rate, engaged cohort +14 pts

They segmented the monitor group into a re-engagement campaign, won back the readers who were still there, and suppressed the rest.

See how the verdicts are made

03Across every list

The patterns we see across every list

After analyzing millions of addresses, the same problems keep showing up. Not because senders are careless. Because validation tools do not look for these signals.

  • Validation cannot see the risks under a 250 OK.

    An address that accepts mail is not the same as an address that is safe to mail. Security filters at Proofpoint, Mimecast, and Microsoft Defender accept then quarantine. Quarantine doesn’t mean your mail is bad, though. It means the filter is strict and something about your send is triggering it: sender authentication that doesn’t quite align, the wrong kind of subscriber attracted by the wrong kind of signup path, an audience mismatch between who signed up and who you’re now sending to. The fix isn’t just list cleaning. It’s cleaning the list plus setting up authentication properly plus adjusting who your signup paths are attracting. Freemail throwaways from services no validator library has cataloged yet return 250 OK the same way real inboxes do. Long-abandoned government and enterprise addresses accept and never open, dragging your engagement score month over month. Every one of these looks like a valid delivery to a bouncing-address check.

  • B2B list decay runs 25 to 30 percent annually.

    People change jobs, companies get acquired, domains lapse. In the lists we analyze, roughly a quarter of B2B addresses go bad every year even with no purchased data and no list hygiene mistakes. Time alone does the damage.

  • Double opt-in does not equal clean.

    Every list we have analyzed, no matter how carefully it was built, had addresses that should not be mailed. Double opt-in confirms intent at the moment of signup. It does not protect against domain churn, mailbox abandonment, or the slow drift from engaged to dormant.

  • Role addresses can be a quiet reputation tax.

    info@, sales@, support@ pass validation every time. But they are often monitored by multiple people or nobody, have higher complaint rates, and some receivers (notably Gmail and Yahoo) treat a high role-address ratio as a negative quality signal independent of engagement.

  • One bad segment drags the whole sender reputation.

    Mailbox providers do not evaluate your email in isolation. Reputation is scored on the per-IP-and-domain pair over rolling 30 and 90 day windows. If 5 percent of your list generates bounces and complaints, that segment poisons placement for the other 95 percent, and the recovery window can take weeks once the damage is done.

What is hiding in your list?

Upload a sample. We will return a labeled CSV plus a summary by risk category, ready to feed back into your ESP suppression list. 1,000 credits free. No credit card. Credits do not expire.

Clean your list free

Twice a month. Real changes only.

New questions, deliverability shifts at Gmail and Outlook, and any tool that landed since the last issue. No fluff, no upsell.

Every one of these lists had been called clean. The spam folder disagreed. That’s not a mystery. It’s what a validation can’t see.

Your numbers can look like these. Upload the list. I’ll sign the report.

Yanna-Torry Aspraki, sitting at a desk, looking at the camera. She’s the founder.

Yanna-Torry Aspraki

Founder, Review My Emails

Copying this almanac isn't permitted. Contributing to it is welcomed and rewarded.
We credit the experts who help us improve answers, we link back to their work, and we put their name in front of our readers. We work with people, not around them.