Photo by Stephen Phillips on Unsplash
There’s a moment in every deliverability investigation that goes like this. You run your campaign through mail-tester.com. You get a 9.7/10. You hit send. Half your list lands in Promotions and the other half lands in spam.
What gives?
Spam score checkers are useful. They are also profoundly misleading if you treat the number as a verdict. They measure what can be measured by reading a single email. They cannot measure most of the things that actually decide whether your message hits the inbox.
Here’s the honest breakdown.
What a spam score checker actually measures
Almost every popular checker — mail-tester, Postmark’s spam check, IsNotSpam, GlockApps’ content tab — is some flavor of SpamAssassin. SpamAssassin is open-source rules engine that’s been around since 2001. It scans an email and runs it against a few hundred regex-based rules. Each rule that matches adds (or subtracts) points. The total gets translated into a 0–10 score.
The rules look at things like:
- Header sanity. Does the From address match the Return-Path? Are dates valid? Is the Message-ID well-formed?
- Authentication. SPF pass/fail. DKIM signature present and valid. DMARC alignment. (More on this in a minute.)
- Content red flags. ALL-CAPS subject lines, excessive exclamation marks, the word “free” in the subject, certain HTML patterns, broken HTML, suspicious link patterns.
- Link-to-text ratio. A two-line email with twenty links scores worse than a five-paragraph email with two.
- Image-only content. If your email is one big image with no text, that’s a classic spammer pattern.
- Blacklist hits. Your IP or domain showing up on Spamhaus, SORBS, Barracuda, or URIBL.
These are real signals. Real spammers really do send broken HTML and link-stuffed messages from blacklisted IPs without authentication. Catching them is genuinely useful.
But there’s a generation gap.
The three things spam score checkers do well
1. Auth verification. If your SPF record is missing, your DKIM signature is broken, or your DMARC alignment fails, every checker on earth will catch it. This is the highest-value thing they do. Fixing these gaps moves the needle more than anything else you can change without altering your list.
If you don’t know whether your auth is set up correctly, run our free Email Authentication Checker — it’ll show you SPF, DKIM, and DMARC status in one click, plus a record generator if DMARC is missing.
2. Header hygiene. Broken headers are a strong spam signal because legitimate ESPs almost never produce them. Bad timestamps, missing Message-IDs, mismatched From/Return-Path — these get flagged correctly.
3. Blacklist surfacing. Most spam testers cross-reference your sending IP against the top RBLs. If you’re on one, you find out. (Whether you can do anything about it is another article.)
That’s the value. Those three things, well-checked, are worth the 30 seconds it takes to run a message through.
Now the gap.
The four things they can’t see
1. ISP filtering is no longer rule-based. This is the big one. Gmail filters using a machine-learning model trained on billions of signals. Outlook uses its own ML model plus SmartScreen. Yahoo, AOL (now both under Yahoo Mail), Apple — same deal. SpamAssassin-style rule scoring captures part of what these models see. It does not capture the part where Gmail looks at how many people who got an email like this in the last 6 hours marked it as spam. That signal doesn’t exist in your test send. It only exists in the real send.
The implication: passing every rule-based check tells you the email is technically clean. It tells you nothing about whether Gmail’s classifier will trust it.
2. Engagement signals. Gmail and Outlook care a lot about whether the recipients of past emails from your domain opened them, replied, marked them important, dragged them out of spam, or marked them as spam. None of this is visible in a test send. If your domain has a track record of senders complaining, no spam score checker will tell you. You’ll learn from your inbox placement.
3. List quality. A spam score checker reads one email. It doesn’t see your list. It can’t see that 20% of the addresses are dead, that 3% are spam traps, that you bought a chunk from a “B2B contacts” vendor last year. These are the things that actually generate hard bounces and spam complaints — the two signals that destroy sender reputation faster than anything else.
We’ve seen people get a 10/10 mail-tester score and then watch their domain reputation collapse over the next 14 days from a single bad list import. Score didn’t matter. The list killed them.
4. Warm-up state and sending history. A new IP sending 50,000 emails in a single afternoon is a different animal from an established IP doing the same. Gmail and Outlook both bake your sending history into their classifier. Spam score checkers can’t see your sending history. They look at the email in isolation.
Why a 10/10 score still lands in spam
Now you can name the gap.
A spam score checker tells you whether your email is clean. The inbox decision is made on whether your sender is trusted. Those are different questions.
You can have a perfectly clean email — auth in place, no red-flag content, no blacklist hits — and still land in spam if:
- Your domain has a complaint history. (Past sends, not this one.)
- Your IP is too new and you’re sending too much.
- Your list contains spam traps and you just tripped one.
- Recipients on past sends marked you as spam.
- Your engagement is low (long open windows, mass deletion without opening).
- You imported a list you bought.
None of those show up in a SpamAssassin scan. They are precisely the things that decide deliverability.
How to use a spam score checker without lying to yourself
Spam score checkers are a pre-flight checklist, not a delivery prediction. Use them this way:
- Run every campaign through one before send. Catches auth regressions, broken templates, accidental blacklist hits. Quick and cheap.
- Treat 8.0+ as the floor, not the goal. Above 8.0, the remaining points are usually trivial (a single keyword in the body). Below 8.0, fix something real.
- Don’t chase 10/10. The score difference between 9.5 and 10/10 has almost no impact on actual inbox placement. The score difference between 6/10 and 8/10 is significant.
- Pair it with seed inbox tests. GlockApps, MailReach, and others run your message through actual mailboxes at Gmail, Outlook, Yahoo, etc. and report where it landed. That’s a closer proxy for real deliverability than any rule-based score.
- Pair it with list hygiene. This is the one most people skip and the one that matters most.
That last point is the whole game. A clean list with a 7/10 score will out-deliver a dirty list with a 10/10 score every single time. Reputation > content.
A spam score checker is a checklist. Deliverability is a system.
We’re building a spam score checker of our own — same rules engine you’d expect, with one difference: it’ll tell you when your score is fine but your list is the actual problem. Watch this space.
But you don’t need to wait. The single highest-leverage thing you can do for deliverability today isn’t getting a better score. It’s getting a cleaner list.
Verify your list, remove the dead addresses, get rid of the role accounts you should never have been emailing, and watch your bounce rate drop. That single move resets your sender reputation faster than any content tweak.
Verify your first 100 emails free → — no credit card, no commitment. The honest fix for the problem your spam score checker can’t see.
