What’s the difference between spam score and placement test?

Spam score and placement test answer two different questions. One asks "how does my email look on paper?" The other asks "where does it actually land?"

Spam score: a content audit

A spam score runs your message through a rulebook. The classic tool is Apache SpamAssassin, which checks roughly 750 rules covering content patterns, header structure, authentication results, URL reputation, and HTML quirks. Each rule has a point value. The points add up. If you cross the threshold (default 5.0), SpamAssassin tags the message as spam.

Other tools follow the same pattern. Mail-Tester gives you a score out of 10. GlockApps Spam Tester runs SpamAssassin plus Symantec and Barracuda checks. Postmark's Spam Check API exposes the SpamAssassin score directly.

What a spam score tells you:

• Did you forget DKIM signing?
• Is your HTML-to-text ratio weird?
• Are you using a phrase that pattern-matches to known spam ("act now," "100% free," excessive exclamation marks)?
• Does your From domain align with your Return-Path?

What it does not tell you: whether Gmail, Outlook, or Yahoo will actually deliver to the inbox. Those mailbox providers run their own filters, and they weight reputation signals that no public scoring tool can see.

Placement test: a delivery audit

A placement test (also called an inbox placement test or seed test) sends your real message to a panel of seed mailboxes at Gmail, Outlook, Yahoo, AOL, iCloud, and others. It then reports where each copy landed: inbox, promotions tab, spam folder, or missing entirely.

GlockApps, Mailgun's Inbox Placement, Litmus, and Validity Everest all run this style of test. You get a percentage breakdown per provider. Gmail 87% inbox, 13% spam. Outlook 100% inbox. Yahoo 0% inbox, 100% missing. That last one means your IP or domain is probably blocked outright, which a spam score will never reveal.

Why the two disagree

Mailbox providers care more about who you are than what you sent. Gmail's postmaster documentation is explicit: domain reputation, IP reputation, authentication, and recipient engagement (opens, replies, "not spam" clicks) drive placement. SpamAssassin cannot see any of that. It does not know your complaint rate at Gmail. It does not know how many of your subscribers archived your last campaign without opening it.

This is why a message can pass SpamAssassin with a 0.1 score and still hit the spam folder for 40% of Gmail users. Reputation tanked, content was fine, the rulebook had nothing to say about it.

The reverse happens too. A message scores 4.8 on SpamAssassin (lots of links, image-heavy, marketing copy) but lands in the inbox because the sending domain has years of clean history and a high engagement rate.

For more on what filters actually weigh, see what signals do spam filters look at and how filters decide if an email is spam.

How to use both

Run a spam score check before you send a new template. Fix the easy stuff: missing authentication, broken HTML, link shorteners, dodgy phrasing. Aim for under 3.0 on SpamAssassin, under 5 missed points on Mail-Tester.

Then run a placement test on a real send (or a representative seed send) to see what providers actually do with you. If Gmail says spam folder, that is a reputation problem. Open Google Postmaster Tools, check your domain and IP reputation, and start fixing engagement, not subject lines.

Related reading: how filters evaluate sender behavior and spam filtering vs reputation systems.

Use the spam score to catch content mistakes. Use the placement test to catch reputation mistakes. They are not interchangeable, and skipping either one leaves a blind spot.