An email spam checker scores a test message, or the headers and body you paste into it, against content rules and authentication checks, usually in the style of SpamAssassin. Treat it as a lab test. Do not treat it as a prediction. A clean score does not put you in the inbox, because the factors that matter most, authentication, complaint rate and sending reputation, are either DNS records you can verify yourself or numbers only the mailbox provider sees. This covers what the score measures, what Gmail, Yahoo and Outlook publish, how to check each record directly, and a seven-row pre-send checklist with an order of operations for when you are already in spam.
What an email spam checker does
Most checkers use one of two patterns. Some give you an address and ask you to send a test message, then report on what arrived. Others take pasted source, or a subject line and body, then analyse that content. Either way, the report has the same shape: a list of tests that fired, a point value for each test, and a total score.
The tests split into two groups. Content tests inspect the words, links, HTML and formatting. Authentication tests inspect the headers: whether SPF passed, whether a DKIM signature exists, and whether the From domain lines up. That is the whole job. The checker inspects one message, from one sending path, at one moment.
The caveat is simple: a checker tells you whether that message trips common filters. It does not tell you how Gmail, Yahoo or Outlook judge your domain after thousands of sends and a pile of recipient reports.
How the spam score is worked out
SpamAssassin is the reference point for most scoring tools, and its documentation describes the scoring model plainly. Every named test gets a score, the number of points added when that test hits. Add the points, and once the total reaches required_score the message is considered spam. The SpamAssassin configuration docs give the default as 5.
The useful detail is in the same documentation. The docs call 5.0 "quite aggressive" and say it suits a single-user setup. They suggest ISPs set something more conservative, like 8.0 or 10.0. Even the people who wrote the tool do not treat 5 as a universal boundary.
A few other details from that page matter for how you read a report:
- A test with no configured score gets 1.0, except tests whose names start with
T_(rules in testing), which get 0.01. - Setting a rule's score to 0 disables it.
- A test can carry up to four scores, chosen by whether Bayes and network tests are enabled (score sets 0 to 3).
Put those facts together and the operational point becomes clear. Two installs running the same rules can score the same message differently, and a tool can disable rules or change their weights. I cannot tell you what any given checker has configured, and that is the problem. When one tool shows a low number and another shows a high one for the same email, they are not necessarily disagreeing. They are different instruments. Compare a score only with the same tool's own scale, and never read any threshold as a promise of inbox placement.
What Gmail, Yahoo and Outlook actually require
The providers publish their own requirements, and they do not match each other exactly. Quote them separately. There is no single "industry standard" here, only three sets of requirements with a lot of overlap.
| Sender group | Authentication and DNS | Spam rate | Unsubscribe | Source |
|---|---|---|---|---|
| Gmail, all senders (from February 1, 2024) | SPF or DKIM for sending domains; valid forward and reverse DNS (PTR) | Below 0.3% in Postmaster Tools | Not required at this tier | Gmail sender guidelines |
| Gmail, more than 5,000 messages a day | SPF and DKIM plus DMARC (policy can be none); From domain aligned with SPF or DKIM domain | Below 0.30% in Postmaster Tools | One-click unsubscribe plus a clearly visible link in the body for marketing and subscribed mail | Gmail sender guidelines |
| Yahoo, all senders (enforcement from February 2024, rolled out gradually) | SPF or DKIM at minimum; valid forward and reverse DNS for sending IPs | Below 0.3% | Not required at this tier | Yahoo sender best practices |
| Yahoo, bulk senders | SPF and DKIM; valid DMARC policy of at least p=none that must pass; relaxed alignment acceptable | Below 0.3% | Functioning list-unsubscribe header supporting one-click (RFC 8058 Post method highly recommended, mail-to acceptable); visible link in body; unsubscribes honoured within 2 days | Yahoo sender best practices |
| Outlook.com consumer addresses, domains sending over 5,000 emails a day | SPF must pass; DKIM must pass; DMARC at least p=none and aligned with SPF or DKIM (preferably both) | No figure given in the announcement | Functional unsubscribe link listed as a recommendation | Microsoft announcement (April 2025) |
Two things in that table need a closer look.
First, Outlook's enforcement changed. The original plan was to route non-compliant mail to Junk after May 5, 2025. The April 29 update struck that out: Microsoft decided to reject instead, with the message "550; 5.7.515 Access denied, sending domain [SendingDomain] does not meet the required authentication level", taking effect May 5th. For that group, a failure is a bounce, not placement in the spam folder.
Second, Gmail's own pages disagree on spam rate. The sender guidelines say below 0.3%. The Top 10 Gmail sender issues page tells senders to keep it below 0.1%.10% and avoid 0.30% or higher, especially for sustained periods. Treat 0.3% as the line you must not cross and 0.1% as the target.
The Gmail FAQ also says what happens when you miss. A spam rate above 0.3% or a missing DMARC record (minimum p=none) means delivery support or mitigations are unavailable. Failures of alignment, SPF plus DKIM, forward and reverse DNS, TLS or RFC 5322 format lead to temporary or permanent failure codes, or spam foldering.
The authentication checks, one at a time
These are the checks you can verify yourself, directly, without trusting a score.
SPF
SPF is a DNS record that lists which servers may send for your domain. Gmail says the record should include every sender for the domain, and mail from third-party senders missing from it is more likely to be marked as spam. In a checker report, a failure appears as an SPF fail or softfail line. In raw headers, read the Authentication-Results header for spf=pass or a different result.
One common trap: Microsoft's FAQ says exceeding 10 DNS lookups in an SPF record might cause SPF to fail. Every include: for a newsletter tool, CRM and helpdesk uses part of that allowance. Run the domain through HostList's SPF checker and it will show what the record actually says, not what you think you configured.
DKIM
DKIM signs the message so the receiver can verify that the domain owner sent it. Failure appears as dkim=none (no signature) or dkim=fail (signature broken, often because something modified the message in transit). Key length matters too. Gmail requires 1024 bits or longer for personal Gmail accounts and recommends 2048 if your domain provider supports it. Yahoo recommends a minimum 1024-bit key.
You need the selector, which sits in the s= tag of the DKIM-Signature header, to look up the public key. HostList's DKIM checker performs that lookup.
DMARC alignment
This is where teams get caught. SPF passing and DKIM passing do not mean DMARC passed. DMARC needs at least one of them to pass and the authenticating domain to match the domain in the visible From header. The Gmail FAQ says only one of SPF or DKIM needs to be aligned for the bulk-sender requirement, and alignment isn't required for forwarded or mailing list messages. Yahoo accepts relaxed alignment. Outlook would prefer both aligned.
The classic failure looks like this: an email platform sends on your behalf, SPF passes for the platform's own domain, DKIM is signed with the platform's domain, and neither domain matches your From address. Everything says "pass" and DMARC still fails. A missing record is also a failure. Check with HostList's DMARC checker, then confirm in a real message's Authentication-Results line (dmarc=pass).
If you need to read the raw headers of a test message, paste them into HostList's email header analyzer and inspect the authentication results line by line. If you are drafting the three records from scratch, this walkthrough on generating SPF, DKIM and DMARC with AI, then checking them covers the checking half properly.
Reverse DNS (PTR)
Forward and reverse DNS must both exist. Gmail wants the sending IP to match the IP of the hostname in the PTR record. Yahoo wants PTR records that are valid, meaningful, non-generic, reflect your domain name, and do not look like a dynamically assigned IP. The usual culprit is a default hosting-provider PTR that reads like a numbered machine name.
Gmail points to the Google Admin Toolbox Dig tool for checking PTR. You usually cannot edit a PTR yourself, because the party that controls the IP block controls the PTR, normally your host or ISP. The fix is usually a support ticket. If you are moving mail hosts, HostList's MX record lookup confirms which host currently receives your mail before you change anything.
Blocklists: what a listing means
A blocklist entry means an operator has flagged an IP address or domain as a source of spam or abuse, and receiving servers that query that list may treat your mail more harshly. Gmail says that if you use a shared IP you should make sure it isn't on any internet blocklist, because messages from blocklisted IPs are more likely to be marked as spam.
The important part is what you can't do about a listing. Spamhaus's SBL FAQ states that removals are governed by its removals policy only, and SBL removals are the sole decision of The Spamhaus Project. There is never any charge, and an offer to remove a listing for a fee is a scam. Deliverability consultants have no special privilege to influence, expedite or modify SBL listings. SBL removal requests must come from the ISP in charge of the listed IP. If you're on shared hosting, that's your host, not you.
PBL is different. It lists IP space that shouldn't send email directly to the internet, and per Spamhaus, PBL listings don't prevent sending unless the user's email program isn't authenticating correctly (SMTP Authentication). A PBL hit on a home or office connection often points to a settings problem, not a reputation problem.
To check, run the sending IP and domain through HostList's blacklist checker, and use Spamhaus's own lookup at check.spamhaus.org, which shows IPs and domains currently listed. After a removal, Spamhaus suggests waiting 1 to 2 hours for networks to catch up (general FAQ). The SBL zone itself is rebuilt and reloaded every 5 minutes, so the delay mostly comes from other networks' caches.
One-click unsubscribe headers
If you send marketing or subscribed mail in volume, this is a hard requirement. Gmail asks for one-click unsubscribe plus a visible link in the body once you pass 5,000 messages a day. Yahoo asks bulk senders for a functioning list-unsubscribe header. Both point at RFC 8058 (January 2017).
The mechanics are narrower than people expect:
- Put exactly one
List-Unsubscribeheader and oneList-Unsubscribe-Postheader in the message. - The
List-Unsubscribeheader must contain one HTTPS URI (it may also carry others, likemailto:). - The
List-Unsubscribe-Postheader must contain the single pairList-Unsubscribe=One-Click. - The message must carry a valid DKIM signature that covers at least those two headers (they must appear in the
h=tag). - The sender must not return an HTTPS redirect.
The RFC's own example pairs these header lines: List-Unsubscribe: <https://example.com/unsubscribe/opaquepart> and List-Unsubscribe-Post: List-Unsubscribe=One-Click. (The URI there is my reconstruction to match the POST path in the same example; check the RFC's section 8.1 for the original.) The receiving system then sends POST /unsubscribe/opaquepart HTTP/1.1 to Host: example.com, with Content-Type: application/x-www-form-urlencoded, Content-Length: 26, and the body List-Unsubscribe=One-Click.
The RFC uses POST for a reason. Anti-spam software often fetches every resource in a mail's header fields automatically, and a sender has no mechanical way to tell that fetch from a human click. A plain link would unsubscribe people by accident.
Gmail's FAQ is clear on two traps. A one-click link that lands on a web page doesn't comply with RFC 8058. Gmail also doesn't automatically reject or spam-folder messages that fail this requirement, but unwanted mail without it is more likely to be reported as spam, and only bulk senders meeting every requirement are eligible for mitigation. A spam checker may never inspect these headers at all. Read them in the raw message.
What a spam checker cannot see
A checker analyses a message. Most of what decides placement lives outside the message.
- Complaint rate. Gmail and Yahoo both set 0.3% as the ceiling, and both measure it on their side. Yahoo notes that its spam rate is calculated on mail delivered to the inbox, which matters when you compare it with Complaint Feedback Loop data or work the figure out in your own system.
- Domain and IP reputation. Gmail says user spam reports can lower a domain's reputation over time. No test message carries that history.
- Engagement history. Yahoo's advice is to send to an engaged audience, remove invalid recipients promptly and watch inactive ones. That is a list problem, not a copy problem.
- Per-recipient filtering. What lands in one person's inbox is not what lands in another's. Outlook's FAQ adds that its Safe Sender list won't be honoured for the new enforcement, so the old "ask them to whitelist us" fix doesn't cover authentication failures there.
For Gmail, use Postmaster Tools. Per Google's Top 10 page, it shows spam rates, authentication rates, delivery errors and domain and IP reputation. There is one catch: if your volume to personal Gmail accounts is too low, the dashboards might not include data for your domain. Small senders often have no spam-rate number to inspect. In that case, put more weight on the records you can verify.
Outlook's FAQ puts the issue in one line: authentication proves the email is from you, but user perception still matters. If recipients report your mail despite passing everything, review content, frequency and the opt-out process.
The pre-send checklist
These are the seven checks I'd run before a campaign or a host move. The last column is the honest part: whether a spam checker can see the item at all.
| # | Check | What it catches | Free way to verify | Can a spam checker see it? |
|---|---|---|---|---|
| 1 | SPF | Missing sender in the record, record over 10 DNS lookups | HostList's SPF checker | Partly: only the result for the test message's sending path |
| 2 | DKIM | No signature, broken signature, key under 1024 bits | HostList's DKIM checker, plus the DKIM-Signature header | Partly: it sees the signature on one message |
| 3 | DMARC alignment | No DMARC record, From domain not matching the SPF or DKIM domain | HostList's DMARC checker, plus the Authentication-Results header | Partly: it sees one message, not your policy history |
| 4 | Reverse DNS / PTR | Missing, generic or mismatched PTR | Google Admin Toolbox Dig tool | Depends on the tool |
| 5 | Blocklist status | Sending IP or domain listed | HostList's blacklist checker; check.spamhaus.org | Depends on the tool; it is a lookup, not a message test |
| 6 | One-click unsubscribe headers | Missing List-Unsubscribe-Post, no DKIM coverage of those headers, redirect or landing page | Read the raw headers of a real test message | Only if the tool has a rule for it |
| 7 | Content score | Spammy phrasing, link and HTML patterns | A SpamAssassin-style report | Yes, this is its core job |
Notice what is absent: complaint rate. No row can cover it, because only the provider has the number.
If you are already in spam
Work in this order. It goes from cheapest and most binary to hardest.
- Authentication first. Confirm SPF, DKIM and DMARC pass and align on a real message, not just in DNS. Fix missing third-party senders, short keys and the PTR record.
- Blocklists next. Check the sending IP and domain. If it's an SBL listing on an IP your host controls, the removal has to come through them, and nobody can pay it away.
- List hygiene and complaint rate. Pull inactive and invalid addresses, confirm opt-ins, add one-click unsubscribe, and check Postmaster Tools if your volume is high enough to show data.
- Content last. Only after the above, work through the content tests in a checker report.
Content comes last because it is the easiest to tinker with and the least likely to be the actual cause.
Frequently Asked Questions
Do I need DMARC if I send fewer than 5,000 emails a day?
Gmail's hard requirement at that volume is SPF or DKIM, with DMARC required from the 5,000 a day tier up. Gmail still recommends setting up SPF, DKIM and DMARC for your domains regardless, and Microsoft's FAQ says all senders benefit from the same practices even though enforcement first targets large senders. A p=none policy is enough to start.
My SPF, DKIM and DMARC all pass. Why am I still in spam?
Passing authentication only proves the mail is from you. Gmail says user spam reports lower a domain's reputation over time, and Outlook's FAQ says to review content, frequency and the opt-out process if recipients keep reporting you. Check complaint rate and list quality next.
Should I use a 2048-bit DKIM key?
Gmail requires 1024 bits or longer for sending to personal Gmail accounts and recommends 2048 if your domain provider supports it. Yahoo recommends a minimum of 1024 bits. If your provider supports 2048, use it. If it doesn't, 1024 meets both providers' stated minimums.
How long after a Spamhaus removal will my mail recover?
Spamhaus says to wait 1 to 2 hours after a recent removal for networks to catch up. That only clears the listing itself. If the cause was complaints or poor list quality, the underlying reputation problem remains.
A spam checker score is a lab result for one message. The records behind it, SPF, DKIM, DMARC alignment and PTR, can be verified directly. The number that matters most, your complaint rate, is visible only to Gmail and Yahoo, with 0.3% as the line to stay under.
Follow HostList for new rankings, original research, and changes across the hosting industry.



