Mail-Tester for Cold Email: How to Score 10/10 (and Why Cold Email Needs More)

Mail-tester.com is the most popular free email deliverability scoring tool on the internet — they get over 60,000 monthly searches just on their brand name. Score out of 10 in 30 seconds. Free, no signup, instant feedback.

For cold email, it's a useful first-pass check but not a sufficient one. Mail-tester catches obvious DNS, authentication, and content-pattern errors. It doesn't catch the issues that actually determine whether cold email lands in primary inbox vs spam at scale — IP reputation, domain reputation history, real-world inbox placement across multiple receivers.

This guide is the cold-email-specific reading of mail-tester: how to interpret the score, what each deduction category means, what mail-tester misses, and how to fill those gaps.

TLDR — mail-tester for cold email in 5 bullets:

• Mail-tester is a setup checker, not an outcome checker. It catches obvious DNS, authentication, and content-pattern errors in 30 seconds. It does not measure reputation or real inbox placement.
• Target: 9.5+/10. Below 8 means real setup issues. Above 9.5 means setup is fine; the rest of deliverability is about reputation.
• Do not overfit. Stripping all links/images/formatting to chase 10/10 destroys real-world reply rate. Score for setup correctness, not for SpamAssassin gymnastics.
• Pair with /tools/email-deliverability-test for DNS + DNSBL + TLS coverage, and with Google Postmaster Tools for ongoing reputation.
• ColdRelay-provisioned domains pass 9.5+/10 out of the box because SPF/DKIM/DMARC, PTR, MX, and TLS are auto-configured on isolated Azure tenants.

Run mail-tester for setup verification, then /tools/email-deliverability-test for the broader scan.

Table of Contents

• The 30-second answer
• How mail-tester actually works
• Reading the score for cold email
• Common cold email deductions and fixes
• What NOT to do — overfitting is a trap
• Mail-tester vs /tools/email-deliverability-test
• Does 10/10 predict inbox placement?
• Beyond mail-tester: the full picture
• How ColdRelay's infrastructure scores on mail-tester
• FAQ

The 30-second answer

Mail-tester scores your email out of 10 based on roughly 7 categories:

Target score for cold email: 9.5/10 or higher. Anything below 9 has something fixable.

Run mail-tester, or use ColdRelay's free deliverability test which runs the same DNS checks plus blocklist scanning across 6 major DNSBLs.

How mail-tester actually works

Three steps:

1. You visit mail-tester.com. They generate a unique random email address (e.g., test-abc123@srv1.mail-tester.com).
2. You send your cold email — from the actual mailbox and sending setup you'd use in production — to that address.
3. You go back to the mail-tester page and see your score plus a per-category breakdown.

The test runs in the background on their server. They check the headers of the message you sent, run SpamAssassin against the body, query DNS for SPF/DKIM/DMARC on your sending domain, look up the From address's PTR record, and check any URLs in the body against their list of known spam-affiliated domains.

The output is a 10-point score with per-category deductions explained line-by-line. Each deduction is actionable — you know exactly what to fix.

Reading the score for cold email

9.5–10/10: solid setup

What it means: your DNS, authentication, and basic content patterns are correct. You're cleared on the obvious setup errors.

What it doesn't mean: that your cold email will inbox-place at Gmail / Outlook / Yahoo. Mail-tester confirms the setup is right; it doesn't measure reputation history or real inbox classification at major providers. A 10/10 mail-tester score with bad Postmaster Tools Domain Reputation = your mail still goes to spam.

8.0–9.5/10: minor fixable issues

What it means: one or two specific deductions. Read the breakdown — usually a single fix gets you to 9.5+.

Common single-deduction causes for cold email:

• SpamAssassin: URI_HEX — your message contains a URL with a hex-encoded path. Don't use shortened URLs or hex-encoded tracking links.
• SpamAssassin: HTML_FONT_LOW_CONTRAST — your HTML uses gray-on-gray text that's hard to read. Cold email should be text-first; remove the gray styling.
• SpamAssassin: MISSING_HEADERS — your message is missing standard headers like Date: or Message-ID:. Common when sending via custom SMTP without a properly-configured sending tool.
• DKIM not signing all messages from this domain — your sending tool's DKIM config is misaligned with what's published in DNS.

Below 8.0/10: real setup problems

Indicates multiple deductions or major failures. Common pattern:

• SPF missing or ~all instead of -all
• DKIM not signing at all
• DMARC at p=none or absent
• Multiple SpamAssassin content triggers (URL shorteners, suspicious phrases, image-heavy HTML)

For cold email, anything below 8.0 means you should fix the setup before sending real campaigns. The score below 8 reliably predicts spam-folder placement at major providers.

Common cold email mail-tester deductions and fixes

URIBL_BLOCKED — your link is on a blocklist

Your message contains a URL whose domain is on URI blocklist (URIBL, SURBL, or similar). Common causes:

• Tracking domain that another customer also uses got listed
• Shortened URL pointing to a known spam destination
• Calendar link domain (Calendly, etc.) that's been flagged

Fix: replace shortened URLs with full URLs to your own domain. If your tracking domain is shared (typical of bundled-SMTP setups), move to a dedicated tracking domain.

RDNS_NONE — your sending IP has no reverse DNS

Receiving servers check that your sending IP has a PTR (reverse DNS) record pointing back to a hostname. Missing PTR = RDNS_NONE deduction.

Fix: configure PTR on your sending IP. ColdRelay-provisioned mailboxes have PTR records configured automatically (pointing at mail.<your-domain>).

MISSING_HEADERS — message lacks standard SMTP headers

Your sending tool isn't including all the standard headers (Date:, Message-ID:, MIME-Version:, etc.). Usually a tool-configuration issue, not a sending-setup issue.

Fix: configure your sending tool to add the missing headers. Most modern sending platforms (Instantly, Smartlead, Lemlist) do this automatically; older or custom SMTP setups might miss them.

HTML_IMAGE_RATIO_* — too many images relative to text

Your message body is mostly images with little text. SpamAssassin reads this as marketing-email pattern.

Fix: cold email should be text-heavy with minimal or no images. If you need an image (a screenshot, a chart), keep it small and balance with substantial body text.

BODY_8BITS — non-ASCII characters in the body

Your message uses extended characters (em-dashes, smart quotes, accented characters) without proper encoding.

Fix: ensure your sending tool sends UTF-8 encoded with Content-Type: text/plain; charset=utf-8 header. Modern tools do this correctly by default.

What NOT to do — overfitting to mail-tester is a trap

The biggest mistake cold email senders make with mail-tester: optimizing the score at the cost of real-world reply rate. The SpamAssassin rules behind the score were tuned against marketing-email patterns from the early 2010s, and several of them actively penalize cold-email best practice. Chasing 10/10 by stripping signals that real humans actually want destroys the campaigns the score is supposed to protect.

Do not strip every link to avoid URI_* rules. A short cold email with one calendar/booking link is normal and outperforms zero-link campaigns in real reply rate. SpamAssassin will sometimes ding a single legitimate link; the deduction is cosmetic, the link is the call-to-action that produces revenue. Keep the link. Move on.

Do not remove all images everywhere. Some onboarding/welcome messages benefit from a small product screenshot. SpamAssassin's HTML_IMAGE_RATIO_* rules will tick up; the deduction is small, the screenshot is valuable. For pure cold outreach the right answer is still text-only — but for warm follow-ups, do not gut the message to hit 10/10.

Do not rewrite plain-English copy to avoid every spam-word. Mail-tester flags certain phrases ("free trial," "limited time," "guarantee," "click here"). The right call: drop genuinely-spammy patterns; keep words that natural cold email actually uses. "I'd love to show you" is not spam. SpamAssassin sometimes thinks it is. Trust the human read.

Do not optimize for 10/10 on a single test send. Reading the per-category breakdown is more valuable than the headline score. A 9.2/10 with three small SpamAssassin nudges is better than a 9.8/10 reached by gutting the message. The setup-level checks (SPF, DKIM, DMARC, PTR, MX) are the ones that actually matter — score those at 10/10, accept 0.5 deduction on content for a message a human would actually read.

Do not run mail-tester repeatedly until you hit 10/10. Each retest from the same domain sends another message into mail-tester's seedboxes. At cold-email volumes this is fine, but obsessive retesting is wasted engineering time. Hit 9.5+, ship the campaign, monitor real Postmaster Tools and reply-rate signals instead.

The right mental model: mail-tester is a smoke test for the setup. Pass the smoke test, then optimize against actual sender-recipient signals, not against SpamAssassin's heuristics.

Mail-tester vs ColdRelay's /tools/email-deliverability-test

Both tools answer the question "is my setup right" — but they overlap, not duplicate. The honest comparison:

The split is roughly: mail-tester is the deepest SpamAssassin-rule scanner you can get for free, and /tools/email-deliverability-test is the deepest DNS + DNSBL + TLS scanner you can get for free. Most cold email senders should run both — they take a combined 35 seconds and cover non-overlapping ground.

Use mail-tester when you need to know whether your message content triggers content-pattern filters. Use email-deliverability-test when you need to know whether your domain setup and IP reputation are clean — including ongoing blocklist exposure that mail-tester doesn't scan for. For full programmatic coverage, also consider the free DKIM generator, SPF generator, and DMARC generator.

Does a 10/10 mail-tester score predict inbox placement? The honest answer

It does not. A 10/10 mail-tester score predicts that your setup is correct. Whether you actually land in primary inbox is a function of three other signals that mail-tester cannot measure:

1. Domain reputation history. A domain that's been sending well-behaved mail for 12 months will inbox-place even on a 9.0/10 message. A brand-new domain at 10/10 will land in spam more often. Google Postmaster Tools' Domain Reputation column is the canonical signal here — mail-tester can't see it.

2. IP reputation history. Same logic at the IP layer. A new IP from a clean ASN inboxes better than a recycled IP with a Spamhaus history, regardless of the mail-tester score. The Postmaster Tools IP Reputation view tells you this; mail-tester doesn't.

3. Per-recipient engagement. Gmail, Outlook, and Yahoo all weight whether the specific recipient has engaged with your domain before. A 10/10 message to a stranger lands in spam more often than a 9.0/10 message to someone who has replied to you twice this quarter. Mail-tester runs in isolation — they have no engagement history with your domain.

Empirical reality check: at ColdRelay we've seen domains pass mail-tester at 10/10 and still land in spam consistently — because Postmaster Tools showed Low Domain Reputation from a prior bad campaign. We've also seen domains at 9.2/10 inbox-place at >90% because they had clean reputation and high recipient engagement. The score correlates with placement but does not predict it.

The cleaner mental model: mail-tester score sets a floor. Below 8.0/10 you'll miss inbox even with perfect reputation. Above 9.5/10 your setup isn't the bottleneck — reputation and engagement are.

Beyond mail-tester: the full deliverability picture

Mail-tester is one layer. The complete cold email deliverability stack involves at least five signal sources, each measuring something mail-tester can't:

1. Google Postmaster Tools. Daily Domain Reputation (High/Medium/Low/Bad), Spam Rate, IP Reputation, authentication pass rates. The ground truth for Gmail placement. (How to read it →)

2. Microsoft SNDS. Outlook's equivalent. Color-coded IP reputation buckets, complaint rate trend, filter classification ratio. Required reading if your prospect base includes Outlook addresses.

3. Blocklist monitoring. Mail-tester sometimes flags a URIBL hit on a body link; it doesn't scan your sending IP across the major DNSBLs. Pair it with /tools/blacklist-checker or read the full IP blacklist guide and the MXToolbox blacklist guide for the broader scan.

4. Seed-list inbox placement. Real test sends to a panel of 70+ seed inboxes across providers. The inbox placement tester is the closest free equivalent to GlockApps.

5. Bounce rate and complaint rate at scale. Mail-tester sees a single test message; the actual reputation lever is your aggregate behavior over thousands of sends. Watch bounce rate below 1% and complaint rate below 0.1% on your production campaigns.

For ColdRelay customers, items 1, 3, and 5 are folded into the dashboard with automated alerts. The mail-tester layer remains a useful smoke test, but it stops being a primary monitoring tool the moment you're sending at real volume.

What mail-tester doesn't tell you

Mail-tester is a setup checker, not an outcome checker. It misses:

1. IP reputation history. Mail-tester runs in isolation — they don't have your sending IP's historical reputation context. An IP that's been sending spam for 6 months still tests 9.5/10 if the message itself is clean.

2. Real inbox placement. Mail-tester's "score" doesn't predict whether your message will land in primary inbox at Gmail or Promotions or Spam. That depends on reputation + content classification by each receiver, neither of which mail-tester measures.

3. Per-provider variation. Gmail, Outlook, Yahoo, Apple Mail each have different spam classifiers. Mail-tester gives one score; the providers might rate you very differently.

4. Engagement signals. Replies, opens, time-spent-in-inbox — all measured by inbox providers and weighted into reputation. Mail-tester sees the single test message; it doesn't see your sending pattern over time.

5. Volume + warmup state. A brand-new domain with 100% perfect mail-tester scores will still get filtered to spam if you start sending 1,000+ messages/day from day one. Mail-tester doesn't know what volume you're sending at.

How to fill the gaps

Mail-tester catches setup errors. For everything else, you need:

Google Postmaster Tools — Gmail's own view of your reputation. Updates daily, shows Domain Reputation (High/Medium/Low/Bad), Spam Rate, Authentication pass rates. (How to read it →)

Microsoft SNDS — Outlook's equivalent. Color-coded IP reputation, complaint rate, filter classification.

Hourly blocklist monitoring — DNSBLs (Spamhaus, Barracuda, etc.) can list you anytime. Mail-tester is a one-off snapshot; you need continuous monitoring. ColdRelay's hourly DNSBL checks across 6 major lists alert on listings within an hour. (Full blocklist guide →)

Seed-list inbox placement tests — Real test sends to a panel of 70+ seed inboxes across providers. GlockApps is the gold standard; ColdRelay folds daily seed tests into the base subscription.

Bounce rate monitoring — Mail-tester doesn't see your bounce rate. Cold email's healthy bounce rate is below 1%; above 2% degrades reputation regardless of mail-tester score. (Bounce rate breakdown →)

A cold email mail-tester routine

When to use mail-tester:

Pre-launch on every new domain. Send one test, hit 9.5+/10, fix any deductions before turning on real campaigns. 5 minutes well spent.

After any sending-setup change. Switched to a new sending tool, changed mailboxes, modified SPF/DKIM — re-run mail-tester to confirm the change didn't break anything.

When debugging a placement drop. If reply rates dropped and you don't know why, mail-tester is the cheapest first check. If the score is still high, the problem isn't setup — it's reputation or content.

Not useful as ongoing monitoring. A mail-tester score doesn't change unless your setup changes. Running it weekly tells you the same thing each time. Use Postmaster Tools, SNDS, and blocklist monitoring for ongoing watch.

How ColdRelay's infrastructure shows on mail-tester

Every ColdRelay-provisioned domain passes mail-tester at 9.5+/10 out of the box. Specifically:

• ✓ SPF record with -all (hard fail)
• ✓ DKIM signing with 2048-bit RSA, signature verifies
• ✓ DMARC at p=quarantine, properly aligned
• ✓ MX record pointing at the dedicated mail server
• ✓ PTR record matching the sending hostname
• ✓ TLS on every SMTP connection
• ✓ Dedicated IPs (no shared-pool reputation drag)

The 0.5 deduction that occasionally appears is content-specific (a particular URL, a SpamAssassin rule on the body) — sending-setup-side we're at 10/10. (How we configure SPF/DKIM/DMARC →)

FAQ

Why is my mail-tester score 8.5 but my cold email still goes to spam?

Mail-tester catches setup errors; it doesn't measure reputation. The remaining 1.5-point deduction is fixable, but even at 10/10 you'd still hit spam if your domain reputation is Low in Postmaster Tools or your IP is on a blocklist. Check those next.

Should I run mail-tester from every mailbox or just one per domain?

One per domain is usually enough — the DNS-level checks (SPF, DKIM, DMARC, MX, PTR) are domain-level, not mailbox-level. The SpamAssassin check is content-level. If your mailboxes share the same sending-tool configuration, they'll get the same score.

Mail-tester says my DKIM is broken but my sending tool says DKIM is configured. What gives?

Common situation. Causes:

• Your sending tool is signing with a key that doesn't match the public key in your DNS.
• The DKIM selector (e.g., default._domainkey) in DNS doesn't match what your sending tool is using.
• DNS hasn't propagated yet (rare; usually resolves within an hour).

Check dig +short TXT default._domainkey.yourdomain.com and compare the public key against your sending tool's configured private key.

What's a 10/10 cold email message look like?

Plain text, no images, one or two short paragraphs, a question at the end, no URL shorteners, no spam-trigger phrases, properly-signed by DKIM, aligned with DMARC. The setup matters more than the content gymnastics — once your setup is right, sub-1.5-point deductions are usually fine.

Is mail-tester biased toward marketing email?

Slightly — its content checks were tuned against marketing-email patterns and some rules (image ratio, font contrast) don't quite fit cold email. Still useful; just don't optimize away cold email's natural patterns (short text, conversational tone) chasing a marketing-email-shaped 10/10.

Does mail-tester store my test email?

They keep test results for 1 week, then delete. The actual email body content is visible to anyone with the test URL (which is generated randomly and not publicly listed) during that week.

Mail-tester is the right tool for the question "is my cold email setup obviously broken?" It's not the right tool for "will my cold email reach the inbox?" — that depends on reputation and provider classification, neither of which mail-tester measures.

Run the free deliverability test (DNS + blocklist + TLS in 30s) → /tools/email-deliverability-test · Cold email infrastructure that passes mail-tester at 9.5+/10 out of the box → Try ColdRelay free