Free tool
Cold email checker
Paste a cold email to see spam trigger words highlighted inline, a 0–100 copy score and the three fixes that matter most. Checks run in your browser as you type. Want a second opinion? Get an AI review and rewrite that keeps your merge tags.
Example email loaded. Paste yours over it.
- Phishing & spam patterns
- Money & pricing
- Hype & overpromising
- Pressure & urgency
- Template-speak & formatting
- ALL CAPS
- Broken tag / shortener
- Merge tag
Copy score
Needs work
4 checks to look at.
- Words
- 98
- Grade
- 4.5
- Links
- 1
- Questions
- 2
Top fixes
- 1Rephrase the 6 highlighted phrases, starting with "money back".
- 2Replace shortened links with the full URL on your own domain.
- 3Rewrite "we/I" sentences around the reader ("you", their team, their problem).
Trigger phrases (6)
Money & pricing
1Money talk in a first touch reads like a promotion and matches classic content-filter rules. Lead with their problem, not your price.
Hype & overpromising
3Superlatives and guarantees cost trust instantly. Specific, modest claims get more replies.
Template-speak & formatting
2Filler openers and stacked punctuation scream mass email. They rarely trip filters alone but kill reply rates.
Structure checks
Trigger phrases — Fail
6 found
Trigger phrases are a tie-breaker for filters and a trust problem for readers. Highlighted in the editor.
Link shorteners — Fail
bit.ly
Public URL shorteners hide the destination and are heavily abused by spammers, so filters and security gateways distrust them.
"You" vs "I/we" — Fail
4 you · 9 I/we
The email is mostly about you and your company. Prospects reply to emails about their situation.
Opt-out line — Warning
missing
People who can't easily opt out hit "Report spam" instead, and complaint rates are a top filtering signal. Many jurisdictions (e.g. CAN-SPAM in the US) also require an opt-out.
Length — Pass
98 words
In the 50–125 word sweet spot for a first touch: long enough to be specific, short enough to read on a phone.
Subject line — Pass
43 chars · 6 words
Short enough to show in full on mobile (~30–40 characters are visible) and reads like a normal email.
Fake Re:/Fwd: — Pass
none
No fake reply or forward prefix.
Links — Pass
1 link
One link is fine. Make sure it's on a domain with a good reputation (ideally your own), not a tracking redirect.
Images & attachments — Pass
none
Plain text, no attachments or images mentioned.
ALL CAPS words — Pass
none
No shouting. Normal acronyms (CEO, SaaS, ROI…) are ignored.
Exclamation marks — Pass
1
One is fine. More starts to sound like marketing.
Call to action — Pass
2 questions
There's a clear ask. Low-friction ("worth a look?") beats high-friction ("30 minutes Tuesday?") on a first touch.
Reading level — Pass
grade 4.5
Easy to read (Flesch–Kincaid grade 8 or below). Busy people skim; plain words win.
Personalization — Pass
{{company}} {{firstName}}
Merge tags present. Remember: a first name alone isn't personalization; a reason this person, now, is.
Broken placeholders — Pass
none
All merge tags look well-formed.
Spintax — Info
none
No spintax. Optional: variations help avoid identical messages at volume.
AI review
Line-by-line feedback and a rewrite that keeps your merge tags.
What the checker looks at
- Trigger phrases
- ~250 curated phrases in five groups: money, pressure, hype, phishing patterns and template-speak. Not an 800-word list that flags "free" in "feel free".
- Length
- 50–125 words is the sweet spot for a first touch. We warn past 125 and fail past 175.
- Subject line
- 2–6 plain words, under ~50 characters, so it shows in full on mobile. Fake "Re:" and "Fwd:" fail.
- Links & shorteners
- Zero or one link in the first email. Public shorteners (bit.ly, tinyurl…) always fail.
- Attachments & images
- Mentions of attachments, file names and embedded images. Strangers' attachments look like malware.
- Shouting
- ALL CAPS words (common acronyms like CEO or SaaS are ignored) and exclamation marks, especially in the subject.
- One clear ask
- At least one question or call to action, and not three competing ones.
- "You" vs "I/we"
- Whether the email is about the reader or about you.
- Reading level
- Flesch–Kincaid grade. Grade 8 or below reads fast; grade 12+ reads like a whitepaper.
- Merge tags & spintax
- Personalization tags present, broken ones like {first_name}, [Company] or {{firstName} caught, and spintax validated. Length is measured on one spin variation.
- Opt-out line
- A plain way to say no, which lowers spam complaints.
The score starts at 100 and loses points per issue, weighted by how much it hurts replies or placement. Broken placeholders, fake replies and shorteners cost the most because a prospect notices them instantly.
What actually decides spam vs. inbox
Spam word lists are the most searched and least important part of deliverability. Mailbox providers decide placement mainly on who you are and how people react to you: domain and IP reputation, SPF/DKIM/DMARC alignment, sending volume and ramp, bounce rate, and above all spam complaints. Google tells senders to keep their user-reported spam rate below 0.1% and never reach 0.3%.
That's why the copy still matters. Hype, fake urgency, "Dear Sir/Madam" and a bit.ly link don't just score points in a content filter. They tell a busy reader this is a blast, and the reader's "Report spam" click is the signal that sinks your domain. Write so a skeptical buyer would believe one person sent it to them on purpose.
Copy scores well and you still land in spam? The problem is upstream. Check the SPF, DKIM & DMARC checker, then work through the cold email deliverability checklist and the deliverability fundamentals.
A first email that gets replies
- Open with them, not you. Skip "My name is" and "I hope this finds you well". Start with the observation that made you write: a hire, a launch, a stack change.
- One problem, one proof point. Specific and modest beats "10x your pipeline".
- Ask one easy question. "Worth a look?" asks for less than "Do you have 30 minutes Tuesday?", and it's easier to reply to.
- Let them say no. "Not relevant? Reply 'no' and I won't follow up." Fewer complaints.
- Vary it at volume. Hundreds of identical messages look like bulk mail. Merge tags plus a few spintax variations in the opener and CTA keep messages distinct.
Starting from scratch? The cold email generator drafts a first touch you can then run through this checker.
FAQ
Do spam trigger words really send cold emails to spam?
Rarely on their own. Gmail, Yahoo and Microsoft weigh sender reputation, authentication and recipient engagement far more than words. Trigger phrases are a tie-breaker for borderline senders, and they make readers more likely to report the email, which damages the reputation that does decide placement.
How long should a cold email be?
50–125 words for the first email: enough for one reason you're writing, one proof point and one question, short enough to read on a phone. Follow-ups can be two or three sentences.
Should I put links in a cold email?
Zero or one in the first touch. Every link is a domain filters check and a reason to click instead of reply. Never use public shorteners like bit.ly; they hide the destination and filters distrust them.
Is it OK to start a cold email subject with "Re:"?
No. A fake "Re:" or "Fwd:" pretends there is an existing conversation. It may lift opens once, but people notice the trick and report it. A plain 2–6 word subject works better.
Does this tool test inbox placement?
No, it checks the copy. Inbox placement depends mostly on domain and inbox reputation, authentication, volume and complaint rate. Use our SPF, DKIM & DMARC checker for the authentication side.
Is my email stored?
The instant checks run entirely in your browser; nothing is sent anywhere. If you click "Get AI feedback", the subject and body are sent to our server and an AI model to generate the review. We don't store them.
Related tools and guides
Try Clement free for 14 days
Good copy is half the job. Clement writes, personalizes and sends cold email campaigns from your own Gmail or SMTP inboxes, so the other half is handled too.
Start free trial