Spam Trigger Words in Cold Email: What to Avoid (And What Really Gets You Flagged)

Search for “spam trigger words” and you will find lists of 500+ forbidden terms — free, guarantee, act now, limited time. The implication is that one wrong word sends your cold email straight to junk. The reality in 2026 is more nuanced: modern spam filters are machine-learning systems that weigh hundreds of signals, and word choice is one of the smaller ones. But smaller is not zero, and certain patterns still reliably hurt B2B senders. Here is what actually matters, what does not, and how to write cold emails that clear the filters.

How Spam Filters Actually Score Your Email

Gmail, Outlook and corporate gateways do not maintain a simple blacklist of words. They score each message on a blend of signals, roughly in this order of weight:

1. Sender reputation — your domain and IP history, complaint rates, and spam-folder placements from past sends.
2. Authentication — whether SPF, DKIM and DMARC pass.
3. Engagement — do recipients open, reply, and move your mail out of spam, or delete and report it?
4. List quality — bounce rates and spam-trap hits.
5. Content — structure, links, formatting, and yes, wording.

This is why a trusted sender can write “free trial” and land in the inbox while a new domain gets flagged for a plain-text message. If your deliverability fundamentals are weak, no word list will save you; if they are strong, an occasional risky phrase rarely matters.

Content Patterns That Still Get B2B Senders Flagged

Within the content signal, filters look for combinations that resemble known spam. The patterns that still cause trouble:

• Money-plus-urgency stacking: “Guaranteed ROI — act now — limited spots.” One of these phrases is fine; three in one email pattern-matches to spam.
• ALL CAPS and multiple exclamation marks in subject lines or openers.
• Link-heavy messages: more than one or two links in a short cold email, especially link shorteners (bit.ly and friends are heavily penalized).
• Image-only or HTML-heavy emails with little real text — keep cold emails plain-text-like.
• Deceptive subject lines: “RE:” or “FWD:” on a first touch. Filters catch it, and recipients report it, which is worse.
• Attachment on first contact. Unsolicited PDFs and docs trip corporate gateways.

Words to Use With Care (Not Ban Outright)

Rather than memorizing 500 words, watch these categories:

• Financial promises: guarantee, risk-free, double your, no obligation, cash, $$$.
• Artificial urgency: act now, urgent, expires today, last chance, don’t miss out.
• Overclaiming: #1, best-ever, revolutionary, miracle, breakthrough.
• Pharma/lottery vocabulary: anything that sounds like the spam folder you already know.

Each is a small negative weight. The fix is usually better copy anyway: “guaranteed 10x pipeline” is both spammy and unbelievable, while “our last three SaaS clients booked 12–18 meetings in the first month” is specific, credible, and filter-safe.

A 5-Point Pre-Send Content Check

1. Read the email aloud. Does any sentence sound like a late-night infomercial? Rewrite it as a plain statement of fact.
2. Count links — one maximum in a first touch, none is even better. Ask for a reply instead, as we recommend in our cold email outreach guide.
3. Check the subject line: lowercase or sentence case, no punctuation stunts, relevant to the body. More on this in our subject line rules.
4. Strip heavy formatting — no images, no buttons, minimal bold.
5. Send a test to a seed inbox on Gmail and Outlook and see where it lands before launching the campaign.

Fix the Bigger Levers First

If your emails are hitting spam, word choice is rarely the root cause. Audit in this order: authentication records, domain age and warmup status, bounce rate, sending volume per inbox, and reply rate. A verified list and a properly warmed-up inbox move placement far more than deleting the word “free.” Once the foundation is solid, clean copy stops being a risk factor and starts being what it should be: the reason a prospect replies.