Email spam trigger words are mostly a myth in 2026. Gmail, Outlook, and Yahoo do not blacklist individual words like "free" or "act now." They evaluate authentication records, sender reputation, recipient engagement, and overall message context. Writing your cold email around a list of forbidden words is a distraction from the factors that actually determine inbox placement. Below is what filters care about today, plus the 100 historical "trigger words" so you can stop worrying about them and fix the things that move the needle.
Short answer: No single word will send your email to spam. Modern filters decide inbox placement based on SPF, DKIM, DMARC, domain reputation, engagement history, and overall content patterns. The 2002-era list of trigger words still circulates because it is cheap content to republish, but avoiding "guarantee" or "discount" in 2026 will not move you out of spam if your authentication or reputation is broken.
Why the Old Trigger Word List Is Mostly Myth
The list of spam trigger words you have seen on a hundred marketing blogs traces back to SpamAssassin rules written in the early 2000s. SpamAssassin is an open-source filter that scored emails by adding points for matched patterns: the word "viagra" added points, the phrase "act now" added points, a missing reply-to address added points. If the total crossed a threshold, the email was flagged. That was a rule-based, deterministic system.
Modern filtering is different. Gmail processes hundreds of billions of messages per day. Outlook is similar. Both use machine learning models that evaluate dozens of signals in combination, weighted by what the recipient typically engages with. A single keyword has almost zero filtering weight unless it appears in combination with other signals (poor reputation, no authentication, mass-send patterns, low engagement). The word "free" appears in legitimate marketing emails from Apple, Spotify, and your bank every single day. They land in the inbox because their authentication and reputation are healthy.
The reason the trigger word list still circulates is that it is easy to write about, sounds authoritative, and gives readers a sense of control. "Avoid these 100 words" feels actionable. "Build a long-term sender reputation" does not. The truth is closer to the latter.
What Actually Matters in 2026
Here are the four factors that decide inbox placement, ranked roughly by weight.
Authentication (SPF, DKIM, DMARC)
Since February 2024, Gmail and Yahoo have required SPF, DKIM, and DMARC for bulk senders. Microsoft has been moving in the same direction. If your DNS records are missing or misconfigured, your email is treated as untrusted regardless of content. Authentication is the first signal evaluated and the most binary: it either passes or it does not. Fix this before you touch anything else.
Sender Reputation
Every sending domain and IP accumulates a reputation score over time at each mailbox provider. Gmail tracks it via Postmaster Tools. Microsoft tracks it via SNDS. The score is built from spam complaint rate, bounce rate, sending volume consistency, and the engagement your past emails generated. A domain with months of clean sending and high engagement can use the word "guarantee" in every email and still hit the inbox. A two-week-old domain sending cold outbound at high volume will hit spam even with the most polished prose.
Recipient Engagement
Modern filters watch how recipients interact with your past messages. Opens, replies, archives, stars, and forwards are positive signals. Deletes without opening, marks as spam, and no-engagement-then-delete are negative signals. If your past sends to a specific recipient or to similar recipients (same domain, same segment) generated low engagement, future messages get filtered more aggressively. Engagement is also why list hygiene matters: stale lists with high non-response rates train the filter to treat you as low-value.
Content Patterns and Density
Content does play a role, but not at the single-word level. Filters look at patterns: ratio of text to images, density of links, link reputation (are you linking to a domain on a blocklist?), presence of tracking pixels from shared domains, HTML-to-plain-text ratio, attachment types, and structural cues like a missing unsubscribe link. A single instance of "free trial" in a 90-word email is invisible. A 600-word HTML email with 12 links to shortened URLs, no plain-text version, and three images is suspicious regardless of which words it contains.
Spam score in modern email filtering is the probabilistic output of a machine learning model, not a sum of keyword-rule points. The model weighs authentication state, sender history, recipient relationship, and dozens of content signals. A high score means the message looks similar to past messages the recipient (or recipients like them) treated as spam. It is not a checklist.
When Words Actually Do Matter
There are edge cases where content language genuinely affects placement. Knowing these helps you stop worrying about the wrong things.
- Density and stacking. One use of "act now" is invisible. Twelve uses across a short email, combined with ALL CAPS, three exclamation points, and a giant red button, is a pattern that matches known promotional templates. The stack matters, not any individual word.
- Phishing-shaped language. Phrases that mimic banking, security alerts, or government notices ("Your account has been suspended", "Verify your identity now", "IRS final notice") trigger phishing-specific classifiers, especially if the sending domain does not match the brand being impersonated.
- Regulated industries. Words associated with pharmaceuticals, gambling, cryptocurrency, adult content, and certain financial products carry more weight because those categories have high historical spam rates. "Casino" in a cold sales email from a SaaS founder is still fine. "Casino bonus winner" in a multi-link HTML blast from a fresh domain is not.
- Subject lines under heavy scrutiny. Subject lines get more filtering weight than body content because they are shorter and more signal-dense. A subject line that screams promotional ("FREE!!! Limited time!!!") will hurt you more than the same phrase buried in paragraph three of the body.
Reality check from running cold campaigns: in our experience, swapping "free" for "complimentary" or removing the word "guarantee" from a sequence has produced no measurable change in inbox placement when authentication and reputation are healthy. Fixing a missing DMARC record or rotating out a burned inbox produces immediate, visible improvements. Spend time on the latter.
The 100 Historical Trigger Words
For reference, here is the list people are usually referring to when they talk about email spam trigger words. Treat it as historical context, not a checklist. Using any of these in normal cold email copy with healthy infrastructure is fine.
Money and offers (25)
- Free
- 100% free
- Cash
- Cash bonus
- Cheap
- Big bucks
- Earn money
- Make money
- Million dollars
- Pre-approved
- Save big
- Lowest price
- Best price
- Discount
- Promotion
- Special offer
- Limited offer
- Bonus
- Investment
- Loans
- Credit
- Refinance
- Mortgage
- Consolidate debt
- Get out of debt
Urgency and pressure (20)
- Act now
- Limited time
- Hurry
- Don't delete
- Order now
- Order today
- Apply now
- Buy now
- Buy direct
- Get started now
- Click here
- Click below
- Sign up free
- Subscribe
- Visit our website
- Once in a lifetime
- Final notice
- This won't last
- Expires today
- While supplies last
Guarantees and trust (20)
- Guarantee
- Guaranteed
- 100% satisfied
- Risk free
- No risk
- No catch
- No obligation
- No fees
- No hidden costs
- No cost
- Money back
- Certified
- Promise you
- Trusted
- Approved
- Endorsed
- Award winning
- Bestseller
- Top seller
- Number one
Emotional and hyperbolic (15)
- Congratulations
- You're a winner
- Winner
- You have been selected
- Amazing
- Incredible deal
- Unbelievable
- Miracle
- Perfect
- Free gift
- Free sample
- Free trial
- Free quote
- Free consultation
- Free preview
Lifestyle and regulated categories (20)
- Dear friend
- Dear sir or madam
- Hi friend
- Work from home
- Be your own boss
- Online business
- Online degree
- Online pharmacy
- Increase sales
- Increase traffic
- Search engine listings
- Lose weight
- Weight loss
- Diet
- Viagra
- Cialis
- Prescription
- Casino
- Gambling
- Lottery
Look at this list with fresh eyes. "Discount", "free trial", "limited offer", "approved", "award winning" - these are words that appear in legitimate marketing emails from major brands every single day. They land in the inbox. The list is a relic.
How to Actually Test Inbox Placement
Stop guessing whether a word will trigger spam. Test it.
- GlockApps or MailReach seed tests. Send a test email to a pool of seed inboxes at Gmail, Outlook, Yahoo, and others. The report shows exactly where each copy landed: inbox, promotions, spam, or rejected. Run this before launching any new campaign template.
- Mail-tester.com. Free spot check that gives a score out of 10 and explains which specific signals are weak. Useful for catching authentication issues and major content problems quickly.
- Google Postmaster Tools. If you send enough volume to Gmail recipients, Postmaster Tools shows your domain reputation, IP reputation, spam rate, and feedback loop data directly from the source. This is the closest thing to a ground truth signal you can get.
- Microsoft SNDS. The equivalent for Outlook and Hotmail recipients. Shows complaint rate, trap hits, and reputation tier per IP.
Spending an hour configuring these is worth more than a year of rewording subject lines.
If you have time to do one thing: fix authentication. Add SPF, DKIM, and DMARC records to every sending domain and verify they pass with a test send to mail-tester.com.
If you have time to do two things: add a continuous email warm-up service on every sending inbox so reputation builds in the background even on weeks when you do not run active campaigns.
If you have time to do three things: clean your list with an email verification service before every send, accepting that 10-20 percent of any cold list will be invalid and removing them protects your bounce rate.
What to Do if You Are Hitting Spam Now
If your cold email is landing in spam today, the fix is almost never in the copy. Work the diagnostic stack in this order.
- Check authentication. Run a test send to mail-tester.com. Confirm SPF, DKIM, and DMARC all pass. If any fail, fix that first.
- Check blacklists. Run your sending domain and IP through MXToolbox. If you are on Spamhaus, SORBS, or any major blacklist, request delisting and stop sending from that domain until the listing clears.
- Check reputation tools. Look at Postmaster Tools and SNDS. If domain reputation is "low" or "bad," you need to back off volume, clean your list aggressively, and rebuild over weeks, not days.
- Check sending behavior. Are you sending too much volume from a young domain? Did you skip warm-up? Are you blasting 500 emails on a Monday morning from one inbox? Drop daily volume by 70 percent and rebuild from there.
- Then, and only then, look at copy. If everything above is healthy and you are still in spam, look at content density: trim links to one per email, kill heavy HTML, remove tracking pixels from shared domains, simplify formatting. This is the last knob to turn, not the first.
If you do this in order, you will almost always find the problem in step 1, 2, 3, or 4 long before you reach step 5. The word "free" in your second paragraph is almost never the issue.
Frequently Asked Questions
Will using the word "free" send my email to spam?
No. "Free" appears in legitimate emails from major brands every day and lands in the inbox. Modern filters do not blacklist individual words. The word will only contribute to a spam classification if it appears alongside many other risk signals (poor authentication, low domain reputation, high content density, many links) which together form a recognizable spam pattern.
Do spam trigger words still matter at all?
They matter in two narrow situations. First, when they stack at high density across a single email (twelve promotional phrases plus all caps plus three exclamation marks). Second, in subject lines, which get more filtering weight than body content because they are shorter and more signal-dense. Outside those cases, single-word avoidance is a waste of time.
Why do these trigger word lists still appear on marketing blogs?
Because they are easy content to write and they sound authoritative. "Avoid these 100 words" feels like actionable advice a reader can immediately apply. The reality, which is to fix authentication and rebuild sender reputation over weeks, is harder to write engagingly and harder for readers to act on. The lists persist because they sell, not because they work.
Is there a tool that tells me if my email will hit spam?
Yes. GlockApps and MailReach run seed tests across Gmail, Outlook, Yahoo, and other providers and report exactly where your message landed. Mail-tester.com gives a free quick score. Google Postmaster Tools and Microsoft SNDS show ground-truth reputation data for your sending domain. Use these instead of guessing about word choice.
What about subject line trigger words specifically?
Subject lines get more filtering weight than body content. Avoid all caps, multiple exclamation points, promotional stacking ("FREE!!! Limited time!!! Act now!!!"), and obvious phishing patterns ("Your account has been suspended", "Verify identity"). A natural subject line that happens to contain "free" or "discount" is fine. A subject line built entirely from promotional triggers is not.
How quickly can I recover if I am hitting spam now?
If the issue is authentication or a recent blacklist event, recovery can take 24 to 72 hours once the underlying problem is fixed. If the issue is sender reputation built up over weeks of bad sending behavior, recovery takes 2 to 6 weeks of disciplined, low-volume, high-engagement sending plus continuous warm-up. There is no shortcut. The fastest path is usually to pause the burned domain, warm a fresh secondary domain in parallel, and migrate sending while the original recovers.