Spam placement is almost never one problem. It is a score, assembled from authentication, sender reputation, recipient engagement, list quality and content — and a single weak input can outweigh four strong ones. This guide works through the nine causes in the order worth checking them, because the cheapest fixes are also the ones people skip.
Authentication → blacklists → list quality → volume → engagement → content.
The first four are binary and fixable today. The last two are gradual and take weeks. Diagnosing them in the wrong order is why people spend a month rewriting subject lines while a broken DKIM record quietly fails every message.
Since 2024, Google and Yahoo require SPF, DKIM and a published DMARC policy from anyone sending in volume. Missing records are not a soft signal — they are grounds for rejection or automatic spam placement, regardless of how good the rest of your sending looks.
The subtler failure is alignment. SPF can pass while DMARC still fails, because
DMARC requires the domain in the visible From: address to match the domain that SPF
or DKIM authenticated. Sending through a third-party platform with its own bounce domain is the
usual cause.
Check: run your domain through the
free SPF, DKIM and DMARC checker, then send a
message to a Gmail account and inspect Show original — you want PASS on all
three lines, not just SPF.
A listing on a major RBL such as Spamhaus overrides everything else. It does not matter how well warmed the mailbox is; some receiving servers reject or quarantine on the listing alone. Shared IPs make this worse — you can be listed for someone else's behaviour on the same infrastructure.
Check: look up your sending IP and domain on Spamhaus, Barracuda and SURBL. Delisting requests are usually free and processed within a few days, but only after the behaviour that triggered the listing has stopped. Warmup will not help until you are delisted.
Hard bounces are one of the fastest ways to damage a sending domain. A bounce rate above roughly 2% signals to providers that you are not maintaining your list, and scraped or purchased lists routinely bounce at 20% or more. Spam traps — abandoned addresses recycled by providers specifically to catch senders using stale data — do worse damage than bounces because hitting one is proof you did not obtain the address legitimately.
Fix: verify every list before sending, remove role addresses
(info@, support@, admin@), and never re-import a list you
bought. No amount of warmup outruns a 15% bounce rate.
A mailbox that sends five messages a day for a month and then sends 400 in one afternoon has produced the clearest spam signature there is. Providers model normal human sending patterns, and a step-change in volume from a low-history sender is treated as a compromised or newly abused account.
| Week | Safe daily volume (new mailbox) | What you are proving |
|---|---|---|
| Week 1 | 10–20 | The mailbox is real and gets replies |
| Week 2 | 20–40 | Sending is consistent, not bursty |
| Week 3 | 40–80 | Engagement holds as volume rises |
| Week 4+ | 80–150 | Sustained reputation under real load |
Those numbers are a floor, not a target. A ramp is defined by the shape of the curve, not the ceiling — doubling weekly is fine, decupling in a day is not. The full ramp schedule and per-provider sending limits go into more detail.
Gmail in particular scores senders on what recipients do: opening, replying, marking as important, moving out of spam. Cold outreach is structurally bad at all of these, which is the core problem — a campaign with a 1% reply rate teaches Gmail that your mail is unwanted even when every technical check passes.
This is the one cause that warmup exists to solve. Automated warmup sends real messages between real mailboxes that open them, reply to them and pull them out of spam, so the engagement average your provider sees is not composed entirely of ignored cold email.
A complaint rate above 0.3% will get you filtered by Gmail; above 0.1% is already a warning sign. Complaints are weighted far more heavily than any content signal, and they persist in your reputation history for weeks.
Fix: make unsubscribing trivial (one click, no login), stop mailing anyone who has not engaged in 90 days, and narrow your targeting. Relevance is a deliverability control, not just a conversion one.
Content matters less than most advice claims, and matters most when reputation is already weak. Individual "spam words" are largely obsolete as a filtering mechanism. What still hurts:
Plain, short, personalised messages with one link to your own domain outperform designed HTML templates for cold outreach, both in placement and in reply rate.
Sending cold outreach from your primary company domain risks the domain your invoices and support replies depend on. The standard pattern is a separate lookalike domain for outreach, forwarded to the main site, with its own authentication and its own warmup.
Equally, a domain registered last week has no history at all. Volume from an unknown domain is filtered aggressively on principle. Let a new domain age one to two weeks after DNS setup before warming it, and another three to four weeks before real outreach.
Reputation is a rolling window, not a permanent score. A mailbox that stops sending for two months does not keep its standing — it drifts back toward unknown, and the next campaign starts from a worse position than the last one ended. This is why warmup is usually worth leaving running at reduced volume between campaigns rather than switching off once placement improves.
If placement has not moved after four weeks of clean sending, the problem is almost always one of the binary items in the first column that was checked carelessly — usually DMARC alignment, or a blacklist listing on a shared IP nobody thought to look up.
Related reading: what email warmup is and how long it takes, the SPF, DKIM and DMARC checker, and how warmup tools compare — Lemwarm, MailReach, Warmup Inbox, Instantly and Mailwarm.
Why do my emails go to spam in Gmail but not Outlook?
Providers score senders independently. Gmail weights recipient engagement heavily, so a mailbox with few replies and opens gets filtered there while Outlook — which leans more on IP and domain reputation plus content heuristics — still delivers. Divergence between providers usually points at engagement rather than authentication, because a broken SPF or DKIM record would hurt everywhere at once.
How long does it take to get out of the spam folder?
Two to six weeks. Authentication fixes apply immediately, but reputation is a rolling average of recent behaviour, so providers need weeks of clean, engaged sending before placement moves. Stopping sending altogether does not reset anything — it just stops the clock.
Does the word "free" in a subject line send email to spam?
Not on its own. Filters are statistical and reputation-weighted, so a trusted sender can use words that would sink an unknown one. Spam-word lists are largely obsolete advice. Content matters most when reputation is already weak, and even then link quality, image-to-text ratio and URL shorteners matter far more than individual words.
Will email warmup alone fix spam placement?
No, and any tool claiming otherwise is overselling. Warmup fixes the reputation and engagement half of the problem. It cannot fix a missing DKIM record, a blacklisted IP, a purchased list generating complaints, or a failing DMARC policy. Fix authentication and list quality first, then let warmup rebuild reputation.
Should I ask recipients to move my email out of spam?
It genuinely helps — a message dragged from spam to the inbox is one of the strongest positive signals a provider records. It just does not scale for cold outreach, which is precisely what automated warmup replicates: a network of real mailboxes that open, reply to and rescue your messages every day.
Fix your DNS records yourself, then let warmup handle reputation. One mailbox, 10 warmup emails a day, permanently free — no card required.
Start free