You can’t stop spam emails in Gmail completely. Gmail’s filter already blocks close to 15 billion unwanted emails every day, more than 99.9% of the spam, phishing, and malware aimed at its users, by Google’s own count, and your inbox still isn’t clean. What’s left is the residue that survived all that, and the filter that caught everything else can’t touch it, because it doesn’t look like spam. Here’s how to cut the residue down, and what to do about the part that won’t cut.
Which method stops which mail
The table runs in order of payoff, top to bottom. If one specific kind of mail is driving you up the wall, find its row and skip to that section.
| Method | Stops | Doesn’t stop |
|---|---|---|
| Report spam | Mass spam patterns, for you and everyone | One-to-one mail that looks legitimate |
| Block sender | One persistent address | The same sender on a new address |
| Unsubscribe | Legitimate lists you signed up for | True spam (it confirms you’re alive) |
| Filters | Any pattern you can describe in a rule | Senders who change patterns weekly |
| Protect your address | Future list-building against you | Lists your address is already on |
1. Report spam: the button that trains Gmail
Gmail’s spam filter learns from reports. Google’s help pages are explicit about this: as you report more spam, Gmail identifies similar emails as spam more efficiently. Every report is a vote against that sender, for you and for everyone else.
On desktop: select the message (checkbox, don’t open it), then click the Report spam icon (the stop sign with an exclamation mark) in the toolbar.
On mobile: open the message, tap the three-dot menu in the top right, tap Report spam.
Do this instead of deleting. Deleting tells Gmail nothing; reporting moves the message to Spam and trains the filter. Everything in Spam is deleted automatically after 30 days, and when you report a message, Google receives a copy and may analyze it for abuse protection. If a sender is impersonating a company or asking for credentials, use Report phishing (same menu) instead: it routes the report to a different, more aggressive system.
2. Block the senders who keep showing up
Blocking sends every future email from that specific address straight to Spam.
On desktop: open the message, click the three-dot menu next to Reply, choose Block “sender”.
On mobile: open the message, tap the three-dot menu next to the sender’s name, tap Block.
Blocking works per email address, not per company. Serious spammers and cold-outreach tools rotate addresses constantly (jake@companyname.com today, jake@trycompanyname.com next week, then jake.smith@companyname.io), so your block list only ever covers the addresses they’ve already burned. Block the persistent individual senders; don’t expect it to slow a rotating operation. We wrote a full guide to what blocking does and doesn’t do.
3. Unsubscribe: from lists that are actually legitimate
For newsletters, retail promotions, and anything you at some point signed up for (knowingly or not), unsubscribing works better than it used to. Since February 2024, Google has required anyone sending more than 5,000 messages a day to Gmail accounts to support one-click unsubscribe and keep their spam complaint rate below 0.3%, or their mail stops getting delivered at all. That requirement is the Unsubscribe link next to the sender’s name at the top of the message, and it’s the reason legitimate senders now honor it fast: their deliverability depends on it.
Only unsubscribe from senders you recognize. Clicking “unsubscribe” in a true spam email can confirm to the spammer that your address is alive and actively read, which makes it more valuable. Our guide to unsubscribing safely covers how to tell a legitimate list from a spammer fishing for confirmation.
4. Build filters for the patterns you keep seeing
Some senders survive both reporting and blocking: the domain changes, the pitch doesn’t. Filters are the Gmail feature built for them. You write a rule that matches on sender, subject, or keywords, and every message that matches gets deleted, archived, or labeled before you see it.
To create one: click the gear icon → See all settings → Filters and Blocked Addresses → Create a new filter. Enter your criteria (for example, from:(@spammy-domain.com) or subject words like “exclusive offer”), click Create filter, and pick an action: Delete it or Skip the Inbox are the useful ones for spam.
The weakness of filters is maintenance. You’re writing static rules against senders who change domains, subject lines, and wording every week: from:(@spammy-domain.com) dies the week that sender re-registers as spammy-domain.io, the same move that defeats blocking. Keep up and you’re back in settings every couple of weeks, maintaining a small rules engine as a hobby. Fine if you enjoy it. If you’d rather not, tools exist that maintain the rules for you (more on that below).
5. Stop leaking your address in the first place
Everything above deals with spam after it arrives. The cheaper fix is staying off the lists: one leaked signup can seed the broker lists spammers buy from, and nothing on this page removes you from a list you can’t see.
- Use a secondary address for signups, receipts, and anything commercial. Set it to forward to your main inbox if you want one place to check, or just skim it once a month.
- Use Gmail’s
+trick:yourname+shopping@gmail.comdelivers to your normal inbox but tells you exactly who sold your address. When a tag starts drawing junk, create a filter matchingto:(yourname+shopping@gmail.com)with Delete it (same click-path as section 4) and the whole leak dies at once. - Don’t post your address in plain text on public pages. Scrapers harvest those constantly.
- Check whether your address appears in data breaches (haveibeenpwned.com). You can’t remove an address from a breached list, so a hit means two things: change the password on the breached account, and expect more spam and phishing at that address from now on, which is a good reason to move future signups to a fresh one.
Why spam keeps coming back anyway
Do everything above and your spam volume drops fast, then stops dropping. What survives is a steady trickle: expect a few cold emails a week and the occasional phishing attempt to keep landing, however many rules you write.
Why the floor? Three reasons:
Your address is on lists that get resold. Once an address is on one broker’s list, it spreads. No amount of blocking removes you from lists you can’t see.
Cold outreach isn’t spam, technically. The steady drip of “quick question,” “following up,” and recruiter mail comes from real people at real companies using tools built to land in inboxes: authenticated domains, clean sending reputations, personalized first lines. Gmail’s filter is looking for mass-mail fingerprints, and this mail has none. It sails through while actual junk gets caught.
Phishing is designed around Gmail’s checks. Fake invoices and lookalike login pages are crafted specifically to beat them, and the volume is industrial: Kaspersky alone blocked over 554 million attempts to follow fraudulent email links in 2025, and that’s one security vendor counting only what it caught. Americans reported losing $12.5 billion to fraud in 2024, a 25% jump over the prior year, and email was the most common way scammers made first contact. Gmail catches most phishing. The misses are the messages that can empty an account.
The automated option: a second filter behind Gmail’s
Cold outreach and phishing get through by not looking like spam, and that’s the mail we built Sieve to catch. It’s a second filter behind Gmail’s own: cold outreach is labeled Solicitor, phishing is labeled Malicious, and both are archived out of your inbox before you see them. Labeled, not deleted: you can audit everything and drag anything back.
You’re handing a tool access to your mail, so here is exactly what Sieve does with that access:
- Mail from people you know (your domain, your contacts, anyone you’ve actually corresponded with) comes straight through, and its contents are never read by AI. The Privacy Policy walks through exactly what is and isn’t stored.
- It learns your patterns instead of making you write rules: drag a message it filed back to your inbox and the correction is recorded, so future classifications improve.
- Sieve stores metadata only, never message bodies, and nothing is used to train models.
- It also includes Smart Unsubscribe: recurring list mail gets grouped into one queue where you unsubscribe or keep each list, no footer hunting.
It’s Gmail-only, from $10/month (15% off yearly), with a 7-day free trial. If the free methods above get you where you want to be, do that; the rest of our email guides go deeper on each one. If you want to shop around, we compared the major inbox-cleanup tools head to head and wrote a standalone look at SaneBox, the closest thing to a default choice in this category. Sieve is for people who did all five and still spend every morning hand-clearing the trickle.
Sources
- Kaspersky Securelist: Spam and phishing report for 2025 (spam = 44.99% of global email traffic; 554M fraudulent link attempts; 144M malicious attachments)
- Google: New Gmail protections for a safer, less spammy inbox (99.9% blocked; ~15 billion unwanted emails/day; bulk-sender rules)
- Google: Email sender guidelines (5,000 messages/day threshold; 0.3% spam-rate limit; one-click unsubscribe)
- Gmail Help: Mark or unmark emails as spam (reporting trains the filter; Spam auto-deletes after 30 days)
- Gmail Help: Block or unsubscribe from emails (blocked senders go to Spam)
- FTC: New data show a big jump in reported fraud losses to $12.5 billion in 2024
- FTC: Consumer Sentinel Network Data Book 2024 (email the most-reported fraud contact method)