A Subscriber Marked You as Spam: How to Get Back to the Inbox

A click on “Report spam” costs a sender far more than an unsubscribe. An unsubscribe loses you one person. A spam complaint damages your domain's reputation and starts pulling everyone else's mail into the spam folder too.

If your open rate is sliding for no obvious reason and a red complaint-rate bar just appeared in Postmaster Tools, you're already in this situation. Here's how to get out of it.

What happens when a subscriber clicks “Spam”

The mechanics are simple. Someone sees your message in their inbox, doesn't recognize the sender (or recognizes you but is tired of the frequency), and clicks “Report spam.” The mailbox provider logs it as a complaint. The message moves to spam and the complaint is added to your sending domain's statistics.

One complaint by itself means nothing. What providers watch is the share of complaints among messages that reached the inbox. Google asks senders to keep it under 0.1%, fewer than one complaint per thousand messages, and never to reach 0.3%. Past that point it cascades: mail lands in spam, open rate drops, the provider gets another signal that your messages are unwanted. A closed loop.

Yahoo publishes a 0.3% ceiling for every sender. Microsoft's SNDS FAQ calls a complaint rate under 0.3% a good bar for an IP. Mail.ru scales its limit with volume: 1.1% for up to 10,000 messages a month, 0.3% above 50 million. Google, for its part, warns that a lower spam rate takes a while to show up in how your mail gets filtered.

FBL: how to find out who is complaining

FBL stands for Feedback Loop. It's the mechanism through which mailbox providers report complaints back to the sender. Not every provider supports it, and not every ESP configures it automatically. Without FBL you're flying blind: complaints accumulate and you have no idea.

The report format is ARF, defined in RFC 5965. Each report is an email in three parts: a human-readable note, machine-readable fields, and a copy of the original message. Providers may redact the recipient address from that copy for privacy reasons. RFC 6449 suggests the same trick senders use for bounces with VERP: embed your own subscriber identifier in the message somewhere a provider is unlikely to strip it. Then a complaint still maps to a contact when the address is gone.

ESPs usually enroll your domain and suppress complainants for you. Yahoo's FAQ says as much and tells senders to confirm the enrollment with their provider. If you run your own SMTP server, FBL needs manual setup. Without it, complainants keep receiving your messages and keep filing complaints.

A practical rule: if you can see a complaint rate in Postmaster Tools but don't know who is complaining, FBL is only partially configured. Sort out Yahoo, Microsoft and Mail.ru first, since they report each complaint. Gmail won't give you names, but the aggregate view and the Feedback-ID breakdown are enough for monitoring.

Complaint addresses: what to do with them

A complaint address belongs to a subscriber who hit “Spam” or “Junk” on at least one of your messages. You learn it from an FBL report or from your ESP's stats. Gmail doesn't hand out such a list at all.

There is exactly one rule. The address goes on your suppression list right away, automatically, with no “one last send” first. That person has already told their mailbox provider they don't want your mail, and the next message can easily bring a second complaint. Unsubscribes are neutral for reputation. A complaint reads differently: either there was no consent, or the mail annoyed someone enough that they went to their provider instead of your unsubscribe link.

The expensive mistake happens during a migration. Switching ESPs or moving to your own SMTP? Export the suppression list and load it into the new system before the first send. Skip that, and everyone who ever complained hears from you on day one.

uChecker can't see the complaint history of an address; that data stays with mailbox providers. You or your ESP keep the complainant list. Address verification handles the neighboring problem (dead mailboxes, dead domains, disposable addresses), which hurts the same reputation.

Why subscribers complain (even ones who opted in themselves)

The uncomfortable truth: people click “Spam” not because you're actually a spammer. They click it because it's easier than hunting for an unsubscribe link. Or because they forgot they ever signed up. Or because they're simply fed up.

There's another scenario: the person never subscribed at all. The address entered your database through scraping, a purchased list, or an automatic CRM import without consent. Those contacts complain at an extremely high rate. If your list contains addresses that never went through an opt-in, that's a ticking bomb.

What complaint rate is and how it's calculated

Complaint rate (Postmaster Tools calls it spam rate) is the share of your messages that recipients reported as spam. The denominator is narrower than most people assume. It isn't your list size, and it isn't every delivered message either. Google divides complaints by messages that landed in the inbox of active recipients, and only counts DKIM-authenticated mail. Yahoo also works from inbox-delivered mail and reminds senders of that when they compute the rate from their own CFL reports.

Say 10,000 messages reached the inbox and 30 people complained. That's 0.3%, right at the red line. Unsubscribes stay out of the formula: 2,000 of them on the same campaign wouldn't move the number at all.

The formula has an odd side effect, and the Postmaster Tools help spells it out. Once Gmail is filtering most of your mail to spam, hardly anyone sees it in the inbox, so hardly anyone complains, and the rate looks low. A low spam rate next to a poor domain reputation is a warning, then.

What complaint rate is acceptable

Postmaster Tools recalculates spam rate daily, so a single bad campaign shows up on the chart right away, however clean things looked before. After you fix the cause, Google says to allow up to seven days for the rate to get back within limits, and spam classification catches up later still.

Other providers give you less to go on. Yahoo states its ceiling but points senders to CFL reports for tracking complaints, and Microsoft shows a complaint rate in SNDS only for IPs you've proven are yours. On an ESP's shared IP you won't see it, so you're reading indirect signals: rejections that mention reputation, rising soft bounces, falling opens (Apple Mail Privacy Protection skews them) and falling clicks.

Step-by-step plan: how to get back to the inbox

If your complaint rate is already in the red zone, work through these steps in order. Skipping steps is like patching holes while the ship is sinking.

How to stay out of spam going forward

Recovery drags on; prevention costs less. Here is what helps.

Special case: flagged as spam by corporate mail filters

B2B mail sometimes ends up in spam not because subscribers complained but because of corporate gateway filters. Barracuda, Mimecast, Proofpoint: large companies run their own gateways with their own rules, and Postmaster Tools won't show you what they decided.

A different set of actions helps here. Check your domain against blacklists (MXToolbox, MultiRBL). If your domain or IP is listed, submit a delisting request: every blacklist has its own procedure, usually a form on their site. Verify that SPF, DKIM, and DMARC are correctly configured. If you send from a shared IP, someone else's behavior may have gotten you listed.

For B2B sends, list validation is especially critical. Corporate addresses often tie to domains that change during reorganizations, contractor switches, or hosting migrations. An address that worked in January may not exist in March.

The short version

The sequence is short: diagnose through Postmaster Tools, clean the list, simplify unsubscribe, reduce frequency, re-warm gradually. Then set up monitoring so you don't end up back at the same point.

Start with what takes five minutes: check your complaint rate in Postmaster Tools and run your list through a validator.