Spam email is unwanted, irrelevant, deceptive, or abusive email, usually sent at scale. In email deliverability, “spam” can mean both the unwanted message itself and the outcome when a mailbox provider or recipient sends a message to the spam folder. Legitimate senders avoid spam by earning permission, sending relevant mail, authenticating their domains, and making opt-outs easy.

What is spam email?

At its simplest, spam email is email a recipient did not ask for, does not want, or reasonably perceives as misleading, excessive, or unsafe. The classic examples are fraudulent offers, phishing attempts, malware delivery, and bulk messages sent without meaningful permission. But a message does not have to be criminal or obviously deceptive to be treated as spam: a legitimate brand can create a spam problem when it emails people too often, targets the wrong audience, obscures its identity, or makes leaving the list difficult.

That distinction matters. Spam is partly a content and consent issue, but it is also a recipient-experience issue. A customer may have subscribed six months ago and still mark a daily promotion as spam because it is no longer useful, was unexpected, or arrived too frequently. Mailbox providers use those choices, along with technical and behavioral signals, to protect inboxes.

Spam is therefore not synonymous with “marketing email.” A password-reset email, invoice, account-security alert, newsletter, product announcement, or abandoned-cart reminder can all be legitimate. Their chances of reaching the inbox depend on whether they are expected, properly identified, technically authenticated, and sent to an audience that actually wants them.

In the United States, commercial email must comply with the CAN-SPAM Act. The Federal Trade Commission says the law applies to commercial messages broadly, not just bulk mail, and requires accurate header information, non-deceptive subject lines, a physical postal address, a clear opt-out method, and prompt honoring of opt-out requests. Legal compliance is essential, but it is not a guarantee of inbox placement: mailbox providers also make their own filtering decisions based on recipient feedback and sender reputation. (ftc.gov)

Spam versus the spam folder versus a spam complaint

These related terms are often used interchangeably, which makes troubleshooting harder. They describe different things.

Spam email

This is the message category or recipient judgment: unwanted, deceptive, harmful, or low-value email. A mailbox provider may classify a message as spam automatically, even if the recipient never sees it in the inbox.

Spam-folder placement

Spam-folder placement is a delivery outcome. The receiving server accepted the message, but its filtering system placed it in Junk, Spam, or a similar folder instead of the inbox. The message was technically delivered, yet it was unlikely to be seen or acted on.

This is why a high delivery rate does not prove good deliverability. An email platform can report that a message was accepted by a recipient server, while the recipient’s mailbox provider filters it away from the inbox. Delivery measures acceptance; inbox placement measures where accepted mail is likely to appear.

Spam complaint

A spam complaint is a direct negative action from a recipient, such as selecting “Report spam” or “Mark as junk.” It is one of the clearest signals that a recipient did not want the message. Complaint data can be available through a sending provider, a feedback loop, or a mailbox-provider reporting tool, but visibility varies by provider and sending volume.

For Gmail traffic, Google describes spam rate in Postmaster Tools as the percentage of messages delivered to engaged recipients’ inboxes that recipients manually mark as spam. The dashboard is Gmail-specific, is based on DKIM-authenticated traffic, and is not a universal measure of all messages a business sends. (support.google.com)

Blocked or rejected email

A block or rejection is different again. The receiving server may refuse a message during SMTP delivery because of an authentication failure, poor reputation, an IP or domain blocklist issue, suspicious traffic patterns, or a policy violation. A blocked message never reaches the mailbox folder structure at all.

A sender can have all four problems at once: some mail is rejected, some is accepted into spam, some reaches the inbox but is reported as spam, and some continues to perform normally. The right response depends on which condition is happening and at which mailbox provider.

Why spam email matters for deliverability and campaign performance

Spam affects much more than one campaign. It affects whether mailbox providers trust future mail from the same domain, subdomain, IP address, and sending pattern.

Inbox placement and visibility

The immediate effect is obvious: recipients rarely see mail in the spam folder. A campaign may show a healthy server-level delivery rate while generating weak opens, clicks, replies, conversions, or renewals because it is not landing where recipients look.

For transactional messages, the consequence can be even more serious. If a login code, password reset, booking confirmation, invoice, or security alert is filtered, users may believe the product is broken. They may repeat requests, contact support, abandon checkout, or lose trust in the service.

Sender reputation compounds over time

Mailbox providers assess patterns, not only individual messages. Repeated complaints, poor engagement, stale addresses, sudden volume changes, suspicious links, and inconsistent identity can reduce confidence in a sender. Once reputation declines, even mail that is useful and properly formatted may face more filtering or throttling.

Reputation is also contextual. A domain with strong transactional-email behavior can still harm itself by using the same identity for aggressive promotions. Separating streams—for example, receipts, product alerts, lifecycle email, and newsletters—does not create immunity, but it makes traffic easier to control, measure, and isolate when a campaign goes wrong.

Lower conversion and misleading reporting

Spam placement can make campaign reporting misleading. A marketer may conclude that subject lines, pricing, or creative are failing when the real issue is that fewer recipients saw the message in the inbox. Similarly, a product team may read low engagement on an onboarding sequence as lack of product interest when the audience was poorly sourced or overmailed.

Open tracking is also an imperfect diagnostic because privacy features and image loading can distort opens. Use a broader picture: complaints, unsubscribes, bounces, clicks, conversions, replies, inbox-placement tests, authentication results, and mailbox-provider data where available.

Higher operational and compliance risk

Spam complaints can trigger suppression workflows, provider reviews, sending limits, or account restrictions. They also identify a more fundamental problem: the sender’s expectation of the relationship does not match the recipient’s expectation.

Commercial-email rules are not identical worldwide, so legal review should reflect the recipients’ locations and the nature of the program. As a baseline for U.S. commercial mail, CAN-SPAM prohibits false or misleading headers and deceptive subject lines, requires a valid postal address and a clear opt-out mechanism, and gives recipients the right to stop future marketing messages. (ftc.gov)

Is spam a metric? Understanding spam rate and complaint rate

“Spam” is a category and an outcome. Spam rate and complaint rate are metrics used to quantify negative recipient feedback. They are related, but the exact calculation depends on the data source.

A practical complaint-rate calculation

A useful internal campaign metric is:

Complaint rate (%) = (spam complaints / delivered messages) × 100

Use delivered messages rather than attempted messages when your system can reliably distinguish them. A message that hard-bounced never gave its intended recipient a chance to complain, so including it in the denominator can make the complaint rate look artificially low.

Worked numeric example

Suppose a campaign attempts to send 80,000 emails:

  • 2,000 messages hard-bounce or are otherwise not delivered.
  • 78,000 messages are delivered.
  • 39 recipients mark the email as spam.

The campaign complaint rate is:

(39 / 78,000) × 100 = 0.05%

That is 0.05%, or one complaint for every 2,000 delivered messages. The number alone is not a clean bill of health: look at the trend, the list source, the message type, the mailbox-provider mix, and whether complaints are concentrated in one segment.

Gmail’s spam-rate measurement is narrower

Do not expect your platform’s complaint rate to exactly match Gmail’s reported spam rate. Google says its Spam Rate dashboard reflects the percentage of mail delivered to engaged Gmail inbox recipients and then manually reported as spam. Mail that Gmail had already sent to spam is not simply treated as ordinary inbox-delivered mail for this calculation; Google also notes that a message initially placed in spam and later marked “not spam” can count toward inbox-delivered messages. (support.google.com)

The important operational lesson is that every metric has a scope. Your sending platform may calculate complaints from delivered messages across all mailbox providers. Gmail’s view applies to eligible Gmail traffic. A seed-list tool estimates placement for a small test audience. None is a complete substitute for the others.

What rate should senders target?

Treat 0.3% as a ceiling to avoid, not as a performance target. Google’s sender guidance says bulk senders should keep spam rates below 0.3%, and its FAQ says bulk senders are ineligible for mitigation while the user-reported spam rate exceeds that level. Yahoo likewise says senders should keep spam rates below 0.3%. (support.google.com)

A prudent program aims materially lower than that threshold. A rate can rise suddenly after a list import, reactivation campaign, frequency increase, misleading subject line, or sending-identity change. Waiting until a visible threshold is crossed means waiting until recipients have already signaled a problem.

Use rates with counts and segments

A 0.10% rate may represent one complaint from 1,000 deliveries or 1,000 complaints from one million deliveries. Both deserve attention, but the operational response may differ. Small datasets are noisier; large-volume complaints can damage more recipients and reveal a broader systemic issue.

Always break reporting down by:

  • mailbox provider, especially Gmail, Yahoo, Microsoft, and corporate domains;
  • sending domain and subdomain;
  • campaign, template, subject line, and message category;
  • acquisition source and signup cohort;
  • geography and language where relevant;
  • recipient engagement segment; and
  • send date, volume, cadence, and time zone.

That segmentation turns “our spam rate increased” into a testable diagnosis, such as “complaints rose only among contacts imported from a webinar list who had not engaged for 180 days.”

Common causes of spam complaints and spam-folder placement

There is no single spam trigger word or one universal filter rule. Modern filtering considers identity, infrastructure, message structure, historical behavior, recipient feedback, and context. The most common causes usually fall into a few repeatable categories.

1. No meaningful permission or weak list provenance

Buying, renting, scraping, harvesting, or loosely co-registering addresses is a high-risk practice. Even if some addresses are technically valid, people may not recognize the sender or remember agreeing to receive mail. That creates complaints, invalid addresses, spam traps, and poor engagement.

Permission should be specific. “Download this guide” does not automatically mean “send daily product promotions indefinitely.” Explain what people are signing up for, how often they should expect mail, and what type of content they will receive. Keep records of the signup source, timestamp, form language, and consent status.

2. Sending too frequently or without a predictable cadence

Frequency is not an absolute number. A daily deal email can be appropriate for a subscriber who explicitly asked for daily alerts and inappropriate for someone who expected a monthly product update. The issue is whether frequency matches the expectation set at signup and the recipient’s observed engagement.

Surprise volume spikes are risky for both people and infrastructure. A brand that has sent one newsletter a month for a year and suddenly sends three promotions per day can generate complaints even if the creative is polished. High volume can also expose a weak reputation faster because more recipients are asked to react at once.

3. Irrelevant targeting and poor segmentation

Email is more likely to be reported when it is irrelevant. A customer who bought a product once may not want unrelated promotions; a trial user may need setup help, not repeated upgrade pressure; an inactive subscriber may need a repermission message or suppression, not the same promotional sequence sent to active customers.

Segmentation should change the actual send decision, not merely alter a first name in the greeting. Use product usage, stated preferences, purchase history, location, language, lifecycle stage, recency, and engagement to decide whether a recipient should receive the message at all.

4. Confusing or deceptive identity

A recipient should be able to quickly recognize who sent the message and why. Confusing display names, frequently changing From addresses, reply-to addresses that do not work, misleading domains, and unrelated link domains all make a message look suspicious.

Subject lines matter too. Curiosity can help, but a subject line that promises something the body does not deliver creates a mismatch that drives complaints and violates U.S. rules against deceptive subject lines. The FTC specifically requires header information and subject lines to be accurate rather than false or misleading. (ftc.gov)

5. Broken authentication or domain misalignment

SPF, DKIM, and DMARC do not make unwanted email wanted. They do, however, show mailbox providers that the sender is authorized to use the visible domain and help protect against spoofing. Authentication failures, inconsistent domains, and poor alignment make it harder for providers to associate legitimate traffic with a stable identity.

For bulk senders, Yahoo requires SPF and DKIM, a valid DMARC policy, and alignment between the From domain and either the SPF or DKIM domain. It also requires valid forward and reverse DNS for sending IPs. (senders.yahooinc.com)

6. Difficult or ignored unsubscribes

When people cannot easily unsubscribe, they often use the spam button as the fastest exit. A tiny footer link, a login wall, multiple confirmation screens, unclear preference controls, or an unsubscribe that fails all create frustration.

Yahoo’s sender guidance calls for a functioning List-Unsubscribe header that supports one-click unsubscribe for marketing and subscribed messages, a visible body unsubscribe link, and honoring unsubscribes within two days. (senders.yahooinc.com)

7. Stale, invalid, or disengaged lists

Lists decay. People change jobs, abandon old addresses, create inbox rules, forget subscriptions, or stop caring about a topic. Continuing to send to long-inactive recipients increases the chance of bounces, complaints, and poor engagement.

A valid mailbox is not the same as an interested recipient. Before a large campaign or import, use an email address verification tool to reduce avoidable invalid-address risk, then apply an engagement policy to decide which valid contacts should still receive marketing email.

8. Abrupt changes in sending behavior

New domains, new IPs, new vendors, new subdomains, sudden volume increases, and major template changes can all alter how mail is evaluated. A sender with little reputation should not jump from a few hundred messages to hundreds of thousands overnight.

Start with recipients who are most likely to recognize and value the mail, keep cadence consistent, and expand gradually as positive performance is established. This is commonly called warming, but it is not a trick for bypassing filters. It is a way to avoid exposing an unproven identity to a huge audience before the program has demonstrated healthy behavior.

How mailbox providers decide whether email is spam

Mailbox-provider filtering systems are deliberately not fully transparent. If every rule were public and fixed, abusive senders could optimize around them. Still, the general model is clear: providers combine many signals to estimate whether a message is wanted, safe, and authentic.

Technical identity signals

Providers inspect whether SMTP traffic follows expected standards and whether domain authentication passes. Common signals include:

  • SPF authorization for the envelope sender;
  • DKIM signatures and whether the signed domain is consistent;
  • DMARC alignment with the visible From domain;
  • TLS use where supported;
  • valid forward and reverse DNS for sending IPs;
  • stable HELO/EHLO and sending infrastructure; and
  • domain, IP, and link-domain reputation.

Authentication makes impersonation harder and provides a dependable identity layer for reputation. It does not compensate for sending unwanted mail. A perfectly authenticated email can still be reported as spam, while a valuable transaction email with a broken authentication path can be distrusted or rejected.

Recipient-behavior signals

Mailbox providers can observe actions individual recipients take over time. Signals may include spam reports, moving mail to spam or out of spam, deleting without reading, replying, adding a sender to contacts, and repeated positive engagement. Providers do not disclose a simple formula, and no responsible sender should rely on one behavior metric alone.

The practical takeaway is human: send mail recipients recognize and expect. Relevance, frequency control, and a clear way out are not merely marketing niceties; they reduce the incentive to report spam.

Message and campaign context

Providers also assess the message itself: URL reputation, attachment behavior, HTML structure, deceptive wording, image-heavy or malformed markup, redirect chains, and inconsistencies between the displayed brand and actual domains. Context matters too. A receipt with an order number, customer-service details, and a predictable sender identity is evaluated differently from an unsolicited “urgent” offer containing unfamiliar tracking links.

Avoid simplistic “spam-word” checklists. Words such as “free,” “urgent,” or “sale” are not automatic causes of spam placement. The risk comes from the full pattern: a poor sender reputation, an unrecognized identity, aggressive wording, suspect links, and a low-quality audience can reinforce one another.

How to fix a spam problem: a practical recovery plan

The fastest durable fix is rarely a template redesign. Start with the people receiving the mail, then verify technical identity, then adjust content and volume.

Step 1: Stop making the signal worse

If complaints spike or inbox placement drops sharply, pause the campaign or reduce the affected segment immediately. Do not respond by resending the same email to non-openers, increasing volume to “push through,” or moving to a fresh domain to evade the issue. Those actions often deepen the reputational damage.

Separate marketing traffic from essential transactional traffic where possible. Do not stop a necessary password reset or order receipt solely because a promotional campaign performed poorly, but verify that message types have distinct routing, domains or subdomains when appropriate, and suppression logic.

Step 2: Find the change that preceded the spike

Compare the last healthy send with the first unhealthy one. Review:

  1. List source: Was there an import, partner list, event list, purchased data set, or old segment reactivated?
  2. Audience: Did you expand beyond engaged subscribers or remove suppression rules?
  3. Frequency: Did recipients receive more messages than usual?
  4. Identity: Did the From name, address, reply-to, link domain, or sending domain change?
  5. Technical setup: Did SPF, DKIM, DMARC, DNS, or a sending provider change?
  6. Content: Did the subject line, offer, sender framing, or landing-page behavior become more aggressive or confusing?
  7. Volume: Did the campaign ramp too quickly or hit a new mailbox-provider mix?

This is an investigation, not a guessing game. Use event data and campaign-level comparisons rather than changing ten things at once. If you alter audience, domain, content, cadence, and template simultaneously, you will not know what improved or worsened the result.

Step 3: Suppress risky recipients

Immediately honor unsubscribes, complaints, hard bounces, and provider suppressions. Then define a recency and engagement policy for marketing mail. For example, a program might stop regular promotions to recipients who have not opened, clicked, purchased, logged in, or otherwise engaged for a defined period, depending on the business model and privacy constraints.

A re-engagement campaign can be useful, but it must be restrained. Send one clear message explaining why the recipient is hearing from you, provide a meaningful choice to stay subscribed, and suppress non-responders rather than repeatedly chasing them. Re-engagement is not permission to keep emailing indefinitely.

Step 4: Repair authentication and sending identity

Confirm that every active sending path is authenticated and aligned. This includes application mail, customer-support platforms, billing systems, marketing tools, and any third-party service that uses your domain in the visible From address.

A basic DMARC monitoring record can look like this in DNS:

Host/name: _dmarc.example.com
Type: TXT
Value: v=DMARC1; p=none; rua=mailto:dmarc-reports@example.com

This example is for monitoring, not an instruction to use the exact address or policy in every environment. Review aggregate reports, identify authorized senders, correct SPF/DKIM alignment failures, and only then consider more restrictive DMARC policies. RFC 7489 documents this record structure and provides an example using a rua=mailto: aggregate-report destination. (rfc-editor.org)

For implementation details, consult your sending provider’s email API and setup documentation, especially when different applications, subdomains, or third-party systems send on behalf of the same organization.

Step 5: Make unsubscribing genuinely easy

Every marketing email should have a visible, understandable unsubscribe path. Do not require a password, make recipients answer a survey, or hide the action behind a series of confirmations before stopping mail. Preference centers are useful when they offer real choices, but they must include an uncomplicated option to stop all marketing email.

For one-click unsubscribe, Yahoo provides this header example:

List-Unsubscribe-Post: List-Unsubscribe=One-Click
List-Unsubscribe: <https://example.com/unsub/recipient-token>

The endpoint should process the one-click POST request for that recipient token without requiring additional interaction. RFC 8058 defines the one-click signaling mechanism, and Yahoo explicitly documents the header pair in its sender guidance. Keep a conventional unsubscribe link in the message body as well. (rfc-editor.org)

Step 6: Resume with the most engaged recipients

After fixing the underlying issue, restart gradually with recipients who recently subscribed, opened, clicked, purchased, or otherwise demonstrated a clear relationship with your brand. Keep sends predictable and watch complaints, unsubscribes, bounces, conversion, and mailbox-provider data after each expansion.

Do not infer that a new template “fixed deliverability” after a single small test. Reputation recovery takes time because providers evaluate behavior over time. Consistency is more persuasive than a brief burst of clean metrics.

Preventing spam before a campaign launches

The best spam-reduction strategy is built into program design, not added after a sender reputation problem.

Build a permission-first signup flow

Tell people exactly what they are joining. A strong signup flow includes the sender name, purpose, approximate frequency, and a link to relevant terms or privacy information. Consider confirmed opt-in for high-risk acquisition sources or when a clear proof of consent is particularly valuable.

Store consent data in a way your team can audit later. Useful fields include the address, signup timestamp, IP or source context where appropriate, form or campaign identifier, consent language version, and preference selections. This data is valuable when investigating complaints or deciding whether an old contact should be mailed again.

Set expectations and meet them

The welcome email is an opportunity to restate what recipients will receive and how often. If the program offers weekly tips, do not pivot into daily promotions without a clear preference choice. If recipients registered for a product, distinguish essential service mail from optional marketing rather than blending everything into one noisy stream.

Create a campaign preflight checklist

Before launching a large campaign, ask:

  • Is every recipient eligible and reasonably likely to recognize this sender?
  • Was this audience acquired with appropriate permission for this message type?
  • Does the From name clearly identify the business?
  • Does the subject line accurately describe the content?
  • Does the visible From domain align with authenticated sending?
  • Are links HTTPS, on expected domains, and free of unnecessary redirect chains?
  • Does the email render clearly in common clients and on mobile devices?
  • Is there a visible unsubscribe link and working one-click handling where applicable?
  • Are unsubscribe, complaint, and bounce suppressions active before the send?
  • Is the volume appropriate for the domain’s recent sending history?

A preflight process is especially important when teams move quickly. It turns vague deliverability advice into repeatable release checks that product, marketing, engineering, and compliance teams can share.

Common myths about spam email

“If it is legal, it will reach the inbox.”

No. Compliance and deliverability overlap but are not the same. Meeting legal requirements helps reduce risk and respects recipients, but mailbox providers still filter based on authentication, reputation, engagement, and complaints.

“A delivered email reached the inbox.”

No. Delivered usually means the receiving system accepted the message. It may still land in Spam, Junk, a tabbed category, quarantine, or another folder. Measure inbox placement separately when it matters.

“Spam filters only care about certain words.”

No. Content contributes to a message’s risk profile, but modern filtering is multi-signal. A familiar, authenticated sender with a wanted message can use promotional language successfully; an unrecognized sender with poor list quality can be filtered even with plain text.

“Using a new domain fixes bad reputation.”

Usually not in a durable way. A new domain starts without established trust and can quickly develop the same negative signals if the list, sending cadence, identity, and opt-out experience remain unchanged. Fix the program, not just the sending label.

“Unsubscribes are worse than spam complaints.”

Unsubscribes reduce list size, but they are a healthy and explicit preference signal. Forcing unhappy recipients to remain subscribed is more likely to create spam complaints and harm future inbox placement. Make unsubscribing easy, then focus on retaining people through relevance.

The bottom line on spam email

Spam email is not just fraudulent bulk mail. For responsible senders, it is a warning that recipients or mailbox providers consider a message unwanted, untrustworthy, or poorly matched to its audience. That warning has direct consequences for inbox placement, campaign results, sender reputation, and essential transactional communication.

The durable solution is straightforward in principle: send only to people who reasonably expect the message, make the sender identity clear, authenticate every sending path, maintain clean suppression data, keep volume and frequency predictable, and offer an immediate exit. Monitor spam complaints as an early-warning signal, but investigate them alongside bounces, engagement, authentication, and audience source. When recipients have a clear reason to want your mail, the inbox becomes much easier to earn.

FAQ

What is considered spam email?

Spam email is unwanted, irrelevant, deceptive, abusive, or unsafe email. It can include fraud and phishing, but legitimate marketing email may also be treated as spam when recipients did not expect it, do not recognize the sender, receive it too frequently, or cannot unsubscribe easily.

What is a good spam complaint rate?

Aim well below 0.3%, not merely below it. Google and Yahoo identify 0.3% as an important upper boundary for sender requirements, so use any rise toward that level as a reason to investigate list quality, targeting, frequency, and unsubscribe handling. (support.google.com)

Does a high delivery rate mean my email did not go to spam?

No. Delivery rate generally measures whether the recipient server accepted the message. It does not tell you whether the email landed in the inbox, spam folder, a tab, quarantine, or another destination.

Why are customers marking legitimate emails as spam?

Common reasons include unclear consent, forgotten subscriptions, too much email, irrelevant offers, confusing sender identity, deceptive-looking subject lines, and difficult unsubscribe flows. Review the specific audience and campaign change that occurred before complaints increased.

How do I stop emails from going to spam?

Start with permission and list quality, then verify SPF, DKIM, and DMARC alignment. Send consistently to engaged recipients, avoid abrupt volume spikes, use recognizable sender and link domains, maintain suppression lists, and provide visible plus one-click unsubscribe options for applicable marketing mail.