- Email throttling occurs when a receiving mail server limits the rate at which it accepts messages from your IP or domain, resulting in temporary deferrals (4xx SMTP codes) rather than permanent rejections.
- Throttling differs from bouncing: throttled emails are delayed but can be retried and eventually delivered, while bounced emails are permanently rejected.
- Common triggers include sending too much volume too fast, poor sender reputation, new or unwarmed IPs, sudden volume spikes, and high complaint rates.
- Gmail, Outlook, and Yahoo each throttle differently, but all use engagement signals, volume patterns, and authentication status to determine throttling thresholds.
- Self-throttling (proactively limiting your own sending rate) is the best prevention strategy for avoiding server-side throttling.
You send a large email campaign and everything looks fine at first. Then your MTA logs start filling with 4xx response codes. Messages queue up. Delivery slows to a crawl. Your emails are not being rejected outright; they are being throttled.
Email throttling is one of the most common, yet least understood, deliverability challenges. Unlike a hard bounce that tells you something is permanently wrong, throttling is a temporary speed limit imposed by the receiving server. Understanding why it happens and how to respond is critical for anyone sending at volume.
What Is Email Throttling?
Email throttling is when a receiving mail server intentionally limits the number of messages it will accept from a specific sender within a given timeframe. Instead of accepting or rejecting your email immediately, the server responds with a temporary deferral, telling your sending server to try again later.
These deferrals show up as 4xx SMTP response codes in your MTA logs, most commonly 421 or 451. The response typically includes a message indicating rate limiting, such as "too many connections" or "try again later."
Throttling = temporary slowdown (4xx codes, emails queued for retry). Bouncing = permanent rejection (5xx codes, emails returned undeliverable). Throttled emails can still be delivered if your MTA retries properly. Bounced emails cannot.
Throttling vs. Bouncing vs. Deferral
These three terms describe different failure points, and confusing them leads to wrong fixes:
| Type | SMTP Code | What Happens | Recoverable? |
|---|---|---|---|
| Throttling | 4xx (421, 451, 452) | Server limits your sending rate; queues messages for retry | Yes, with proper retry logic |
| Soft bounce | 4xx | Temporary issue like full mailbox or server unavailable | Yes, usually resolves itself |
| Hard bounce | 5xx (550, 553, 554) | Permanent rejection: invalid address, blocked domain, policy violation | No |
The critical distinction is that throttling is volume-driven. The receiving server is not saying your email is bad; it is saying you are sending too much, too fast for the trust level it has assigned to your IP or domain.
Why Mailbox Providers Throttle Email
Mailbox providers use throttling as a protective mechanism. It serves several purposes:
Infrastructure Protection
Receiving servers handle billions of messages daily. Throttling prevents any single sender from consuming disproportionate resources. Even legitimate high-volume senders can overwhelm a server if they dump hundreds of thousands of messages in minutes.
Reputation Evaluation
When a mailbox provider sees a spike in volume from an IP or domain it does not fully trust, throttling buys time to evaluate the content and engagement signals from the initial batch of accepted messages. If those first messages generate opens, clicks, and no complaints, the provider gradually increases the acceptance rate. If they generate spam complaints, the throttle tightens or turns into outright blocking.
Spam Prevention
Spammers send as fast as possible to maximize delivery before they get blocked. Throttling is an effective countermeasure because legitimate senders will respect retry signals and slow down, while spammers typically abandon the attempt and move to another target.
Common Throttling Triggers
Volume Spikes
Sending 500,000 emails when you normally send 50,000 will trigger throttling at virtually every major mailbox provider. Providers track your sending patterns and flag sudden increases as suspicious. This is especially common during holiday campaigns, product launches, or end-of-quarter pushes when marketers increase volume significantly.
New or Unwarmed IPs
A new dedicated IP has no sending history. Mailbox providers start with very conservative acceptance rates for unknown IPs and gradually increase them as positive engagement data accumulates. Skipping the warmup process and sending at full volume from day one guarantees throttling.
Poor Sender Reputation
If your IP or domain reputation is low due to previous complaints, spam trap hits, or high bounce rates, mailbox providers will throttle your sending more aggressively. A sender with strong reputation gets higher rate limits than one with a questionable history.
High Complaint Rates
Google enforces a strict complaint rate threshold of 0.3%, with a recommended target below 0.1%. When your complaint rate rises, Gmail and other providers tighten throttling limits and may begin rejecting messages entirely.
Authentication Failures
Missing or misconfigured SPF, DKIM, or DMARC records make mailbox providers less willing to accept your mail at high rates. Properly authenticated senders get higher rate limits than unauthenticated ones.
How Major Providers Handle Throttling
Gmail
Gmail is particularly strict with new senders. It starts by accepting a small number of messages and increases limits as engagement signals accumulate. If your initial batch of messages gets good open rates and low complaints, Gmail ramps up quickly. If engagement is poor, limits stay tight or drop further.
Gmail's throttling responses typically include messages like "Our system has detected an unusual rate of unsolicited mail" or "The IP you're using to send mail is not authorized." These indicate varying levels of throttling severity.
Microsoft Outlook/365
Microsoft uses connection-level throttling in addition to message-rate limits. They limit concurrent SMTP connections from a single IP and cap the number of messages per connection. Microsoft recently introduced tenant-level outbound email limits (TERRL) for Exchange Online, calculated based on the number of email licenses a tenant holds.
Important: Microsoft's new Tenant External Recipient Rate Limit (TERRL) caps the number of external recipients a tenant can send to per day. This applies to all outbound mail from the tenant, including messages relayed from on-premises servers. Check the Exchange Admin Center for your tenant's current limit.
Yahoo/AOL
Yahoo limits concurrent IMAP and SMTP connections per IP. Excessive connections (often caused by multiple sending threads or devices) trigger deferrals. Yahoo's error messages are generally clear about the throttling reason, making diagnosis straightforward.
How to Identify Throttling in Your Logs
Throttling shows up in your MTA (Mail Transfer Agent) logs as temporary failures. Here are the key patterns to look for:
421 4.7.28 Our system has detected an unusual rate of unsolicited mail originating from your IP address.
451 4.7.1 Service unavailable - try again later.
452 4.5.3 Too many recipients.
421 Too many concurrent SMTP connections from this IP address.When you see clusters of 4xx codes targeting a specific provider, that is throttling. Your MTA should be configured to retry these messages with increasing intervals (exponential backoff). Most well-configured MTAs handle this automatically, but you should verify that your retry settings are reasonable: start with 5-minute intervals and increase gradually over 24-72 hours.
Monitor your retry queues regularly. If messages to a specific provider are sitting in the retry queue for more than 24 hours, the issue is likely more severe than standard throttling. Check your sender reputation and blacklist status for that IP.
How to Prevent Email Throttling
1. Self-Throttle Your Sending
The most effective strategy is proactive self-throttling: limiting your own outbound sending rate to stay within each provider's acceptable thresholds. If you do not throttle yourself, the receiving server will do it for you, on its terms, not yours.
General guidelines for self-throttling:
- Gmail: Start at 500 messages per hour for new IPs; ramp up gradually based on engagement feedback.
- Microsoft: Limit concurrent connections to 10 or fewer per IP; keep messages per connection under 50.
- Yahoo: Limit to 100 messages per minute for established senders; lower for new IPs.
2. Warm Up New IPs Properly
Every new dedicated IP needs a warmup period of two to six weeks. Start with your most engaged subscribers and gradually increase volume. Consistent sending patterns build trust faster than aggressive ramp-ups.
3. Maintain Consistent Volume
Avoid large swings in sending volume. If your normal daily volume is 50,000 emails, jumping to 300,000 for a promotion will trigger throttling. Instead, spread large campaigns over multiple days or use multiple IPs to distribute the load.
4. Keep Lists Clean
High bounce rates and spam trap hits degrade your reputation, which leads to tighter throttling limits. Regular list hygiene and email verification before sending prevents these reputation hits.
5. Segment by Provider
Configure your MTA to apply different sending rates for different recipient domains. Gmail, Outlook, and Yahoo each have different tolerance levels, and applying a single rate across all providers is inefficient. Send at the maximum rate each provider will accept without throttling.
6. Authenticate Everything
Ensure SPF, DKIM, and DMARC are properly configured and aligned. Use our SPF checker and DMARC checker to verify your records. Authenticated senders consistently receive higher rate limits from mailbox providers.
Recovering from Active Throttling
If you are currently being throttled, take these steps:
- Reduce volume immediately. Cut your sending rate to 25-50% of your current level. Continuing to push against throttling limits makes the problem worse.
- Check your reputation. Use Google Postmaster Tools, Microsoft SNDS, and your ESP's reputation dashboards to assess your standing.
- Review recent campaigns. Identify anything that may have triggered higher complaint rates: new segments, changed content, reactivation campaigns sent to stale addresses.
- Verify authentication. Confirm that SPF, DKIM, and DMARC records are valid and aligned. A configuration change that broke authentication could be the trigger.
- Ramp back up slowly. Once throttling subsides, increase volume gradually over several days, monitoring deferral rates and engagement metrics at each step.
Gmail's throttling is partially driven by machine learning models that analyze engagement patterns. Senders whose recipients consistently open, click, and reply receive higher sending limits. Senders whose mail is ignored or deleted face tighter restrictions, even at the same volume.
How ESPs Handle Throttling for You
Most email service providers build throttling management into their sending infrastructure. They maintain sending rate limits per provider, implement automatic retry with backoff, and distribute load across multiple IPs. If you use a reputable ESP, much of the throttling prevention is handled behind the scenes.
However, your ESP cannot protect you from throttling caused by poor list quality, high complaint rates, or content issues. Those factors are your responsibility. Your ESP controls the sending infrastructure; you control the list, the content, and the sending strategy.
Frequently Asked Questions
Check your MTA logs for clusters of 4xx SMTP response codes, particularly 421 and 451 errors targeting a specific provider. A growing retry queue for one domain (e.g., gmail.com) while other domains deliver normally is a strong indicator of throttling.
Standard throttling from a volume spike typically resolves within a few hours as your MTA retries and the receiving server accepts messages at its preferred pace. Throttling caused by reputation issues can last days or weeks until the underlying reputation improves. Reducing volume and improving engagement accelerates recovery.
Throttling itself does not directly damage your reputation. It is a symptom, not a cause. However, if you respond to throttling by hammering the receiving server with aggressive retries, that behavior can harm your reputation further. Always respect deferral signals and back off accordingly.
Distributing volume across multiple IPs can increase your aggregate sending rate since throttling limits are often applied per IP. However, each IP must have its own established reputation. Spreading traffic across unwarmed IPs will trigger throttling on all of them. Also, mailbox providers increasingly track sending patterns at the domain level, not just IP level, so adding IPs does not bypass domain-based throttling.
Sender-side throttling is when you or your ESP intentionally limit your outbound sending rate to stay within acceptable thresholds. This is proactive and prevents problems. Receiver-side throttling is when the receiving mail server limits how many of your messages it will accept. This is reactive and indicates you have already exceeded the server's tolerance for your volume or reputation level.