Mailgun Email Deliverability Guide

Google says bulk senders should keep reported spam rates below 0.3%. That single number explains a lot about Mailgun email deliverability, because a message that Mailgun accepts can still land in spam or get throttled if mailbox providers see weak signals.

If you send through Mailgun, the real job is not just SMTP handoff. It is aligning your domain, protecting your reputation, separating traffic types, and keeping engagement strong enough that Gmail, Outlook, Yahoo, and Apple Mail interpret the mail as wanted.

When Mailgun email deliverability drops, the fix is usually operational, not magical. The teams that recover fastest are the ones that can connect Mailgun logs to list quality, campaign cadence, and mailbox-provider behavior.

What Mailgun controls, and what it does not

Mailgun gives you the transport layer, domain authentication support, IP options, event data, and routing controls needed to run a serious program. That matters, because reliable queueing and clean DNS setup reduce avoidable failures before a message ever reaches Gmail or Outlook.

But Mailgun does not decide whether a recipient wanted the message. Mailbox providers do. If your Mailgun account shows a 99% accepted rate and inbox placement still slides, the missing context is usually reputation, targeting, or alignment rather than a platform outage.

That is why deliverability work on Mailgun should be framed as a decision system. Use Mailgun for sending and telemetry, then diagnose outcomes by stream, domain, IP, and mailbox provider.

Get the Mailgun setup right before you scale

Authenticate every sending domain

For Mailgun email deliverability, SPF and DKIM are table stakes, and DMARC is the policy layer that tells mailbox providers how seriously you manage identity. Your visible From domain should align with either SPF or DKIM, and in practice DKIM alignment is usually the cleaner path for marketing traffic.

Use separate subdomains for separate jobs. A common setup is tx.example.com for transactional mail and news.example.com for marketing. That separation protects receipts, password resets, and product alerts from problems caused by promotional campaigns with weaker engagement.

Choose shared or dedicated IPs based on sending pattern

Shared IPs can work well when your volume is moderate and stable, because reputation is pooled and operational overhead is lower. Dedicated IPs make more sense when you send enough volume to establish your own reputation and you can control consistency week after week.

If you send 5,000 one day, 80,000 the next, and then go quiet for a week, a dedicated IP usually creates more work than value. If you send 100,000 to 500,000 messages a day on a predictable pattern, it gives you cleaner control.

Warm volume changes, not just new IPs

Teams often think about IP warm-up only when a new dedicated IP is provisioned. In practice, mailbox providers also react to abrupt volume shifts on an existing IP or domain. A jump from 25,000 messages a day to 150,000 in one afternoon is enough to trigger Gmail temp failures or Outlook throttling, even if the infrastructure itself is healthy.

Mailgun can transmit the mail, but mailbox providers still judge whether the increase looks earned. Ramp up by audience quality first, then broaden volume once engagement holds.

The metrics that actually predict Mailgun email deliverability

Complaint rate is the fastest reputation signal

For Gmail in particular, complaint rate is one of the clearest forward indicators. Google says bulk senders should stay below 0.3%, and many teams try to operate closer to 0.1% to keep more margin. If you are crossing those levels on a Mailgun marketing stream, the problem is usually audience selection, frequency, or expectation setting.

Measure complaints separately for onboarding, promotional, win-back, and product-update mail. A blended account-level average hides the stream that is doing the damage.

Bounces and deferrals tell you whether the issue is data quality or trust

Hard bounces usually point to list hygiene, acquisition quality, or stale CRM data. If hard bounces are consistently above 2%, fix the source of the addresses before you optimize creative or cadence.

Soft bounces and temporary failures are different. Outlook often signals reputation stress through 421 or 451 style deferrals before it blocks outright. Gmail may slow acceptance when volume jumps too quickly. Yahoo can also temp-fail when it sees unfamiliar patterns. In Mailgun, those patterns should be reviewed by domain, IP, and campaign type, not as one blended bounce bucket.

Do not let Apple Mail open data mislead you

Apple Mail Privacy Protection inflated open rates for a large share of recipients, which means opens alone are not a reliable engagement proxy. For Mailgun email deliverability decisions, weight clicks, replies, conversions, recent product activity, and recency of last meaningful action much more heavily.

If a segment has not clicked, replied, or converted in 180 days, sending the same promotional cadence through Mailgun usually adds more reputation risk than revenue. Re-permission or suppress it before it drags down the rest of the stream.

Common Mailgun deliverability problems, and the likely fix

Mailgun accepts the message, but Gmail placement is weak

Start with alignment and complaint rate. Check that the From domain aligns with DKIM or SPF, then review whether the audience expanded recently to colder segments. Gmail often tolerates technically valid mail that it still considers low value, and that is where accepted events can be misleading.

The fix is usually to tighten segmentation, reduce frequency to low-engagement cohorts, and prove positive response again before scaling back up.

Outlook defers mail while Gmail looks normal

This usually points to reputation or cadence rather than authentication. Outlook tends to react strongly to sudden bursts, inconsistent sending windows, and questionable address quality. If Mailgun logs show repeated temporary failures at Outlook but normal acceptance elsewhere, reduce concurrency, smooth the schedule, and inspect recent list-source changes.

It also helps to isolate Outlook-heavy audiences by stream so you can lower pressure without slowing the whole program.

Transactional and marketing traffic are affecting each other

If password resets and invoices share the same domain and reputation path as broad promotional sends, marketing mistakes can bleed into critical mail. Mailgun makes it possible to separate streams by domain and, where appropriate, by IP pool. That is usually one of the highest-impact changes a lifecycle team can make.

The reason is simple, mailbox providers build trust on patterns. Transactional mail is expected, high intent, and often quickly acted on. Promotional mail has a different engagement curve and should not borrow against the same trust bucket.

An operating model that keeps Mailgun deliverability stable

The strongest Mailgun programs treat deliverability as an operating rhythm, not a rescue project. Review performance by mailbox provider each week. Track complaint rate, hard bounce rate, temporary failures, unsubscribes, and engagement by stream. Compare changes in deliverability to changes in audience rules, frequency, and volume, not just subject lines.

For RevOps and lifecycle teams, the practical move is to connect Mailgun event data to your warehouse or BI layer so you can answer specific questions fast. Did Outlook deferrals rise after a new lead source was added. Did Gmail complaints spike after a win-back expansion. Did transactional mail improve once it moved to its own subdomain. Those are the decisions that actually improve inbox placement.

When you test changes, change one major variable at a time. If you alter frequency, audience size, and sending domain together, you will not know which lever improved or hurt Mailgun email deliverability.

Related reading: email deliverability tools and spf and dkim deliverability.

Run your first deliverability test

FAQs

Is Mailgun good for email deliverability?

Mailgun is capable infrastructure, but deliverability depends on how you authenticate domains, manage reputation, separate traffic, and control list quality. The platform can enable good outcomes, but it does not guarantee inbox placement.

Should I use a dedicated IP with Mailgun?

Use a dedicated IP when you have enough stable volume to build and maintain your own reputation. If your volume is low or inconsistent, a shared IP is often the safer choice.

Why do Mailgun accepted events not match inbox placement?

An accepted event means the message entered the downstream delivery path. Gmail, Outlook, Yahoo, and Apple Mail still decide whether to inbox, spam-folder, defer, or silently filter based on trust and recipient signals.

What is the first thing to check when Mailgun email deliverability drops?

Check complaint rate and recent audience changes first, then verify domain alignment and look for ISP-specific deferrals. Those usually explain more than send acceptance alone.

Should transactional and marketing email use different Mailgun domains?

Yes, in most cases. Separate domains or subdomains help protect critical transactional traffic from lower-engagement marketing patterns and make reputation analysis much clearer.