Email deliverability is the ability of your emails to actually land in the recipient’s inbox instead of the spam folder, the promotions tab, or a mailbox provider’s quarantine. It is not the same thing as your delivery rate. An email can be “delivered” (accepted by the receiving server) and still be filtered straight into junk. Deliverability is about inbox placement, not just acceptance.
Most articles on this topic hand you a list of twenty generic tips and wish you luck. This one is built differently. Inbox placement problems almost always trace back to six specific factors, and the symptoms you are seeing (a sudden drop, one provider blocking you, gradual decay over months) point to different culprits. Below you will find the definitions, the benchmarks, and a diagnostic framework to identify which of the six is actually hurting you.
Deliverability vs. Delivery Rate: The Distinction That Trips Everyone Up
Your email platform probably shows you a delivery rate of 98% or 99% and you assume everything is fine. That number only tells you the receiving server did not bounce your message. It says nothing about where the message ended up. (via https://mailchimp.com)
| Metric | What it measures | What it hides |
|---|---|---|
| Delivery rate | Emails accepted by the receiving server (sent minus bounces) | Everything routed to spam or quarantined after acceptance |
| Deliverability / inbox placement | Emails that reach the primary inbox where a human can see them | Nothing important, but it is harder to measure and requires seed testing or Postmaster data |
| Open rate | Proxy for engagement | Distorted by privacy protection features that auto-open images |
The practical takeaway: a 99% delivery rate paired with a collapsing open rate is the classic signature of a deliverability problem. Your mail is being accepted and then filtered.
What Is a Good Email Deliverability Rate?
Benchmarks vary by industry and list quality, but here is how to read your own numbers:
- 95% or higher inbox placement: healthy. Normal maintenance is enough.
- 85% to 94%: a warning zone. Something in the six factors below is degrading. Diagnose now, before it compounds.
- Below 85%: active reputation damage. Roughly one in six paying-attention subscribers never sees you.
- Bounce rate: keep it under 2%. Above 5% and mailbox providers start throttling you.
- Spam complaint rate: stay under 0.10%. Gmail’s published hard ceiling for bulk senders is 0.30%, and hitting it triggers filtering.
The 6 Factors That Decide Inbox Placement
Mailbox providers such as Gmail, Outlook, Yahoo and Apple Mail run every incoming message through a scoring engine. That engine weighs the following six inputs. Fix the right one and placement recovers. Fix the wrong one and you waste a month.
Factor 1: Sender Reputation
Sender reputation is a score that mailbox providers assign to your sending domain and your sending IP address. It is the single heaviest weighted input, and it is built from your history: complaint rates, bounce rates, spam trap hits, how often recipients delete without reading, and how often they move you out of junk.
Domain reputation vs. IP reputation
- Domain reputation follows you everywhere. Change email platforms and it comes with you. This is now the dominant signal at Gmail.
- IP reputation matters most if you send from a dedicated IP. On a shared IP, you inherit part of the reputation of everyone else on that IP, for better or worse.
How to check it
- Open Google Postmaster Tools and look at the Domain Reputation and IP Reputation graphs. Anything below “High” deserves attention. “Bad” means most of your Gmail traffic is going to spam.
- Enroll in Microsoft SNDS for visibility into Outlook and Hotmail traffic.
- Check whether your domain or IP appears on major blocklists (Spamhaus, SpamCop, Barracuda).
Symptom that points here: a broad, cross-provider decline that affects every campaign regardless of content.
Factor 2: Authentication (SPF, DKIM and DMARC)
Authentication proves you are who you claim to be. Since major providers tightened their bulk sender rules, missing or misconfigured authentication is no longer a soft penalty. It is a hard barrier.
| Protocol | What it proves | Where it lives | Common failure |
|---|---|---|---|
| SPF | That the sending server is authorized to send for your domain | DNS TXT record | More than 10 DNS lookups, which breaks the record entirely |
| DKIM | That the message was not altered in transit, via cryptographic signature | DNS TXT record plus signing at the sending platform | Key never published, or a weak 1024-bit key |
| DMARC | What to do when SPF or DKIM fails, and sends you reports | DNS TXT record at _dmarc.yourdomain.com | Left at p=none forever, or misaligned From domain |
The bulk sender checklist you must pass
If you send more than 5,000 messages a day to a single major provider, you are treated as a bulk sender and the following are mandatory:
- SPF and DKIM both configured and passing
- A published DMARC policy (p=none is the minimum entry point, but move toward quarantine or reject)
- Alignment: your visible From domain must match the domain authenticated by SPF or DKIM
- One-click unsubscribe via List-Unsubscribe and List-Unsubscribe-Post headers, honored within two days
- Spam complaint rate kept under 0.30%
- Valid forward and reverse DNS (PTR) records for your sending IP
- TLS encryption for transmission
Symptom that points here: emails rejected outright, a sudden total block at one provider, or authentication failures visible in the email headers. This is the fastest factor to fix and should always be checked first.
Factor 3: List Hygiene
Your list is a liability as much as an asset. Every invalid address, abandoned mailbox and recycled spam trap on it tells mailbox providers you do not manage your data.
What damages a list
- Purchased or scraped lists. There is no safe way to use these. They are loaded with spam traps and will destroy domain reputation quickly.
- Recycled spam traps. Real addresses abandoned for a long period, then reactivated by providers purely to catch senders who never clean their data.
- Pristine spam traps. Addresses that were never human. Hitting one means you scraped or bought data.
- Typo domains. gmial.com, yahooo.com, hotmial.com. Some of these are owned by trap operators.
- Role addresses. info@, sales@, contact@, admin@. High complaint risk, low engagement.
- Old imports. A list with more than two years of accumulated addresses and no re-permission process.
The hygiene routine
- Use double opt-in for every new signup. It costs you a little volume and saves you the reputation.
- Remove hard bounces immediately, automatically, every time.
- Suppress addresses that soft bounce three times consecutively.
- Run a validation pass on any list segment that has not been mailed in six months or more.
- Sunset unengaged subscribers on a fixed schedule (see Factor 4).
Symptom that points here: bounce rate above 3%, a big drop right after importing an old list, or reputation damage that appeared immediately after a one-off campaign to a dormant segment.
Factor 4: Engagement Signals
Filtering has shifted heavily toward behavior. Mailbox providers watch what recipients do with your mail and adjust placement accordingly, per-recipient and per-domain.
Positive signals
- Opening, and especially reading for more than a couple of seconds
- Clicking a link
- Replying
- Moving your message out of spam, or marking as “not spam”
- Adding you to contacts, starring, or filing into a folder
Negative signals
- Marking as spam (the most damaging single action)
- Deleting without opening, repeatedly
- Never opening across a long stretch of sends
- Unsubscribing at an elevated rate
Build a sunset policy
This is the highest-leverage fix most senders skip. Segment by recency of engagement:
| Segment | Definition | Action |
|---|---|---|
| Active | Opened or clicked in the last 90 days | Full sending frequency |
| Cooling | Last engagement 90 to 180 days ago | Reduce frequency, send your best-performing content only |
| Dormant | No engagement in 180 to 365 days | Run a short re-permission sequence, two or three emails maximum |
| Sunset | No engagement after re-permission attempt | Suppress. Keep the record, stop mailing it. |
Symptom that points here: slow, steady decline in open rates over several months with no bounce or authentication problems. Your list is aging faster than you are cleaning it.
Factor 5: Content and Technical Triggers
Content matters less than reputation, but it is the tie-breaker when a filter is undecided. It is also the factor people obsess over first, usually wrongly.
What actually raises flags
- Link domain reputation. If you link to a shortener, a shared tracking domain, or any site with a poor reputation, that reputation transfers to your message. Use a branded tracking domain aligned with your sending domain.
- Image-only emails. A single large image with almost no text is a classic spam pattern. Keep a real text-to-image balance.
- Missing plain text version. Always send multipart messages with both HTML and plain text.
- Broken or bloated HTML. Unclosed tags, code pasted from a word processor, messages over 100 KB that get clipped in Gmail.
- Attachments. Avoid them in bulk mail. Link to a hosted file instead.
- Deceptive subject lines. Mismatch between subject and body is a direct complaint driver.
- Missing physical address or unsubscribe link. Legally required in most jurisdictions and checked by filters.
The “spam words” myth
Modern filters are not keyword scanners. Writing “free” will not send you to junk on its own. What gets you filtered is a low-reputation sender writing like a low-reputation sender: all caps, excessive exclamation points, red bold text, urgency stacking, and hidden text. Clean up the pattern, not the vocabulary.
Symptom that points here: one specific campaign or template underperforms while your others are fine.
Factor 6: Sending Volume, Consistency and Infrastructure
Mailbox providers profile your normal behavior. Deviating from it looks like a compromised account.
The rules of volume
- Warm up new domains and IPs. A brand new sending domain has no reputation, which is not the same as a good one. Start with a few hundred of your most engaged contacts per day and increase gradually over four to six weeks.
- Avoid volume spikes. Going from 2,000 emails a month to 80,000 in a single send is the fastest way to get throttled. Ramp in steps of roughly 30 to 50% per send.
- Send consistently. One email every quarter is worse than one every two weeks. Irregular senders have no stable reputation to draw on.
- Separate your streams. Transactional mail (receipts, password resets, order confirmations) should go on a different subdomain from marketing mail. One bad promo campaign should never jeopardize your password reset emails.
- Use subdomains deliberately. For example, mail.yourdomain.com for marketing and notify.yourdomain.com for transactional, keeping your root domain protected.
Symptom that points here: deferrals and throttling messages in your logs (4xx responses), or a placement collapse right after a large one-off send.
Diagnose It: Match Your Symptom to the Factor
Use this table before you change anything. Guessing wastes weeks, because reputation recovery is slow.
| What you are seeing | Most likely factor | First thing to check |
|---|---|---|
| Emails rejected or bounced with a policy message | Authentication or blocklisting | Read the bounce code, then verify SPF, DKIM, DMARC alignment |
| Fine at Outlook, spam at Gmail (or vice versa) | Provider-specific reputation | Google Postmaster Tools or Microsoft SNDS for that provider |
| Gradual open rate decline over 3 to 6 months | Engagement signals | Percentage of your list that has engaged in the last 90 days |
| Sudden crash after one campaign | List hygiene or volume spike | What segment was mailed, and how it compared to your usual volume |
| Bounce rate above 3% | List hygiene | Source of recent signups, validation process, opt-in method |
| One template underperforms, others are normal | Content triggers | Image-to-text ratio, link domains, HTML weight |
| Landing in Promotions instead of Primary | Content and engagement | This is not spam filtering. Focus on reply rate and personalization. |
| Brand new domain, nothing lands | Volume and warm-up | Your ramp schedule and initial recipient quality |
| Deferrals and 4xx retry messages in logs | Throttling from reputation or volume | Send rate per hour and domain reputation trend |
How to Check Your Email Deliverability (Free and Paid)
- Google Postmaster Tools. Free. Requires domain verification and a meaningful daily volume to Gmail. Gives you domain reputation, IP reputation, spam rate and authentication pass rates. This is the single most useful data source available.
- Microsoft SNDS and JMRP. Free. Covers Outlook, Hotmail and Live traffic, and lets you receive complaint feedback.
- DMARC aggregate reports. Point the rua tag in your DMARC record at a reporting tool and you get daily XML reports on who is sending as your domain and whether they pass.
- Seed list / inbox placement tests. Send to a set of test addresses across providers and see where each one lands. Useful for a snapshot, but a small sample.
- Blocklist monitoring. Check your domain and IP against Spamhaus and similar lists monthly, not only when something breaks.
- Your own metrics. Track inbox placement proxies: open rate by mailbox provider. If Gmail opens are half your Outlook opens, you have a Gmail-specific problem.
A 30-Day Recovery Plan
Week 1: Fix the technical foundation
- Publish or repair SPF, DKIM and DMARC, and confirm alignment
- Add List-Unsubscribe with one-click support
- Set up a branded tracking domain
- Enroll in Postmaster Tools and SNDS
- Check every major blocklist and file delisting requests where needed
Week 2: Clean the list
- Remove all hard bounces and repeat soft bounces
- Purge role addresses and obvious typo domains
- Run validation on any segment untouched for six months or more
- Move to double opt-in on all forms
Week 3: Rebuild engagement
- Mail only your 90-day engaged segment. Nothing else.
- Reduce frequency temporarily and raise content quality
- Ask for replies in at least one send, since replies are a strong positive signal
Week 4: Ramp back carefully
- Reintroduce the cooling segment in small batches
- Increase volume by no more than 30 to 50% per send
- Watch domain reputation daily and stop expanding if it dips
- Run a re-permission sequence for dormant contacts, then suppress the non-responders permanently
Set expectations: authentication fixes can show results in days. Reputation repair typically takes four to eight weeks of consistent good behavior. There is no shortcut, and any vendor promising one is selling you a problem.
Five Deliverability Myths Worth Discarding
- “A 99% delivery rate means everything is fine.” It means nothing bounced. It says nothing about placement.
- “Avoiding spam trigger words fixes it.” Content is a tie-breaker, not the main driver. Reputation is.
- “A dedicated IP is always better.” Below roughly 100,000 emails per month, a dedicated IP often hurts because you cannot generate enough consistent volume to maintain a reputation on it.
- “Bigger lists mean more revenue.” Mailing dead addresses reduces inbox placement for the contacts who actually want to hear from you. A smaller engaged list outperforms a bloated one.
- “Landing in Promotions is a deliverability failure.” It is a tab, not a filter. Users who shop expect to find offers there.
FAQ
What is a good email deliverability rate?
Aim for 95% or higher inbox placement, a bounce rate under 2%, and a spam complaint rate under 0.10%. Below 85% inbox placement, you have an active reputation problem that will keep compounding until you address the underlying factor.
How do I fix email deliverability?
Diagnose first. Verify SPF, DKIM and DMARC are passing and aligned. Then clean your list of bounces and dormant contacts. Then mail only your engaged segment for two to three weeks to rebuild positive signals, and ramp volume back gradually. Fixing content before fixing reputation almost never works.
How do I check my email deliverability?
Use Google Postmaster Tools for Gmail data, Microsoft SNDS for Outlook, DMARC aggregate reports for authentication visibility, and a seed list test for a cross-provider snapshot. Also compare your open rates broken down by mailbox provider, which reveals provider-specific filtering quickly.
Does email deliverability differ for cold outreach?
Yes, and it is far riskier. Cold recipients did not opt in, so complaint rates run higher and engagement runs lower. If you do cold outreach, use a separate domain from your main brand domain, keep daily volume low per mailbox, personalize heavily, and monitor complaints obsessively. Never run cold campaigns from the domain that sends your transactional email.
Why do my emails go to spam only at Gmail?
Each provider maintains its own reputation model. A Gmail-only problem usually means your Gmail-specific complaint rate, engagement, or authentication pass rate is below threshold. Google Postmaster Tools will tell you which of the three within a day of checking. attentive.com makes the same point with more data.
How long does it take to recover deliverability?
Authentication and blocklist issues can resolve in a few days. Reputation damage from complaints or bad list practices typically takes four to eight weeks of clean, consistent sending. Severe damage on an established domain can take longer, which is why prevention is dramatically cheaper than recovery.
Should transactional and marketing emails use the same domain?
They should share the same root domain for brand trust, but be sent from separate subdomains with separate reputations. That way a poorly received marketing campaign never puts your password reset and order confirmation emails at risk.
The Bottom Line
Email deliverability is not a mystery and it is not luck. It is six measurable factors: sender reputation, authentication, list hygiene, engagement signals, content triggers and sending volume. When your emails stop landing, one or two of those are almost always responsible, and the symptom tells you which. Diagnose before you act, fix the technical foundation first, and give reputation the weeks it needs to rebuild.
If you would like a hand auditing your setup or interpreting your Postmaster data, our team at Cadecran can walk through your authentication records, list health and sending patterns with you and hand back a prioritized action list. Get in touch and let us take a look.