A marketing team once celebrated hitting a 98% delivery rate on their latest campaign. The emails were getting through! But when they checked actual results, open rates had cratered to 3%. Something wasn't adding up.
The problem was a fundamental misunderstanding of what "delivered" means. Their emails were being accepted by Gmail and Microsoft—hence the high delivery rate. But they were being routed straight to spam folders, where almost no one ever looked. The emails were technically delivered but practically invisible.
This distinction between delivery rate and inbox placement is one of the most important concepts in email deliverability, and one of the most commonly confused.
What delivery rate actually measures
Delivery rate is the percentage of emails that were accepted by the receiving mail server. It's calculated simply: emails accepted divided by emails sent, minus hard bounces.
When you send an email, the receiving server makes a binary decision: accept or reject. If it accepts, the email counts as delivered. If it rejects (returning a 5xx error), it's a bounce. Delivery rate measures this acceptance rate.
A 98% delivery rate means 98% of your emails were accepted by receiving servers. The other 2% bounced—maybe invalid addresses, maybe blocked by the server, maybe rejected for policy reasons.
This metric is easy to measure because it's based on clear server responses. Your email system knows immediately whether each message was accepted or rejected. No ambiguity, no delay.
But here's the critical limitation: delivery rate tells you nothing about what happens after acceptance. The receiving server might accept your email and then immediately route it to spam. It might accept it and delete it. It might accept it and deliver it to the inbox. Delivery rate doesn't distinguish between these outcomes.
What inbox placement measures
Inbox placement rate measures the percentage of delivered emails that actually reach the recipient's inbox, as opposed to the spam folder or other filtered locations.
This is what actually matters for email effectiveness. An email in the inbox might get opened. An email in spam almost certainly won't. If your goal is engagement—opens, clicks, conversions—inbox placement is the metric that predicts success.
Inbox placement is much harder to measure than delivery rate. You can't directly observe where emails land; that information stays with the receiving server. Instead, you have to infer it through indirect methods.
Seed testing is the most common approach. You maintain a panel of test email addresses across various providers. When you send a campaign, you include these seed addresses. Then you check where the emails landed—inbox or spam. The results from your seed panel estimate your overall inbox placement.
Engagement metrics provide indirect signals. If your open rates suddenly drop while delivery rates stay high, emails are probably landing in spam. This isn't precise measurement, but it's a useful indicator.
Some email providers offer inbox placement data through their analytics. They aggregate data across their customer base to estimate placement rates by receiving domain. This can be more accurate than small seed panels but still involves estimation.
Why the gap matters
The gap between delivery rate and inbox placement can be enormous, and that gap represents lost opportunity.
Consider a scenario: you send 100,000 emails. Your delivery rate is 97%—97,000 emails accepted. But your inbox placement is only 60%. That means roughly 58,000 emails reached the inbox, while 39,000 went to spam. Your delivery rate looks great; your actual reach is mediocre.
This gap tends to widen when you have reputation problems. ISPs increasingly accept emails rather than rejecting them outright, then filter aggressively on the backend. This approach gives them more data for their spam algorithms and avoids tipping off spammers about what triggers rejection.
Gmail is particularly known for this pattern. They'll accept almost everything, then use their sophisticated filtering to sort it. You might have 99% delivery rate to Gmail while having 40% inbox placement. The delivery rate gives false confidence.
Factors that affect each metric
Delivery rate and inbox placement respond to different factors, though there's overlap.
Delivery rate is primarily affected by list quality and authentication. Invalid addresses cause bounces, lowering delivery rate. Failed authentication (SPF, DKIM, DMARC) can trigger rejection. Blacklist listings cause rejection. Severe reputation problems cause rejection.
Inbox placement is affected by everything that affects delivery rate, plus much more. Engagement history matters—if recipients don't open your emails, future emails are more likely to be filtered. Content matters—certain words, patterns, or formatting trigger spam filters. Sender reputation matters—even if you're not rejected, poor reputation routes you to spam. Recipient behavior matters—if people mark your email as spam, similar future emails get filtered.
You can have perfect delivery rate with terrible inbox placement. Your authentication is correct, your list is clean, you're not blacklisted—so emails are accepted. But your content triggers filters, your reputation is mediocre, and recipients don't engage—so emails go to spam.
Measuring inbox placement
Since inbox placement can't be directly observed, you need strategies to estimate it.
Seed list testing remains the gold standard. Services like Validity (Everest), GlockApps, and others maintain panels of real email addresses across major providers. You send to these seeds along with your regular campaigns, then check where they landed. The seed results estimate your overall placement.
The limitation is sample size. A panel of a few hundred addresses can't perfectly represent billions of inboxes. Results can vary between tests. But over time, patterns emerge that are useful for identifying problems and tracking improvement.
Engagement-based inference works for ongoing monitoring. Establish baseline open rates for your list. If open rates drop significantly without other explanation (like subject line changes), suspect inbox placement problems. This won't tell you exact placement rates, but it flags issues quickly.
Provider-specific tools help with major destinations. Google Postmaster Tools shows spam rates for Gmail. If your spam rate there is climbing, your inbox placement is falling. Microsoft offers similar data through SNDS.
A/B testing can isolate placement issues. If you send identical content from two different sending configurations and see dramatically different open rates, the difference is likely inbox placement, not content appeal.
Improving inbox placement
Since inbox placement depends on more factors than delivery rate, improving it requires broader effort.
Engagement is increasingly central. ISPs watch whether recipients open, click, reply to, or move your emails out of spam. High engagement signals wanted email; low engagement signals the opposite. Send to engaged subscribers, re-engage or remove inactive ones, and create content people actually want to read.
Reputation management matters enormously. Monitor your sender reputation through available tools. Address problems quickly. Maintain consistent sending patterns. Avoid practices that damage reputation—purchased lists, spam trap hits, high complaint rates.
Content optimization helps at the margins. Avoid spam trigger words and patterns. Maintain good text-to-image ratios. Include clear unsubscribe options. Make sure your emails render well and provide value.
Authentication should be complete. SPF, DKIM, and DMARC all passing is table stakes. Failures don't always cause rejection, but they do influence filtering decisions.
List hygiene prevents the problems that hurt both metrics. Clean lists have fewer bounces (better delivery rate) and fewer spam traps and unengaged recipients (better inbox placement).
Which metric to prioritize
Both metrics matter, but inbox placement is ultimately more important for business outcomes.
Delivery rate is a prerequisite. If emails aren't being accepted, they certainly aren't reaching inboxes. Delivery rate problems need immediate attention—they usually indicate acute issues like blacklisting or authentication failures.
But once delivery rate is healthy (above 95%), inbox placement becomes the limiting factor. The difference between 95% and 99% delivery rate is minor compared to the difference between 60% and 90% inbox placement.
Focus your ongoing optimization efforts on inbox placement. Monitor delivery rate for sudden drops that indicate problems, but measure success by whether emails actually reach inboxes and drive engagement.
Frequently asked questions
What's a good inbox placement rate?
Above 80% is generally healthy. Above 90% is excellent. Below 70% indicates significant problems. These benchmarks vary by industry and sending type—transactional emails typically see higher placement than marketing emails.
Why do ISPs accept emails just to spam them?
Accepting emails gives ISPs more data for their filtering algorithms. It also avoids revealing to spammers exactly what triggers rejection. From the ISP's perspective, filtering is more flexible than rejection.
Can I see inbox placement in my email platform's analytics?
Most platforms show delivery rate but not true inbox placement. Some offer estimated placement based on aggregated data or seed testing. For accurate measurement, you typically need dedicated deliverability tools.
How often should I test inbox placement?
For regular senders, monthly seed testing provides good baseline monitoring. Test more frequently when making changes to sending infrastructure, content templates, or list management practices. Test immediately if you notice engagement drops.