Hi,
This week I received an interesting question: messages were reaching the inbox in seed mailboxes, but tests in personal accounts and other signals suggested that some emails might be going to spam.
How is that possible?
When you use a tool like InboxAlly, the seeds show where your messages were detected in those monitored mailboxes. They are useful for generating positive engagement signals, improving your sender reputation and email deliverability, tracking trends, and comparing results. But they do not reproduce the exact history, preferences, and interactions of every recipient on your list.
That is why good seed results are encouraging, but they do not end the investigation when there are signs that your readers are having a different experience.
Seed mailboxes and real recipients have different histories
Mailbox providers may consider several signals when deciding what happens to a message. These include sender reputation, message characteristics, and each recipient’s relationship with the emails they receive.
Someone who recognizes the sender and interacts frequently may have a different experience from someone who has never opened the messages or has previously marked them as spam.
Seed mailboxes also have their own history. That makes it possible to observe one result in the seeds and another in a group of personal mailboxes.
This does not necessarily mean that one result is wrong. It means they are showing different samples.
Does a drop in open rates mean you have a deliverability problem?
Not necessarily.
Although open rates are becoming less precise, a decline deserves attention, especially when it occurs alongside other signals. But it does not identify the cause on its own. Changes in the subject line, audience, frequency, and the way opens are recorded can also affect the metric.
Google itself says it does not track open rates and that a low rate is not, by itself, a reliable indicator of delivery or spam classification problems.
To investigate, compare similar campaigns and ask:
- Did the decline happen across all providers, or was it concentrated at Gmail?
- Did the composition of the list change?
- Did the content or sending frequency change?
- Did clicks, replies, and unsubscribes follow the same trend?
- Were there spam complaints, deferrals, or rejections during the same period?
Look at spam complaints
If recipients are marking your messages as spam, that signal deserves attention even when the seeds show good results.
For personal Gmail accounts, Google Postmaster Tools helps you monitor the user-reported spam rate. Google recommends keeping it below 0.10% and avoiding 0.30% or higher.
But there is an important detail: a spam rate of zero does not prove, by itself, that the problem has been resolved. In fact, if your spam rate is consistently zero, as counterintuitive as it may seem, you may have a deliverability problem.
Google explains that if many messages are sent directly to the spam folder, fewer recipients can mark them as spam from their inbox. The spam rate reflects user reports of messages delivered to the inbox. On low-volume days, some data may also be unavailable to protect user privacy.
That is why you should observe the trend over time and compare the rate with the other available signals.
How to investigate in practice
When the results appear to disagree, I would start with these steps:
- Confirm the message and the send. Is the campaign sent to the seeds the same one the recipients received? Was it sent from the same domain and infrastructure? Did it use the same content, links, and configuration?
- Analyze each provider. Separate Gmail, Yahoo, Microsoft, and other providers relevant to your audience. An overall average can hide a problem concentrated at just one of them.
- Compare similar periods and audiences. Check whether the size or composition of your list changed, and allow an equivalent amount of time when comparing campaign performance.
- Run controlled tests. Sending the message to a few personal mailboxes you can access can help you develop hypotheses. Because those mailboxes have their own histories and make up a small sample, their results do not represent your entire audience either. You can also use InboxAlly’s free Placement Tester.
- Check Postmaster Tools and campaign data. Look at complaints, reputation, delivery errors, unsubscribes, clicks, and replies. Do not use a single metric as a complete diagnosis.
If you use InboxAlly, monitor seed results by provider and by message. They can help you investigate trends, but they should be interpreted alongside signals from real recipients.
Do I need to switch domains or platforms?
I would not start there.
If you have spam complaints, for example, switching domains or email service providers without understanding why people are complaining may carry the same problem into the new setup.
First, review whether your contacts consented to receive the communication, recognize the sender, and receive content that matches the expectations set when they subscribed. Also check that unsubscribing is easy.
Depending on the signals, it may make sense to start the next campaign with a smaller group of subscribers who have shown recent interest, monitor their response, and then decide how to expand. This is an iterative process: form hypotheses, run tests, and see whether the results support those hypotheses. You may need to repeat it several times before identifying the root cause.
The signals complement each other
Seeds, personal mailbox tests, Postmaster Tools, and campaign metrics answer different questions. When they appear to disagree, the solution is to understand what each one is measuring.
Good seed results can be part of the answer. Complaints and reader behavior help complete the diagnosis.
Have you ever seen different results between your seeds and your recipients’ mailboxes? How did you investigate?
Until next time,
Fabrício
Sources
- InboxAlly — the email deliverability tool I recommend.
- Gmail sender guidelines
- Gmail sender guidelines FAQ
- Postmaster Tools dashboards
- InboxAlly’s free inbox placement tester
Disclosure: Some links here are affiliate links. You pay the same price, but I may earn a small commission. This helps me continue testing and recommending only what I genuinely trust and use in my day-to-day work.
