Inbox placement: testing where your email actually lands

An inbox placement test sends your actual campaign email to a panel of seed mailboxes across Gmail, Microsoft 365 and other major providers, then reports where each copy landed: the inbox, the spam folder, or nowhere at all. It answers the one question your sending dashboard cannot — whether anyone can actually see your email. A message filed in spam still counts as "delivered", which is why placement testing exists as a separate discipline: it is the final check on the infrastructure chain — authentication, warm-up, volume discipline, list hygiene — set out in the practical deliverability guide.

Why doesn't delivery rate tell you where email lands?

"Delivered" means the receiving server accepted the message. Filtering happens after acceptance: Gmail or Microsoft 365 takes your email in, runs it through reputation and content checks, and only then decides whether a human ever sees it. Your sending platform can report a 99% delivery rate while half the campaign sits in junk folders.

Open rates do not rescue the picture. Privacy features and image pre-fetching now register opens that never happened and hide opens that did, so open data is unreliable in both directions. The result is a campaign that looks healthy on every dashboard while quietly reaching nobody.

This is the most common pattern behind "cold email doesn't work for us". The copy is fine, the list is fine, and almost nothing is being seen — the channel takes the blame for an infrastructure failure, which is the argument of cold email isn't dead; bad cold email is. A placement test is how you separate the two cases before you rebuild the wrong thing.

How does a placement test actually work?

The mechanism is straightforward. When you run a test, the tool gives you a set of seed addresses — real mailboxes it controls at Gmail, Google Workspace, Microsoft 365, Outlook.com and the larger independent providers. When you send your actual campaign email, from your actual sending mailbox, to those seeds, then each seed reports where the message arrived. When the results compile, then you read three numbers: inbox rate, spam rate, and missing rate — messages rejected or silently dropped. When you segment those numbers by provider, then you see the pattern that matters: landing cleanly at Gmail while junking at Microsoft is common, and it points to a reputation problem with one filter stack rather than a general failure.

One caveat. Seed mailboxes have no relationship with you, while some real recipients have engagement history that filters weigh in your favour. Treat a placement test as a smoke alarm rather than a laboratory instrument — imprecise by a few points, decisive about direction.

What score should make you act?

Providers publish no thresholds, so these are working rules rather than gospel. Consistently above 90% inbox across the major providers: the infrastructure is doing its job. High seventies to mid-eighties: investigate — usually one provider is dragging the average down. Below roughly 50%: stop the campaign, because continuing to send teaches filters that your mail belongs in spam.

The trend matters more than any snapshot. A drift from 95 to 85 over three weeks is a louder alarm than a single reading of 80, because reputation decay compounds — the earlier you intervene, the cheaper the fix.

When should you test?

• Before campaign one. A warm-up score — a properly warmed mailbox shows something like 91/100 on the common reputation tools — tells you the mailbox has history; a placement test tells you it lands.
• After any change. New domain, added mailbox, rewritten template, volume increase.
• On a cadence. A fortnightly or monthly check per sending domain is typical, and cheap next to what it catches.
• When replies sag. We hold campaigns to a ~4% positive-reply expectation and treat anything below 3% as broken. Placement is the first check, before anyone touches the copy — rewriting an email nobody sees fixes nothing.

What do you fix when placement drops?

Work the checks in order of likelihood, not preference:

1. Authentication. Confirm SPF, DKIM and DMARC all pass; a DNS change or a misconfigured mailbox breaks them silently.
2. Blacklists. If the domain or IP is listed, follow the recovery protocol before anything else.
3. Volume. Spikes and irregular sending are precisely what filters watch for; hold to the daily sending limits that keep you invisible to filters — 25–40 emails per day per inbox, steady, every weekday.
4. List quality. A rising bounce rate poisons reputation faster than any copy mistake.
5. Content, last. It is usually the smallest factor, and the one people reach for first.

Placement testing is unglamorous engineering — nobody ever won a pitch with a seed-list report. But it is the difference between managing a channel and guessing at one.

---

Next step: the Growth System Audit — £450, seven days, credited against any build — maps where your growth system leaks and what to build first.