Cold Email Inbox Placement Audit: A 2026 Checklist

What inbox placement actually measures

I'm Hlib Storchak. I build and run outbound systems for B2B founders and sales teams, and I've booked over 2000 meetings for B2B clients doing it. Most of that work is not copywriting. It's making sure the email actually gets read before anyone judges the copy at all.

Inbox placement rate is the percentage of sent emails that land in the primary inbox, as opposed to the spam folder, a promotions tab, or nowhere at all. It is not the same thing as delivery rate, which just means the receiving server accepted the message without a hard bounce. A mail server can accept your email and still drop it straight into spam. Delivered and placed are two different outcomes, and most teams only ever measure the first one.

That gap matters because your sending tool's dashboard almost never shows you placement. It shows sends, opens (unreliable on its own), bounces, and replies. None of those tell you what fraction of your list never even saw the message land where a human would check it. You have to go looking for that number on purpose.

The 2026 benchmark: 87.2%, not 84%

Validity's 2026 Email Deliverability Benchmark Report, built from its own inbox monitoring network, puts the global average inbox placement rate at 87.2% for 2025, a 3.7 percentage point improvement year over year. That means roughly 1 in 8 legitimate emails, cold or otherwise, still never make it to a primary inbox globally. The report credits the improvement mainly to fewer emails being blocked or rejected outright, as major mailbox providers got better at distinguishing legitimate senders from spam operations rather than just filtering everyone harder.

The average hides a wide spread by provider. This is the table I actually check against, not the headline number.

That Microsoft number is the one I'd flag first if your ICP skews toward mid-market or enterprise B2B, since that's exactly where Office 365 concentrates. A list that is 60% Microsoft-hosted domains is not going to perform anywhere near an 87.2% average, no matter how clean the rest of your setup is.

The 84% number you might still be quoting

Here's a detail worth catching before you build a checklist off the wrong year. Validity's 2025 average is 87.2%, up 3.7 points year over year. Do the subtraction and the prior year's average lands around 83.5%, close enough to the "roughly 84%" figure that keeps circulating in 2026 cold email content, including in an earlier draft of this article's own topic line. That older number was probably accurate when it was published. It just isn't the current one anymore.

This is the same mistake I've flagged in a few other articles on this blog: a real, sourced stat from one year gets repeated into the next year's roundups without anyone checking whether the underlying report actually refreshed the figure. Before you cite a deliverability benchmark to a client or a boss, check the publish date on the report it traces to, not just the year printed on the blog post quoting it.

Step 1: run a seed test before you scale

A seed test sends your actual campaign email to a panel of real mailboxes across Gmail, Outlook, Yahoo, and a few others, then reports back where each copy landed: inbox, spam, or nowhere. Tools built for this, GlockApps and Mail-Tester are the two I see used most, exist specifically because your own sending platform cannot tell you this on its own.

Run one before you scale a new domain, a new copy angle, or a new sending tool past a handful of accounts. It takes minutes and it's the only step on this list that gives you a direct placement number instead of an inferred one.

What a seed test will not tell you

A seed panel is a sample, not your whole list, and mailbox providers treat known seed addresses somewhat differently than genuinely engaged real-world recipients over time. Treat it as a pre-send sanity check, not a substitute for watching your real campaign's numbers afterward.

Step 2: read Google Postmaster Tools and Microsoft SNDS

Both Gmail and Microsoft publish free tools that show you domain-level reputation and spam-rate data direct from the source, not a third-party estimate. Google Postmaster Tools reports your domain's reputation, spam rate, and IP reputation for Gmail specifically. Google's own bulk sender guidance keeps the spam complaint threshold at under 0.3%, with 0.1% the safer target. Microsoft's Smart Network Data Services (SNDS) and its Junk Mail Reporting Program (JMRP) do the equivalent for Outlook and Office 365.

Set both up once, for every sending domain, even the ones that feel like they're performing fine. These are the closest thing to ground truth you'll get without paying for a monitoring platform, and they're free.

Step 3: segment by provider, not just an average

Once you have a placement number, break it down by mailbox provider before you act on it. A blended 85% average can hide an 95% Gmail rate sitting next to a 60% Microsoft rate. If your list is a mix, that blended number tells you almost nothing actionable. The fix in the first case is nothing, the fix in the second case is a Microsoft-specific authentication and warmup review.

Most cold email platforms let you tag or filter sends by recipient domain type. Pull that cut before you decide anything based on an aggregate number.

Step 4: check blocklist status and the spam folder itself

Two more checks, both quick. First, run your sending domains and IPs through a blocklist checker (MXToolbox is the common free one) on a schedule, not just when something looks wrong. A single listing on a major blocklist can gut placement overnight, and you want to catch it same-day, not a week into a dead campaign.

Second, actually open the spam folder on a real test account you control. Automated tools tell you the rate. Looking at the folder tells you why: which subject lines, which senders, which content patterns are the ones actually getting filtered. That qualitative read is what turns a placement number into a fix.

Step 5: retest after every change that touches sending

New domain, new mailbox, new sending platform, a big jump in daily volume, a new copy template: any of these can shift placement, for better or worse. Run a fresh seed test after each one instead of assuming last month's number still holds. This is the step most teams skip, since it feels like busywork when the last test passed. It's also the step that catches a bad change before it burns three weeks of a domain's reputation instead of three minutes of a test send.

What a bad placement rate actually costs you

Here's a simple way to see why this is worth the twenty minutes a month it takes. All the numbers below are stated assumptions, not researched facts, swap in your own.

Assume you send 5,000 cold emails a month and your list is roughly a third Microsoft-hosted domains, common for a mid-market B2B ICP. At the global 87.2% average, about 4,360 of those land in an inbox. At Microsoft's 77.4% rate applied to that third of the list, with the rest closer to the Gmail and Yahoo numbers, your real blended placement likely sits closer to 82 to 84%, not 87.2%. That gap, roughly 150 to 250 emails a month, is invisible in your sending dashboard because those emails were "delivered," just not placed anywhere a person would see them.

Apply even a conservative reply rate, Instantly's 2026 benchmark puts the cold email platform average at 3.43%, to those missing 150 to 250 emails and you're looking at roughly 5 to 9 replies a month that never had a chance to happen, not because the copy was wrong, but because the message never reached a folder anyone opens. Over a quarter, that's a meaningful chunk of pipeline lost to a problem a seed test would have caught in minutes.

Where the Forge stack fits

This audit works the same regardless of what you send from, GlockApps and Postmaster Tools don't care whether the mail came out of Smartlead, Instantly, or anything else. Where I do have a preference, built from running this for clients, is on the domain and mailbox side underneath whatever sending tool sits on top: I default to Infraforge and Mailforge for provisioning domains and mailboxes, and Warmforge for the actual warmup and ongoing health signal, mainly because the health monitoring surfaces the exact provider-level reputation data this checklist asks for in one place instead of four separate tools. That's a personal preference from running the setup repeatedly, not a verdict on the alternatives, plenty of agencies run a perfectly healthy stack on other infrastructure providers.

The monthly audit checklist

This is the actual sequence I run for clients on a recurring basis, not just at launch.

FAQ