Your sequencer dashboard says 92% delivered. Your warmup pool says 95% inbox placement. Both numbers are technically true and both are nearly useless, because neither one measures where the email actually lands. Jesse Ouellette, founder of the cold-email infrastructure company LeadMagic, published two posts on the LeadMagic blog in late May 2026 making the same point from different directions: a sequencer's "delivered" status just means the receiving server accepted the SMTP handshake, not that a human ever saw the message in their inbox. And warmup pools report inbox placement based on how your domain performs against other domains in the same pool, not against how Gmail, Outlook, or Yahoo route mail to real prospects. An email deliverability audit is the only way to find out which number is lying. Most teams never run one until replies stop coming in.
Why your warmup pool's number is not your real number
Cold email deliverability lives and dies by two dashboards most teams never cross-check. Pooled warmup services work by cycling your domain through a shared network of seed accounts that open, reply to, and star each other's messages. That activity teaches Gmail, Outlook, and Yahoo that your domain sends mail people want, and it works, up to a point. The problem, according to Ouellette's LeadMagic post "Your Warmup Pool Is Lying: Cold Email Delivery in 2026," is that the inbox-placement score the pool reports back to you is measured against those same seed accounts and the pool's own engagement patterns, not against how a stranger's Gmail account, with no relationship to your domain, actually filters your mail. A domain can look "healthy" inside the pool's dashboard while landing in spam for every real prospect on a cold list.
The companion post, "Your Sequencer Says Green, Gmail Says Spam. Audit Now.," makes the same argument from the sending side. A sequencing platform marks an email "delivered" the moment the receiving mail server accepts it over SMTP. That's a real event — it means the message wasn't bounced — but it says nothing about which folder it landed in. Ouellette's framing is useful because it names the exact gap most cold email teams never check: "delivered" and "in the inbox" are two different facts, tracked by two different systems, and almost nobody reconciles them.
The pool mechanism itself is not the problem. Shared seed networks are how most warmup tools build sending reputation quickly, and a domain warmed through a pool generally performs better than one sent cold from day one. The problem is treating the pool's own scoring as your only feedback loop. A pool measures engagement inside a closed system it controls — seed accounts that were configured to open and reply on a schedule. Real prospects on a real list don't behave like seed accounts. They ignore mail, mark some of it as spam, and never open plenty of it even when it lands in the inbox — none of which the pool's dashboard is built to detect, because it isn't looking at your real sends at all.
The reassurance is exactly why teams stop checking. A dashboard that always reads green trains you to stop looking for the problem it isn't built to show you. That's a worse position than having no dashboard at all, because at least a team with no dashboard knows it's flying blind and tends to double-check manually out of habit. A team staring at a 95% placement score has no reason to open a seed inbox and look, so the gap between reported and real can widen for weeks before anyone notices it in reply rates.
This matters more with every ISP crackdown cycle. We've written before about how Microsoft's bulk-sender enforcement escalates automatically based on engagement signals like open rate and spam-complaint rate. If your dashboard says you're sending clean mail while your real inbox placement is quietly degrading, you won't see the throttling coming until send volume drops for reasons your sequencer can't explain.
The 5-check email deliverability audit
Something Inc. runs the same five checks on every cold email account before we scale a client's send volume. None of them require enterprise tooling. All five take longer than reading a pool dashboard, which is the point — a number you can generate in five seconds by refreshing a screen is a number that was never designed to catch a problem, only to reassure you one doesn't exist.
“The dashboard that's supposed to warn you when something breaks is usually the same dashboard hiding the fact that it's already broken.”
What the gap actually looks like
Here's an illustrative example, not a measured result: say a pool dashboard reports 95% inbox placement for a domain you're about to scale. You run the manual seed-account test across Gmail, Outlook, and Yahoo instead of trusting the pool's number, and the real split comes back 71% inbox, 22% spam, 7% promotions. That 24-point gap between the reported number and the measured number is exactly what check one — the manual seed test — is built to catch, and exactly what a pool dashboard has no reason to surface on its own.
| METRIC | POOL DASHBOARD (REPORTED) | MANUAL SEED TEST (MEASURED) |
|---|---|---|
| Inbox placement — illustrative example | 95% | 71% |
| Spam folder — illustrative example | not tracked | 22% |
| Promotions tab — illustrative example | not tracked | 7% |
What to fix when the audit finds a gap
If checks three and four turn up broken or missing authentication, fix those first. An unaligned DMARC record or a flagged IP will sink inbox placement no matter how carefully you warm the domain afterward. Correct SPF and DKIM, move DMARC to at least p=quarantine, and if the IP or domain shows up on a blacklist, work through that provider's delisting process before you send another campaign from it.
If authentication and reputation are clean but the manual seed test still shows mail landing in spam, the problem is usually content or list quality rather than infrastructure — spammy subject lines, tracked link redirects, attachments, or a list with a high rate of invalid or unengaged addresses. Fix the list and the copy, re-run the seed test, and only resume full send volume once inbox placement holds across two or three sends in a row, not one clean result you're eager to trust.
How often to run this audit
Run the full 5-check audit before you scale any new sending domain past initial warmup, and again any time open rates or reply rates drop without an obvious cause — a new list, a new offer, a holiday week. Those are the moments teams blame the copy or the list when the actual cause is inbox placement quietly sliding toward spam.
Volume changes are the other trigger worth building into the calendar. Ramping a domain from a few dozen sends a day to several hundred, adding a second or third mailbox on the same domain, or moving a client from a pilot list to full territory coverage are all moments where reputation that looked stable at low volume can crack under load. Run the audit before the ramp, not after the first week of results comes in soft.
For established domains already sending at volume, check items three and four — authentication alignment and reputation — monthly. They rarely change on their own, but a shared hosting neighbor, an expired DKIM key, or a lapsed DMARC record can break them without any alert firing on your side. Check items one, two, and five — the manual seed test, folder placement, and engagement throttling exposure — weekly during any active campaign. Engagement metrics move fastest, and they're the earliest signal an ISP is starting to throttle you, well before send volume visibly drops.
Assign the audit to one person, not a shared responsibility between whoever manages the sequencer and whoever manages the domains. In practice those are often different people, sometimes different vendors entirely, and a gap between "delivered" and "inbox" is exactly the kind of problem that falls through a seam nobody owns end to end. If you're running this across a portfolio of client domains rather than one company's list, standardize the checklist so all five checks run the same way on domain four hundred as they did on domain one.
None of this is an argument against cold email as a channel — cold email isn't dead, it's just harder to run on autopilot than it was three years ago. It's also an argument for treating the first message in any sequence as the one that has to land clean; we've written about why the first email in a sequence wins most of the replies, and that's doubly true when a chunk of your list never sees it because it landed in spam. For B2B teams running outbound at scale, the audit is cheaper than the pipeline lost when a domain quietly goes dark.
Run this email deliverability audit before your next send
Pick one domain sending live campaign volume right now. Run the manual seed test across Gmail, Outlook, and Yahoo today, not after the next campaign underperforms. Check the DMARC policy — if it's still p=none, move it to p=quarantine this week. Pull last week's open and complaint rates and write them down somewhere you'll actually look at again. If any of the five checks turn up a gap like the illustrative one above, fix authentication and reputation issues first, then re-warm before you resume full volume. If you can't say with certainty what folder your last campaign landed in for a sample of real recipients, you don't have a deliverability number — you have a delivery number, and those are not the same thing. If you want a partner to run this audit and manage the sending infrastructure on an ongoing basis instead of doing it by hand every week, that's what our cold email program is built for.
See where you are cited today
A free snapshot audit of your rankings and AI citations before we ever talk.
Tyler leads work at the intersection of SEO and generative engines at Something Inc., helping B2B brands get ranked and cited across every major AI engine.