DELIVERABILITY METRICS
Cold email deliverability metrics and benchmarks
The numbers that actually tell you whether cold email is working — and the one number that doesn't.
9 min read01
Why metrics matter beyond “did it send”
A send confirmation only tells you the message left your outbox. It says nothing about whether it reached a human, or even reached an inbox at all. Reading the right metrics is the difference between knowing a campaign is working and assuming it is.
Open rate is the metric most people reach for first, and it's the least reliable one available. Apple's Mail Privacy Protection pre-loads email content for Apple Mail users the moment a message arrives — before a human ever looks at it. That pre-load registers as an open. The result is an open rate inflated by every Apple Mail recipient on the list, regardless of whether anyone actually opened the message. Any deliverability read built on open rate alone is built on a number that no longer measures what it used to.
This page is the benchmarks companion to the full cold email deliverability guide — that guide covers how to fix deliverability; this one covers what to measure and what the numbers mean.
02
Sender reputation, defined
Sender reputation is the trust profile a mail provider builds for a sending domain or IP over time. It isn't a single score you can look up in one place — it's an ongoing judgement, built from sending history, recipient engagement, and complaint signals, that a mail provider updates continuously and uses to decide where your next message lands.
That history has to be built deliberately, which is why a new inbox goes through a warmup ramp before it ever sends real cold outreach — see the warmup metrics and ramp schedule guide for what to track while that reputation is being built. Once outreach begins, that same reputation keeps being written to with every send, so it's worth confirming a domain's current standing directly rather than assuming nothing has changed.
Check a domain's current standing → Blacklist Checker
03
Core metrics and thresholds
Inbox placement rate
The percentage of sends that land in the primary inbox — not promotions, not spam. Industry benchmarks put the average at 83–84% across major providers, meaning roughly 1 in 6 legitimate emails never reaches the inbox. This figure is covered in full in the deliverability guide — it isn't re-derived here, only reused.
83–84%
average inbox placement rate across major providers
Bounce rate
A high bounce rate signals sending to addresses that were never verified. The same three-zone thresholds apply here as on the deliverability guide.
Below 2%
Safe zone
2–3%
Caution
Above 3%
Danger zone
Spam complaint rate
Google's own Email Sender Guidelines (support.google.com/a/answer/14229414) set two thresholds for senders emailing Gmail addresses. Senders should keep their spam rate below 0.1%. It should never be allowed to reach 0.3% or higher — that's Google's enforcement threshold for bulk senders (5,000+ emails/day to Gmail), and bulk senders above it lose eligibility for mitigation.
Below 0.1%
Google's recommended target
0.3% or higher
Google's bulk-sender enforcement line
Reply rate as a deliverability indicator
There's no verified industry benchmark figure for reply rate worth citing here, so none is given. What matters is the direction of the signal: an unexpectedly low reply rate is more often a sign of a placement problem — mail landing in spam rather than the inbox — than a messaging problem. Treat a reply-rate drop as a deliverability signal to investigate first, before assuming the copy or targeting is at fault.
04
What spam filters actually weigh
Spam filtering isn't one check — it's several signals evaluated together. These are the ones that matter for cold outreach.
Authentication status
Whether SPF, DKIM, and DMARC all pass for the sending domain. A single failure here undermines every other signal.
List quality and bounce behaviour
Sending to unverified addresses produces bounces, and bounces are one of the strongest negative reputation signals a mail provider tracks.
Complaint reports
Recipients marking a message as spam is a direct, explicit negative signal — and the one Google's spam-rate thresholds are built around.
Sending pattern and volume consistency
Sudden spikes in volume, or sending at machine-precision timing, look automated. Consistent, gradual, human-paced sending looks like a genuine mailbox.
Authentication is worth confirming directly rather than assuming it's still correctly configured — see the DNS Health Checker. The pre-send checklist covers all four of these signals as checks to run before a campaign launches, not just after something goes wrong.
05
Trusting the right numbers
Inbox placement rate is the reliable metric here — not open rate. Open rate is distorted by Apple Mail's pre-loading behaviour in ways that have nothing to do with whether a human read the message. Bounce rate, spam complaint rate, and reply-rate direction all give a clearer read on what's actually happening to a campaign than open rate ever will. When the numbers disagree, trust placement over opens.
ForgeSend tracks these signals per inbox
Per-inbox health scoring covers DNS configuration, bounce rate, and blacklist status — and the pre-send guardrail checklist blocks a campaign from launching if any of them fail.
Start free trial →