What is a good cold email reply rate in 2026? Honest benchmarks from 750K sends
Industry "benchmarks" tend to be inflated and self-reported. Here are the actual numbers from our first 6 months in production - 400 mailboxes, 750K cold sends, measured at our 10-20/day cap.
Rejwan Nirob·May 19, 2026·6 min readReply rate is one of the few cold-email metrics that resists gaming. Open rates can be inflated by pixel pre-fetching. Click rates can be inflated by auto-scanners. A reply requires a human reading the message and choosing to write back. That is what makes the number useful - and what makes it worth getting right.
Here is what reply rate actually looks like at infrastructure-good baseline. Pulled from our first 6 months in production: roughly 400 mailboxes provisioned, ~750K cold sends measured at our 10 to 20 per day per mailbox cap, across roughly 160 domains.
The infrastructure-good baseline
On clean infrastructure (SPF, DKIM, DMARC all green; 2 to 3 mailboxes per domain; 14 days of warmup; 10 to 20 cold sends per day per mailbox), the median cold-email reply rate we see in our customer fleet sits at 2.4 to 3.1 percent. That is reply rate measured as positive-or-negative replies divided by sends. Auto-replies and out-of-office are excluded.
The 25th percentile sits near 1.6 percent. The 75th percentile sits near 4.8 percent. Above 5 percent reply rate is genuinely strong copy on a well-targeted list. Above 8 percent usually means a very tight ICP and a personalized opener; it does not scale.
What kills reply rate, in order
- Broken DNS auth (SPF/DKIM/DMARC). The mail lands in spam. People do not reply to spam.
- Bad list. Hard-bounce rate above 4% signals you bought a list, not built one.
- Generic opener. The first three words decide whether the message is read. "Hope this finds you well" decides zero replies.
- Wrong ICP. Sending CFO-grade copy to AEs gets ignored. Sending AE-grade copy to founders gets blocked.
- Bad timing. Tuesday-Thursday morning is the safe window for B2B; Friday afternoon is the dead zone.
Note the order. Three of the top four are not copy problems. They are upstream problems - infrastructure and list - that cap your reply rate before your copy ever gets a chance to land.
When to suspect infrastructure vs copy
“If your reply rate is below 1% on a list you trust, the bottleneck is almost never the copy. The mail is not being read.”
Run a quick diagnostic before iterating on copy. Send the same campaign to a seed list of 20 inboxes you control across Gmail, Workspace, Outlook, and M365. If placement is below 70% primary inbox on the seed, the campaign is being filtered. Fix that first.
If placement is healthy and reply rate is still under 1%, then it is copy or targeting. At that point A/B the opener line and the call to action; do not waste cycles on subject lines until the opener is dialed in.
What boosts reply rate
- Hyper-targeted ICP - 200 carefully-selected prospects out-replies 5,000 sprayed ones every time
- Personalized opener referencing something specific to the prospect
- Direct CTA - one ask, one verb, one calendar link or one yes/no question
- Short body - under 90 words for B2B cold; the reply rate curve drops sharply above 120
- Follow-up sequence - 60% of replies arrive on follow-up 2 through 4, not on the first send
Reply rate lives downstream of infrastructure quality. Inboxlee handles the upstream - clean DNS, automatic warmup tracking, daily placement testing, blacklist monitoring - so reply rate variance comes from your copy and your list, not your sending stack.
See infrastructure monitoringFrequently asked
What is a good cold email reply rate in 2026?
On clean infrastructure with a well-targeted list, the median cold-email reply rate sits at 2.4 to 3.1 percent (positive or negative replies divided by sends, excluding auto-responders). The 25th percentile is near 1.6 percent; the 75th near 4.8 percent. Above 5 percent is strong copy on a tight list. Above 8 percent is rare and usually does not scale.
Why is my cold email reply rate so low?
Three of the top four causes are not copy. Broken DNS auth (mail in spam, no replies). Bad list (high bounce, low trust). Wrong ICP (right message, wrong audience). Generic openers are the only copy-level cause in the top four. If reply rate is below 1 percent on a list you trust, the bottleneck is almost always upstream - check inbox placement on a seed list before iterating on the opener.
How do I know if my reply rate problem is infrastructure or copy?
Send the same campaign to a seed list of 20 inboxes you control across Gmail, Workspace, Outlook, and M365. If primary-inbox placement on the seed is below 70 percent, the bottleneck is infrastructure - fix that before iterating on copy. If placement is healthy and reply rate is still under 1 percent, the bottleneck is copy or targeting.
Should I focus on reply rate or open rate as the headline metric?
Reply rate. Open rates are heavily distorted by pixel pre-fetching (Apple Mail Privacy Protection inflates them 10-30 percentage points without any human reading the message). Reply rate requires a human to read and write back, which is the actual signal you care about. Use open rate as a directional check only.
How long does a cold email follow-up sequence have to be?
Three to five messages spread over two to three weeks captures the majority of replies. Roughly 60 percent of cold-email replies arrive on follow-up 2 through 4, not on the first send. Beyond five messages the marginal reply rate per message drops sharply and the unsubscribe rate climbs.