Most cold email reply rate benchmark posts are written by companies that sell sending tools. When we decided to test cold outreach at techpotions, we logged every send, stared at a 0% reply rate, and shut the whole thing down on day two.

This is the test log, the infrastructure reality, and the decision rule we walked away with. If you’re a small studio or agency asking whether to fund a cold-email motion, the math here might save you a domain.

What We Actually Sent

33 cold emails. Two days. Five segments. Zero genuine replies.

One hard bounce, which puts the bounce rate at about 3%—normal. Batch 1 was the warmest audience you’ll ever get in a cold list: people who literally posted “email me” inside a Hacker News hiring thread. They still didn’t reply.

We sent no follow-ups to batches 2 through 4, cancelled batch 5, and moved on.

The Infrastructure Excuse Doesn’t Fly Here

Whenever a cold email test returns zero replies, the first objection is always “your deliverability was broken.” We’ll save you the suspense: ours wasn’t, and even the mistakes we made didn’t cause the silence.

• SPF, DKIM (spacemail selector), and DMARC were all verified before sending.
• We ran a Port25 verifier probe and saw no red flags.
• The mailbox was only three weeks old and had zero warmup, and we hit 33 sends on day one. That’s exactly the pattern ESPs rate-limit, so we later set a policy of 15 sends/day in week one and 25 in week two, spaced 60 seconds apart. But here’s the thing: none of our messages bounced except that one hard bounce. They were delivered.
• We also shipped a genuine unforced error: wrong domain in the links across batches 1 through 4. Embarrassing, but it wouldn’t prevent a text-based reply, and none came.

Infrastructure wasn’t the bottleneck. The audiences just didn’t bite.

The Cold Email Reply Rate Benchmark That Matters

Published 2026 benchmarks give a healthy B2B cold reply rate somewhere between 1% and 6%. Instantly’s benchmark report across billions of emails puts the average at 3.43%, with top performers exceeding 10%. InboxKit’s dataset (combining warmup and campaign traffic) shows 33.4%, but that’s heavily inflated by automated warmup replies, not human engagement. Oppora.ai notes that B2B SaaS cold audiences typically see open rates of 25–35%, and agencies in professional services tend to hover around similar reply rates.

Our 0% on 33 sends is not an indictment of those benchmarks. It’s just a completely ordinary draw from a small sample. At 4% expected, 33 sends predicts 1.3 replies. Zero is well within the noise.

That’s the entire statistical point: you cannot kill a channel with 33 sends, but you can learn what it would cost to get a real answer.

The Decision Rule: Cost to Significance

We didn’t retire cold outreach because the data proved it dead. We retired it because the cost of reaching statistical significance—several hundred sends, weeks of warmup, ongoing domain risk on a mailbox we depend on—was higher than the cost of the alternatives we already had.

A small test can’t tell you a channel works. It can, however, tell you what it would take to know. Here’s the rule of thumb:

• At an expected reply rate of 3–4%, you need roughly 200–300 sends to even start seeing a handful of replies.
• Factor in warmup, list building, copy iterations, and the risk of domain reputation damage if you push too fast.
• Compare that total cost (time + money + risk) against the next-best channel you could fund instead.

That comparison is the real output of any small test, and it’s where most teams get stuck. Our Growth Labs service runs these exact small experiments for studios so you don’t gamble a domain on a motion that won’t pay back.

What We Did Instead

We moved budget and attention to channels that don’t require a cold audience to give a damn on day one:

• Warm demand: responding to people already asking for what we build.
• Reddit participation: being genuinely helpful in threads where our expertise matched the question.
• Past clients: reactivating relationships that didn’t need an intro.
• WhatsApp: direct, low-friction conversations with people who already knew us.

Zero cold emails since.

If you’re sitting on the fence about whether to stand up an outbound engine, start with a tiny, logged test and let the cost-to-significance rule make the call.

FAQ

What is a good cold email reply rate benchmark?

A healthy B2B cold email reply rate typically falls between 1% and 6%, with Instantly's 2026 benchmark averaging 3.43% across billions of sends. Top performers can exceed 10%, but that usually involves heavy personalization and list curation.

Does a 0% reply rate on 33 sends mean cold email is dead?

No. At an expected 4% reply rate, 33 sends predict about 1.3 replies, so zero is a completely ordinary statistical draw. A small test cannot prove a channel dead; it can only reveal the cost of getting to a statistically significant sample size.

What’s the biggest mistake small studios make when testing cold outreach?

Assuming infrastructure problems are always to blame. More often, the audience just isn’t interested, and the studio never runs the cost-to-significance math before burning a domain on a channel that doesn’t suit them.