What is a good cold email open rate? (3,000+ emails of real data)

Few people in this industry publish their own cold email data. The reason is simple: if the numbers are good, you're accused of bragging; if they're bad, you lose customers.

We chose to publish, because the data is the best product description there is.

Of the three metrics, the most worth watching is the repeat open rate - 236%, or 23.6x the industry average of 10%. People who opened your email came back to read it again, an average of 3.36 times. Repeat reading is the most direct signal of interest density: reading once versus three times reflects completely different intent.

The unique open rate of 28.21% is above the 25% industry average, but not by a wide margin. Delivery is healthy; that's not the strongest lever.

Put the two together and the conclusion is the opposite of what most expect:

Content quality isn't the bottleneck - the CTA (call to action) is. If the content were garbage, the repeat open rate would be near zero; nobody re-reads a boring email. Every metric beats the industry, which means the content genuinely pulls. The real problem: we never gave readers a low-friction reason to reply.

Where the data comes from

Since TokoAI went live, we've sent over 3,000 traceable cold emails to manufacturing exporters. All send data is pulled directly from the system - no pixels, no third-party trackers.

The system records three layers of metrics:

• Unique open rate: how many distinct people opened the email
• Total open rate: all open events, including repeats
• Repeat open rate: total minus unique, divided by unique - how many times one person opened

The repeat open rate is the most ignored yet most valuable metric. Reading once vs. three times reflects entirely different decision intent - one is a glance, the other is real interest.

From "fix the copy" to "fix the CTA"

At first we shared the common instinct: the problem must be the content. We iterated 7 strategy rounds, wrote 8 copy variants, and tested 5 CTA phrasings.

Only when the system accumulated enough data did the repeat open rate surface and tell a different story:

Recipients signaled with behavior that the content had value (re-reading), and also that the current CTA deserved a rethink.

That changed everything. The optimization focus shifted from "write better content" to "design a more natural call to action."

What we're doing now

• Lower the CTA barrier: the CTA moved from "reply and let's chat" to a more natural ask that lets readers signal interest at the lowest cost. We stay on email because reply data flows straight back into the system to drive the next round.
• Give real value in the content: every email carries an insight specific to the reader's industry, drawn from the system's anonymous accumulation across similar companies, so it feels worth keeping.
• Auto-classify signals: a reply-signal classifier is deployed, separating buying signals, pending replies, and noise so no real opportunity slips through.

If this resonates

If you also run B2B cold email, a few questions worth asking yourself:

• Does your tracking come from an API or from gut feel?
• Are you watching unique opens or repeat opens? The latter tells you if content truly caught attention.
• Would you yourself take the CTA action you're asking for?

We don't have fixed answers - they differ by industry and product. But we have a system: send, track, analyze, iterate. This week beating last week is the signal we're right.