Out-of-office notices, not human replies, are the single biggest category of cold email responses
Automated absence notices outnumber every other kind of reply, including genuine human responses, making them the single largest category in the dataset.
43.3%
Ask a sales team what happens when they send a cold email campaign and most will describe two outcomes: someone replies, or someone doesn't. This study's labelling of 2,418 inbound replies (drawn from 2,419 rows, with 1 excluded) shows a third outcome that is actually the most common of all: an out-of-office autoresponder.
Out_of_office replies, defined in this study's rubric as automated absence notices, accounted for 43.3% of every reply received, 1,046 replies out of 2,418. That makes it larger than genuine human replies (34.6%) and larger than the other catch-all automated category, auto_reply_other, which covers bounces, ticket receipts, and departure notices, at 22.1%. Each reply was classified independently by Claude Haiku 4.5 against a written rubric, not by a human reviewer, over a measurement window from 2025-12-01 to 2026-10-19.
Why out-of-office replies dominate. An out-of-office autoresponder fires immediately and automatically, with no judgment involved, whenever an email lands in an inbox with the feature turned on. A genuine reply requires a person to open the email, read it, decide it is worth responding to, and write something back, a much higher bar. Across any large-enough send, the mechanical trigger will almost always outnumber the deliberate one, especially since autoresponders are often set for extended periods such as parental leave or travel. The sheer size of this category is less about email quality and more about the base rate of how many people, at any given moment, have an auto-reply configured.
Why it is worth separating from genuine replies. A campaign dashboard that reports total replies without separating this category will systematically make a campaign look more engaging than it is. If a list happens to be mailed during a period with more people on leave, say around major holidays, the "reply rate" could spike for reasons that have nothing to do with the email's subject line, copy, or targeting. Teams that want to compare campaigns fairly need to compare genuine reply rates, not total reply rates.
How much this varies. The out-of-office share is not a fixed constant. Across the 74 campaigns studied, it ranged from 18.2% at the low end up to 82% at the high end, with a median of 47.1%. Some of that spread likely reflects when a campaign was sent, holidays and seasonal leave periods would be expected to push this figure up, though this study did not isolate send-timing as a variable, so that explanation remains plausible rather than confirmed. Some of it may also reflect list quality or industry: a list skewed toward roles with predictable travel schedules could differ from one skewed toward desk-based roles. Distinguishing those explanations would require comparing campaigns sent to similar lists at different times of year, which this dataset does not isolate on its own.
What is clear without further analysis is the headline fact: when a cold email gets a response, the single most likely kind of response, more likely than a human reply and more likely than any other category, is an automatic notice that the recipient is not even there to read it.
Other findings
What percentage of cold email replies are actually written by a human?
Only about a third of cold email replies are written by a person
A labelling study of inbound replies to cold outreach campaigns finds that most of what lands in a sales inbox is not a person responding at all.
Do campaign sentiment labels like "interested" or "not interested" line up with whether the reply was actually from a person?
Does labeling a reply "interested" or "not interested" tell you whether a human wrote it?
Replies tagged with an outcome label are almost always genuine, while the large pool of untagged replies is mostly automated, which suggests sentiment tagging in this dataset is already filtering out the noise.
Is the genuine reply rate consistent across campaigns, or does it swing wildly depending on which campaign you look at?
How much does the human-reply rate vary from one cold email campaign to the next?
The share of replies that come from an actual person ranges from under a tenth to over half depending on the campaign, which means a single campaign's reply quality tells you little about the next one.
Want to know how AI answers describe you?
We run the same measurement on your category. Fifteen minutes with founder Omar Jenblat, your own numbers, no deck.
- Your category measured the same way
- Your own numbers, not a sample deck
- Fifteen minutes, no obligation
