An out-of-office reply is not a real reply, and it is not a wasted send either. Count it separately from replies, watch it as a rate, and read it for the return date and any named colleague. Done that way it protects your numbers and sometimes hands you a second contact.
Why shouldn't out-of-office replies count as replies?
They inflate your reply rate and say nothing about interest. I plan against positive replies, as in how many cold emails it takes to get one positive reply, and an auto-responder is not a positive. The outbound numbers nobody shows you makes the same point: reply rate on its own is close to meaningless until you strip the noise.
One external figure, with a caveat. FirstSales, citing Instantly's 2026 benchmark, puts auto-replies and out-of-office messages at 45.1% of cold email replies. That page does not include the underlying data, so I treat it as a reason to check your own number, not a benchmark to plan against.
What out-of-office rate means the calendar is the problem?
My line is 2% of sends. Above that, buyers are away, and pushing more volume into empty rooms is the wrong move.
The real case: I sent 5,000 emails into August and got back over 100 out-of-office replies and zero interested. The copy, list and infrastructure were fine. I had launched into the quietest fortnight of the American business year. The full story is in best time to send cold email, and the out-of-office check is step seven of the diagnostic order.
What do I do with an out-of-office reply?
Treat it as data with four fields:
| Field | Why it matters |
|---|---|
| Return date | Tells you when to resume the sequence |
| Named delegate | A second contact at the same company |
| Reason for absence | Leave and travel are temporary |
| Role change or departure | The address may be dead, so find the replacement |
Then act:
- Pause that contact. Do not delete them.
- Reschedule the next step for one or two business days after the return date.
- If a colleague is named, send them a separate short email saying the original contact is away, and keep the ask small. Do not imply the original contact endorsed you.
- If they have left the company, stop and research the replacement.
- Store the return date as a date field, not a note nobody reads.
This also changes how you count follow-ups. A paused contact has not been skipped, so the schedule in how many follow-ups a cold email sequence should have shifts by the length of the absence instead of ending early.
How do I calculate the rate?
Out-of-office rate = auto-replies received divided by emails sent, not divided by replies. The August campaign worked out like this: 100+ auto-replies on 5,000 sends is 2% or more, right at my line, with zero interest underneath it. The signal was the calendar, so the fix was pausing and relaunching later, not rewriting the email.
Your sending tool may lump auto-replies in with real replies. If it does, tag them with a rule (subject lines containing "out of office", "automatic reply" or "away") and review a sample by hand each week. The rules will miss some, and the sample tells you how many.
Which auto-replies are not out-of-office?
Three other kinds show up in the same bucket and need different handling:
- "I no longer work here" notices. Mark the contact as left, and find the replacement.
- Ticketing system acknowledgements. The address is a shared inbox, not a person. Remove it from the list.
- Vacation replies with no date. Treat them like out-of-office and check back in two to three weeks.
Mixing these up inflates your out-of-office rate, so separate them before you read the number.
Where this advice stops working
Not every sending tool lets you pause one contact and resume later. If yours does not, remove the contact from the sequence and re-add them with a note and a date. And if your out-of-office rate is low but replies are still thin, the cause is elsewhere. Start with placement and authentication in the deliverability audit, not with auto-replies.
To see how I read a whole campaign, start with the outbound numbers nobody shows you.