Cold Email Benchmarks: Reply Rates From a Live Outbound Operation

October is still running: these figures cover 1 to 27 October 2026 and will be replaced by the full month in the next refresh.

Month by Month, Cold Campaigns

Replies per 100 cold emails sent, every step of every sequence counted.

"People contacted" is Smartlead's unique lead count per campaign, summed. It is n/a for September because Smartlead's campaign report timed out on the full month and had to be pulled in three ten-day windows; a person emailed in two windows would be counted twice, so that column is left blank rather than overstated. Sent, replied, bounced and positive are unaffected.

How Much Campaigns Differ From Each Other

An average hides a lot. These are the per-campaign replies per 100 emails, for cold campaigns that sent at least 300 emails in the month.

Replies per 100 emails, so a campaign at 2.0 got two replies for every hundred emails it sent that month, across all steps of its sequence.

What the Replies Said

Smartlead tags every reply with a category. This is the split for October 2026 so far, across all sending (cold plus the warm lists and out-of-office re-sends below), because categorisation happens at the lead level, not the campaign level.

586 categorised replies. 22 of them were a meeting request or a booked meeting.

"Meeting booked" is a reply that turned into a meeting in the diary. "Do Not Contact" is anyone who asked to be removed; they are suppressed across every campaign the same day.

Meeting Requests by Month

Meetings that get booked from a reply by phone or LinkedIn after the email are not in this column; it only counts what was tagged in Smartlead. The real meeting count is higher and will come from the reply-handling system in a later refresh.

All Sending, Including Warm Lists and Re-Sends

For completeness, this is everything the workspace sent in the month: cold campaigns plus out-of-office re-send subsequences, reactivation of people who already know us (past clients, event attendees, webinar registrants, identified website visitors), client nurture lists, automated referral follow-ups and inbox tests.

The last two columns are Smartlead's account-level monthly totals, which count sends slightly differently from the campaign report (they include campaigns since deleted and sit on different day boundaries). Both are shown so you can see the gap rather than wonder about it.

How These Numbers Are Counted

• Source. Smartlead's campaign analytics for the Sales Boutique agency workspace, pulled through the Smartlead API with the dates set to Europe/London. One workspace, every client campaign in it.
• Emails sent counts every email in every step of a sequence, so one person in a three-step sequence can be three sends.
• People contacted is Smartlead's unique lead count for the campaign in that month, summed across campaigns.
• Reply rate per person contacted is replies divided by people contacted. This is the number most people mean by "reply rate". Replies per 100 emails is replies divided by emails sent, times 100, and is the stricter of the two.
• Positive replies are the leads Smartlead marked positive (interested, information request, meeting request, meeting booked). Automated out-of-office replies are counted as replies by Smartlead and sit in the neutral bucket; we don't strip them out of the reply count, so the reply rates here are slightly generous and the positive share is slightly conservative.
• Bounce rate is bounced emails divided by emails sent.
• Cold means net-new prospects. Out-of-office re-sends, subsequences, warm lists, client nurture lists, referral follow-ups and inbox tests are classified by hand from the campaign setup and excluded from the cold figures.
• Open rates are not shown because we send with open tracking off; tracking pixels hurt deliverability and the numbers they produce are mostly bots anyway.
• Sending setup behind these numbers: secondary domains per client (never the main domain), several mailboxes per domain, warmed before use, low daily volume per mailbox, every list verified before upload.