Five times the conversations, and almost everyone gets an answer
The case where the result comes from volume rather than conversion. And where part of the data had to be thrown out.
Calls booked per month
78.3
235 in the last 90 daysAverage over the last 90 days, counted from bookings that carry an actual scheduled call time. No such figure exists for the period before installation, because nobody was recording bookings then.
- People who got a reply within 24 hours79.8%96.7%
- Conversations held1,2897,348
- Conversations with at least one follow-up40.2%46.6%
An account selling courses and mentoring to experts and creators who want to turn services into educational products. Qualification is strict, because the entry bar is set at a specific revenue level, so the inbox exists to filter most enquiries out before anyone reaches the calendar.
Nearly six times the conversations, and the share of people left without any reply fell from one in five to one in thirty.
The starting point: one person in five left hanging
In the year before installation the account held 1,289 conversations in which the lead wrote something in text. An answer within 24 hours reached 79.8 percent of them, meaning roughly one in five got none at all.
Follow-up covered 40.2 percent of conversations at 1.35 reminders per conversation. That is a better figure than the other accounts in this set showed before installation, and the explanation is simple: some replies were already going out automatically back then.
What the measurement showed afterwards
Over the last 90 days the account is holding 7,348 conversations, nearly six times as many, and the share of people answered rose from 79.8 to 96.7 percent. That is the actual content of this case: at several times the traffic, the share of people left without a reply fell from one in five to one in thirty.
Follow-up reached 46.6 percent of conversations instead of 40.2, but reminders per conversation fell from 1.35 to 1.04. Together those two numbers mean more people get a reminder today, just less often several times over. The drop in the average stays on the page, because it moves in the wrong direction.
What we discarded, and why
We do not report reply time for this account. In the before window the median is 17 seconds against a ninetieth percentile of 4.6 hours, and no human workload produces that distribution. It means some replies were already automated back then, probably by another tool, while others waited half a day. Comparing that median against ours would be comparing one automation to another and presenting it as human against tool.
The share of people answered and the follow-up share are immune to that artefact, because they measure whether a reply happened at all rather than how fast. So those stay and reply time goes.
Why this is not one hundred percent
Two groups are left without a reply, both of them on purpose. The first is friends and accounts the client themselves follows: the tool has a setting to leave them alone, because nobody wants an agent pitching their mates. The second is contacts rejected immediately against the disqualification criteria the client sets. On top of that there are conversations a human took over and never finished. One hundred percent on this metric would mean the tool answers everybody indiscriminately, and that is a defect rather than a feature.
Every metric we measured
The full table, including the metrics that did not improve. Sample sizes sit below it.
| Change | Year before Setor | Last 90 days | Change |
|---|---|---|---|
| People who got a reply within 24 hours | 79.8% | 96.7% | 16.9 pp |
| Conversations held | 1,289 | 7,348 | 5.7× more |
| Conversations with at least one follow-up | 40.2% | 46.6% | 6.4 pp |
| Follow-ups per conversation | 1.35 | 1.04 | 1.3× less |
What each of these numbers actually measures
- People who got a reply within 24 hours
- Share of conversations where a person wrote something in text and received a reply within 24 hours of their first message. It counts people, not messages.
- Conversations held
- Conversations in the window with at least three messages where both sides spoke.
- Conversations with at least one follow-up
- Share of threads in the window where the business wrote again after at least thirty minutes without a reply. The denominator is every thread, single-message ones included, so the figure is understated on both sides of the comparison.
- Follow-ups per conversation
- Average number of such reminders per thread.
What these numbers rest on
Before window
January 2025 → January 2026
1,289 conversations in window
3,768 reply pairs
After window
Last 90 days
7,348 conversations in window
23,127 reply pairs
Cutoff point
January 31, 2026
The first trace of Setor on this account.
A conversation counts when the lead wrote something in text. That is the denominator for every percentage on this page apart from the time based metrics. A pair is a lead's message and the business reply that follows it. That is the denominator for the time based metrics.
What we do not measure here
- Response time. A median of 17 seconds against a ninetieth percentile of 4.6 hours reveals that some replies were already automated back then, by another tool.
- Booked call rate. Without a trustworthy response time and with a before window this small, we do not compare it to anything.
How we counted thisSee the report on 1.58 million messagesAI setter versus a human setter
Want to see these numbers on your own account?
On the call we go through your inbox and show you how many messages currently go unanswered.