From two hours to one minute, three in the morning included

The starting point: one reply in five on time

Before installation the median reply time was 1.9 hours, and outside working hours 2.2 hours. Replies within five minutes covered 20.8 percent of messages, one in five. The ninetieth percentile was 16.9 hours, so the slowest tenth of conversations rolled over into the next day.

An answer of some kind within 24 hours did reach 89.8 percent of people, though. This account was not ignoring anyone, it was answering whenever the owner had a moment. Follow-up covered 45.2 percent of conversations at 0.74 reminders per conversation.

What the measurement showed afterwards

The median fell to 67 seconds, and outside working hours to 66 seconds. The second number is the more interesting one: nights and weekends stopped existing as a separate category. Replies within five minutes rose from 20.8 to 97.7 percent, and the ninetieth percentile fell from 16.9 hours to 2 minutes.

The share of people answered went from 89.8 to 99 percent, and follow-up from 45.2 to 71.7 percent of conversations. This is the gap you cannot close by hiring one more person, because a shift ends at five and the messages do not.

Why sample size gets its own section here

The pre installation window covers 166 conversations where the lead wrote something in text. That is few. Time based metrics, however, are computed not on conversations but on pairs of the form lead writes, business answers, and there are 558 of those. The median and the five minute share therefore rest on a reasonable base, especially given that the difference runs from 20.8 to 97.7 percent, a gap sample size does not explain.

Metrics computed across whole conversations are a different matter, meaning the share of people answered and the follow-up rate. There the denominator is 166, so a handful of conversations either way shifts the result by a point or two. Treat those as an indication, not as proof. We do not report a booked call rate for this account at all.

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.

What each of these numbers actually measures

Replies within 5 minutes

Share of pairs where the reply arrived within five minutes.

Median time to reply

The middle of the distribution of time between a lead's message and the business reply that follows it. Pairs longer than 24 hours are dropped, because that is a follow-up.

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.

Median reply outside working hours

The same figure, computed only for messages outside Monday to Friday 9:00 to 17:00, in the seller's timezone.

Ninetieth percentile time to reply

The boundary of the slowest tenth of replies: nine in ten arrive faster than this.

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.

Conversations held

Conversations in the window with at least three messages where both sides spoke.

What these numbers rest on

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

• Booked call rate. The before window holds 121 conversations, so any figure here would be guesswork.