Chatbot speed to lead: why leads still wait in working hours

The team is at work, and the lead still waits

A delay on a night-time message is understandable — nobody is at work. But in many companies the same delay happens during the day, in working hours: salespeople are at their desks, phones are on, and a lead sits unanswered for forty minutes. The cause is not laziness. Messages arrive on three channels, nobody knows which belongs to whom, the salesperson is on a call, and the notification is lost among others.

This article is about cutting sales response time in working hours: why speed matters, where delay comes from, what the bot's first reply should do, how the bot-to-person handoff SLA is built and what to measure. Leads that arrive outside working hours are a separate topic, covered in after-hours leads.

Why speed matters — and why it is not enough alone

The effect of response speed has been measured in independent research: in the MIT Sloan / InsideSales Lead Response Management study, a lead contacted within 5 minutes was about 21 times more likely to be qualified than one contacted after 30 minutes. The figure is about qualification, not sales, but the direction is clear: interest cools.

Yet speed is not a single metric. «Hi, how can I help?» can arrive in one second and change nothing. Speed is worth something when three things come together:

• A correct first reply — a real answer to the customer's question
• Recognised intent — is this a sales lead, a support question or a stray message
• A person on time — a salesperson joining a sales lead within a set time

Where delay comes from

1. 1Fragmented channelsInstagram on one phone, WhatsApp on another, website chat on a laptop. Nobody sees every message in one place.
2. 2Leads without an ownerThe message lands in a shared inbox and everyone assumes someone else will answer.
3. 3The salesperson is busyOn a call, in a meeting, with another customer. The notification arrives and is not seen.
4. 4No priorityA customer ready to order sits in the same queue as someone asking for the address.

None of these four is solved by making people faster. A system solves them.

A first-reply template for the bot

The bot does not replace a person in the first reply — it fills the gap until the person arrives and prepares their work. A good first reply has four parts:

1. 1The answerIf the customer asked about price, stock or address — answer it. Information that is in the knowledge base does not need to wait for a salesperson.
2. 2One clarifying question«When do you need it?», «Which branch is nearest?» — a question that sets the lead's priority.
3. 3When the next step happensFor a sales lead: «A manager will message you within 10 minutes.» The time promised must match the real SLA.
4. 4Channel choice«Shall we call you, or continue here?» — the customer chooses.

The handoff SLA: from bot to person

An SLA — a service-level agreement — is a simple question here: once the bot hands a lead over, how long until a person joins, and what happens if nobody does?

1. 1Target timeFor a warm lead in working hours, for example 5–15 minutes. The figure comes from the company's real capacity, but it has to be written down.
2. 2OwnerThe lead is assigned at once — in rotation, by channel, by workload or by hand. A lead with no owner is bound to break the SLA.
3. 3A visible notificationIn the channel the salesperson actually watches. In Vexvon a new lead goes to the sales group as a Telegram notification and a salesperson takes it with a «Start» button.
4. 4EscalationIf the target time has passed and nobody has taken it — a second salesperson or a manager. Reminders are checked every minute, so a delay does not stay invisible.

Priority: who goes first

Not every lead needs the same speed. The priority rule is based on what the bot collected.

• High: ready to order or book, «today», a specific product
• Medium: interested, timing «this week», an open question
• Low: «just looking», a distant date, a general question
• Support: already a customer with a problem — goes to support, not the sales queue

Priority can also be built as a scoring system — see chatbot lead scoring. To start with, three levels are enough.

Example: 11:20, three channels

1. 111:20 — three messages at onceInstagram: «price?», WhatsApp: «can I come in today?», website: «how do returns work?».
2. 211:20 — the bot answers all threeIt gives the price and asks about timing; shows free slots for today; states the return rule.
3. 311:21 — priorityThe WhatsApp lead is high (today), Instagram medium, the website one is a support question the bot closes.
4. 411:22 — handoverThe WhatsApp lead goes to the sales group; a salesperson takes it in Telegram and writes at 11:26.
5. 511:35 — InstagramThe customer has answered «this week»; the lead is assigned to the next free salesperson.

Dashboard: what to watch

• Average time to the bot's first reply — should be seconds
• Average and maximum time from handover to the salesperson's first message
• Share of leads that break the SLA — by day and hour
• Message volume by hour — which hours are short of salespeople
• Sale rate for fast-answered versus delayed leads — if tracked in the CRM

An hourly heatmap and the busiest hour are in the analytics report — building the shift rota from that data helps more than writing the SLA.

A rota for peak hours

Peak hours break the SLA most often: lunch, the evening hours when ads are shown most, Monday morning. Three practical steps:

1. 1Measure the flowPull two weeks of message volume by hour. The peak usually falls somewhere other than expected.
2. 2Fit the rota to the flowMove lunch breaks out of peak hours, and make sure at least one salesperson watches only messages during ad hours.
3. 3A separate promise for the peakIf volume really is high, the bot promises a time that matches it — «within 30 minutes» — and that promise is easier to keep.

Common mistakes

• Counting the bot's first reply as «speed» and not measuring the salesperson's time
• Sending the lead to a group without assigning an owner
• Promising a shorter time than the real SLA
• Letting support questions into the sales queue and delaying sales leads
• Writing «sorry for the wait» to delayed leads without a time — the customer wants a concrete time

Limits

The bot answers in seconds, but what speeds up a sale is a person joining on time. If there are not enough salespeople for the peak-hour volume, the bot cannot hide it — it only makes the wait more polite. The bot also sets priority only from what was said in the conversation; if a customer hides their intent, the priority can be wrong, and a salesperson must be able to correct it.

Conclusion: speed is a correct answer plus a person on time

A chatbot cuts first-reply time to seconds, but the value of speed lies in a correct answer, recognised intent and a person joining on time. A written SLA, an owner assigned at once, a visible notification, priority and escalation — with those in place, leads stop waiting during working hours.

How a lead is prepared on WhatsApp in three questions is shown in WhatsApp lead qualification; the rest of the section is in this category. To build your own SLA together, ask for a demo.

Frequently asked questions

1. 1Which businesses does chatbot response speed matter most for?Any field where customers write to several companies at once: real estate, renovation, education, clinics, cars, travel.
2. 2Does the bot replace the salesperson's response time?No. The bot gives the first reply and prepares the lead; the sales conversation depends on a person joining on time.
3. 3What should the SLA be?In line with the company's real capacity, written down and measured. For a warm lead in working hours, many companies aim for 5–15 minutes.