Speed to lead: why the first hour decides the deal, and how a small team answers inside it

Speed to lead is the time between a prospect sending an enquiry and receiving a reply that moves the conversation forward. It is one of the few sales metrics with strong research behind it: in data from 1.25 million leads, firms that tried to contact a prospect within an hour were nearly seven times as likely to qualify that lead as firms that waited even an hour longer, and more than sixty times as likely as firms that waited a day or more.

The same researchers audited how fast companies actually answer, and the result was not flattering. Of 2,241 US companies that received a test enquiry through their website, 37% replied within an hour and 23% never replied at all. Among those that did reply within a month, the average wait was 42 hours. Both findings come from "The Short Life of Online Sales Leads" by James Oldroyd, Kristina McElheran and David Elkington, published in Harvard Business Review in 2011.

The study is more than a decade old and buyer habits have shifted toward messengers, but the mechanism has not changed: the moment someone fills in a form or writes to you is the moment they are thinking about the problem. An hour later they are in a meeting, comparing alternatives, or have already booked a call with whoever answered first.

Why most enquiries wait too long

When a team learns its real response time, the first instinct is to ask people to try harder. That rarely works, because the delay is usually structural. Four causes come up again and again.

Nobody owns the enquiry

A form notification goes to a shared mailbox or a group chat. Five people see it, each assumes someone else has it, and it waits. An enquiry with no named owner is an enquiry nobody is responsible for answering.

The queue is sorted by newest

Most inboxes put the latest message on top. That is exactly wrong for response time: the person who wrote three hours ago sinks below the one who wrote three minutes ago, and the longest wait keeps getting longer.

Channels are split across tools

Website forms land in email, Instagram messages on someone's phone, WhatsApp on another phone, chat in a widget dashboard. Each channel has its own response time and nobody sees the total. The fastest channel hides the slowest.

The salesperson has to research before replying

If answering means opening the CRM in another tab to check whether this person is already a customer, has an open deal or talked to a colleague last month, every reply costs several minutes of context-gathering, and the reply waits for a quiet moment that never comes.

The set-up that gets a reply out in minutes

Fast response is a system property. These five pieces, in this order, cover most of the gap for a small or mid-sized team.

1. One queue for every channel. Forms, email, website chat and messengers land in the same list, so the team works one queue with one response time. One queue for every channel explains what that queue has to do.
2. An owner on arrival. Every new enquiry is assigned the moment it arrives, by a rule (round-robin, territory, product, language) or to a named person on duty. An "Unassigned" view exists for exceptions, and someone checks it.
3. The queue sorted by debt. Whoever you still owe a reply sits on top, longest wait first. New messages do not jump the queue.
4. Context beside the reply. The person answering sees the customer's record, open deals and history in the same screen, so there is no detour to another tab.
5. An escalation for silence. If an enquiry has had no reply after a set time, the owner gets a reminder; later, their manager does. Automation enforces the target that a policy only states.

None of this needs a large team. A small team with one shared queue sorted by wait time can out-respond a much larger team working from personal inboxes.

What to automate, and what to keep human

Automation speeds up the parts of a first response that do not need judgement. It should not pretend to be the salesperson.

What a useful first reply contains

Speed only counts if the reply moves the conversation. A good first reply does three things: it answers the question that was actually asked, or says plainly when the answer will come; it uses what the customer already told you instead of asking them to repeat it; and it offers one clear next step, usually a call at a specific time or a booking link with real free slots. "Thanks for reaching out, when are you free for a chat?" is fast and useless. "The kitchen you described usually takes four to six weeks; I can walk you through options on Thursday at 11:00 or Friday at 15:00" is fast and starts a deal.

A bot that answers instantly and then traps the customer in a loop is slower than a person who answers in ten minutes. If you use one, test it before it meets customers: before a support bot talks to customers covers knowledge, boundaries and the hand-over.

How to measure speed to lead properly

Three numbers are enough. Report them weekly, split by channel and by owner.

• Median first-response time. The median, because one weekend enquiry answered on Monday should not hide that weekday replies take eight minutes. Count working hours separately from all hours if your team does not cover evenings.
• Share answered within the target. For example, the percentage of enquiries with a first human reply within 15 minutes during working hours. A target makes the metric actionable.
• Enquiries never answered. A raw count, reviewed one by one. In the research above this was nearly a quarter of companies; in a single company it is often a handful of leads a month, each worth a deal.

Measure from the moment the enquiry arrived to the first reply that moves things forward. An automatic "thanks, we received your message" is useful, but it is not the response.

How Nextora handles it

In Nextora the Inbox is one queue for the whole company: Telegram, WhatsApp, Instagram, SMS, email to your work address and chats with the website bot all land in the same list. The tabs slice the queue by owner (yours, unassigned and the rest), and the Channel chip narrows it to a single source when you need to.

By default the list is ordered by debt, not by time. Everyone you still owe an answer sits on top, longest wait first, and those rows read "waiting 2h" instead of a timestamp. A second ordering, Biggest deal first, lifts the largest open-deal total to the top without overriding the debt rule: someone you owe an answer still outranks someone you do not. Replying happens next to the customer's record, with the deal and history in a side panel.

When the website bot is answering, it carries on while it can. If the customer asks for a human or the conversation stalls, the row is marked "Waiting" and a short recap, the handoff summary, appears above the thread, so the person taking over does not have to read the whole history.

Ownership and escalation are rules. A board rule on the leads board reads like a sentence, for example "a new lead from the website: notify Anna, an hour later give her a call task", and can be written in plain words for the Workshop to build. Forms can include a meeting-booking question that shows real free time, so an enquiry can become a booked call without anyone replying at all. The Conversations page shows the Inbox, and Automation covers rules and workflows.

A two-week plan

1. Day 1: send a test enquiry through every channel you advertise and time the replies. Write the numbers down.
2. Days 2–3: route every channel into one queue and give every new enquiry an owner on arrival.
3. Days 4–5: switch the queue to longest-wait-first and agree a first-response target.
4. Week 2: add an escalation rule for silence and an out-of-hours answer path, then repeat the test enquiries.

Fast first replies only pay off if the enquiry lands somewhere useful. If your forms send notifications but leave the CRM empty, start with the web form that looks like it works, and once enquiries become deals, keep them moving with stages that have exit criteria.