I once watched a company spend eleven weeks and a substantial amount of money rewriting their sales messaging, while their median response time to an inbound enquiry sat at nineteen hours. The new messaging was genuinely better. It made almost no difference, because by the time anyone read it the reader had hired somebody else.
Response time is the least glamorous number in sales and reliably the highest leverage one in a small operation. Here is how to find yours, why it is worse than you believe, and the four fixes that take an afternoon rather than a quarter.
Interest is perishable
The uncomfortable premise underneath all of this: a lead is not a stable object. It is a person in a temporary state.
Somebody fills in your form or answers your email because right now, in this ten minute window, the problem is on their mind. They are at their desk. The tab is open. They have probably just contacted two of your competitors as well, because that is what people do.
An hour later they are in a meeting. The next morning the problem has moved down the list, or someone else has already answered and the decision is effectively made. The message you send at hour twenty is not a late version of the message you would have sent at minute five. It is a different message arriving to a different person.
That is the whole mechanism. Everything below follows from it.
Your number is worse than you think
Ask a small team about their response time and you will hear "pretty quick, usually within the hour". Then measure it and the median is six hours.
The gap is not dishonesty. It is that people remember the fast ones. The enquiry you answered in ninety seconds is a story you tell. The one that sat in a shared inbox over a weekend because everyone assumed a colleague had it is not a memory at all, and there are more of those than you think.
Measure it like this, and do not automate the measurement until you have done it once by hand:
- Take the last thirty enquiries from every channel. Every channel. The form, the general inbox, the personal inboxes, WhatsApp, the phone, and replies to your outbound.
- For each, write down when it arrived and when a human genuinely responded. Not the autoresponder.
- Sort the gaps and take the middle one.
Use the median, not the average. One enquiry that took nine days because it landed during a holiday will drag an average into fiction. The median tells you what a typical person experienced.
Do it now, before reading the rest. It takes twenty minutes and it is the only part of this article that is actually about you.
The delay is a queue problem, not a people problem
Here is what I find nearly every time, and it is never that anybody is slacking.
- The enquiry arrives in more than one place. Website form to one address, WhatsApp to a phone in someone's pocket, replies to outbound in a personal inbox, the occasional call to a mobile. There is no single queue, so there is no way to see what is waiting.
- Nobody owns the first response. Two people can see the shared inbox, which reliably produces zero people answering, because each assumes the other has. A queue with two owners has no owner.
- Nothing announces itself. The only way anyone learns a lead arrived is by going and looking, and going and looking is the first thing that stops happening on a busy day.
- Answering properly takes twenty minutes, so it never happens between jobs. The reply waits for a clear hour, and a clear hour is a rumour.
Every one of those is structural. You cannot fix any of them by asking people to try harder, which is what most teams attempt first.
The four fixes
1. One queue. Every enquiry, every channel, one place. This is the fix that makes the other three possible, and it is the reason we built the unified inbox so that email, WhatsApp and outreach replies thread onto the same lead record. If a message can arrive in four places, someone will always be able to say honestly that they did not see it.
2. One named owner per shift. Not a rota "in principle". A person, today, whose job is the first response. It can rotate weekly. It cannot be "whoever is free".
3. An alert that reaches a pocket. A notification on a device the owner is carrying. If checking is a decision, checking will lose to the job in front of them.
4. Separate acknowledging from answering. This is the one that changes the number fastest. A two line reply within five minutes that says you have seen it, you understand what they are asking, and you will come back properly by four o'clock, does almost all the work of the real answer. It stops the search. It converts you from "a company that has not replied" into "a company that is dealing with it".
Then hit four o'clock. A promised time that slips is worse than no promise.
What to do about outbound reply latency
The same principle, pointed the other way. Someone who answers a cold email is at maximum interest in the minutes after they hit send. They have just spent attention on you voluntarily, which is rare.
Three habits that matter more than any of the clever ones:
- Check for replies at fixed times, several times a day, rather than continuously. Continuously means never, because it depends on mood.
- Have the follow-up half written. Most replies fall into four or five shapes: interested, wrong person, not now, how much, and who are you. If those five responses already exist as drafts you edit rather than blank pages you face, your latency collapses.
- Stop the sequence on reply, automatically. Nothing undoes a good conversation faster than the third automated follow-up arriving after someone has already answered you personally. If you are running sequences by hand, this is the single most dangerous manual step in the whole operation.
An AI draft waiting for approval solves the middle one honestly. It is not writing the reply for you; it is removing the blank page, which is the actual delay. That is what the reply drafts in our AI sales agents are for, and why they wait for a human to press send.
A target worth writing down
Pick one and put it somewhere visible.
| Situation | Target | Realistic if you do the four fixes |
|---|---|---|
| Inbound enquiry, business hours | Acknowledge in 5 minutes | Yes |
| Inbound enquiry, out of hours | Acknowledge next opening hour | Yes |
| Reply to your cold outreach | Answer within 1 business hour | Yes |
| Full answer requiring a quote | Within the window you promised | Only if you promise honestly |
The targets are not the point. Having a number is the point. Teams with a number hit it. Teams intending to be quick discover, six months later, that their median is nineteen hours and their new messaging never had a chance.
If you want the wider frame on why leads go quiet, sales pipeline stages that match how deals actually close covers what happens after the first response, and the follow-up sequence covers how many times to try before you let go.