Fixing your speed to lead: measuring it honestly, then actually changing it

Speed to lead is the time between an enquiry arriving and the enquirer getting a reply they can act on. Most businesses quote a number that is wrong in the same two ways, fix the wrong thing, and conclude the idea is overrated. This guide is how to measure it honestly and what actually moves it.

The number you are probably quoting is wrong

Two mistakes, and nearly everybody makes at least one.

You are measuring from when somebody opened it, not from when it arrived. An enquiry that landed at 7pm and was opened at 9:05am was not answered in five minutes. It was answered in fourteen hours. The clock starts when the customer hit send, because that is when their clock started.

You are counting an acknowledgement as a reply. "Thanks for getting in touch, someone will be with you shortly" is not an answer, and a customer comparing three businesses does not count it as one. Measure to the first reply that contains information they asked for: a price, an availability, a yes or no.

There is a third, subtler one: you are averaging. The average hides the tail, and the tail is where the losses are. Twenty enquiries answered in four minutes and four answered the next morning average out to something respectable and lose you four customers. Use the median and the 90th percentile, or just count how many took longer than an hour.

Measure it before you change anything

You need three timestamps per enquiry: when it arrived, when a person first opened it, and when the first substantive reply went out. If your tools cannot give you those, that is itself the finding.

Do it by hand for two weeks if you have to. Twenty rows in a spreadsheet will tell you more than a dashboard built on the wrong definition.

Then split the result three ways, because the averages conceal which problem you have:

• By hour of day. Almost every business has one number during opening hours and a completely different one outside them.
• By channel. Phone, web form, text, social. They usually differ by an order of magnitude, and the worst one is rarely the one anybody is watching.
• By who was on. Not to performance-manage anybody. To find out whether the problem is a process or a rota.

What the split usually shows

Three patterns, and the fix is different for each.

Fine in hours, terrible out of hours. The commonest. Your team is not slow; they are asleep. Nothing about training or process will move this number, and most of the effort spent on it is wasted for that reason.

Fine on the channel you watch, terrible on the ones you do not. Somebody watches the phone. The web form goes to an inbox two people can see and neither owns. Instagram messages are read on somebody's personal phone when they remember.

Bad everywhere, in hours too. This is a capacity or ownership problem and it is the only one of the three where more staff is the honest answer. Automation applied here makes a busy team busier.

The five changes, in the order they pay off

1. Give every channel an owner. Free, and usually the biggest single improvement. Most slow enquiries are not slow because everyone was busy; they are slow because everyone assumed somebody else had it. One named person per channel per shift, written down.

2. Answer the out-of-hours ones automatically, from facts you approved. This is where automation earns its keep, because there is no human alternative at 9pm. The constraint is that it answers from information somebody wrote down, and stops when it does not know. An automated reply that guesses converts a slow answer into a wrong one, which is a worse trade than it sounds.

3. Make the first reply useful. If the honest answer is "we need to know X before we can quote", ask for X in the first reply. Most businesses spend two messages establishing contact before the message that moves anything.

4. Route by what the enquiry says, not by who is free. An emergency and a price question do not deserve the same queue. This is worth doing after the first three, not before: routing rules over an unowned channel just sort the backlog.

5. Then look at your rota. Only once you know how much of the tail is genuinely outside hours. This is the expensive fix and it should be last.

What to expect, honestly

Reply time is one input into whether somebody buys, and the research everyone quotes about responding within five minutes is mostly drawn from industries that are not yours. Treat it as directional rather than as a target handed down from physics.

What is reliably true: an enquiry answered while the customer is still thinking about the problem converts better than the same enquiry answered tomorrow, and the businesses that improve fastest are the ones that fixed ownership before they bought anything.

What is reliably false: that dropping from four hours to three minutes multiplies your conversion by the ratio. The first hour matters far more than the first minute, and the gap between overnight and morning matters more than either.

How to know it worked

Re-measure the same three timestamps, the same three splits, the same median and 90th percentile. Compare the count of enquiries that took more than an hour, which is the number that moves visibly and the one your team will feel.

Then check the thing that actually pays: of the enquiries answered inside an hour, how many booked, against those that were not. If that gap is not there after a quarter, speed was not your constraint, and it is worth knowing that rather than buying more of it.

Lead management is where we do the routing and the scoring, and Automations is where the out-of-hours reply comes from.