In short: an email response audit measures how long a business takes to give a real human reply to an enquiry from someone it does not know. You run one by sending a genuine enquiry to your own public address, unannounced, and timing from arrival to a useful answer. An automated acknowledgement does not stop the clock.
Most businesses have a number in their head for this. It is usually wrong, and it is wrong in a predictable direction, because the enquiries you remember answering are the ones you answered. The ones that sat until Tuesday do not come to mind, because nothing about them was memorable. They arrived, they waited, and by the time anyone opened them the customer had gone quiet.
An audit replaces the number in your head with one you did not choose.
What the audit measures, and what it does not
The clock starts when the enquiry lands on your public address. It stops when a person sends something that answers part of what was asked.
Both ends of that need to be strict, because both are easy to fudge.
The start is not when someone opened the email. It is when it arrived. A customer waiting for a reply does not care that the inbox was not checked until eleven.
The stop is the part that catches people out. It is not the autoresponder. It is not the internal notification that pinged someone’s phone. It is not a message saying “thanks, we’ll be in touch shortly”, because that answers nothing and sets no date. It is the first message containing something the customer can act on: a price, a time, a question that moves the job forward, or a clear statement of when they will have the answer.
Acknowledgements are worth sending. They stop an enquiry feeling ignored, and they cost nothing to automate. Counting one as a reply is how a business ends up with an impressive response time and no more work than it had before.
Two methods, and they answer different questions
There are two honest ways to run this, and it is worth knowing which question you are asking before you pick one.
The live drill. One enquiry, sent unannounced, timed. This tells you what actually happens to a stranger who contacts you today. It is a single sample, so it is a snapshot rather than an average, and it will catch a bad day as readily as a good one. That is not a flaw. A customer who contacts you on your bad day gets your bad day.
The historical pull. Twenty or thirty real enquiries from the last month, timed from your own mailbox. This gives you a median and a worst case. It is more representative and it takes longer, and it can only measure enquiries you actually replied to, which leaves out the ones you never answered at all.
Run the drill if you want to know whether the process works. Run the historical pull if you want to know how well.
Method one: run the drill yourself
The mechanics take ten minutes.
- Use an address nobody recognises. Not your personal account, not a colleague’s work account. A free account made for the purpose, with a plausible name on it.
- Ask a real question. Something a genuine customer in your area would ask, specific enough to need an answer rather than a brochure. Vague enquiries get vague replies and you will have measured the wrong thing.
- Send it through the public route. The contact form on the website, or the address printed on the van. Not the address you give to people you already work with.
- Send at a normal time. Mid-morning on a weekday is the fair test. Sending at ten at night measures something else, and if you want to know that, run it as a second test rather than the first.
- Tell nobody. This is the part that fails most often.
- Record two timestamps. Arrival, and the first human answer. Note anything that came in between, including the autoresponder, but do not count it.
The awkward part: you know it is coming
If you send the enquiry yourself, you have compromised the test before it starts.
You will check the inbox more often that morning. If you are the person who answers enquiries, you will answer this one faster than you would answer a real one, and you will not notice yourself doing it. Even where a team handles enquiries and you do not, the knowledge tends to leak: you glance at the inbox mid-morning, you see it sitting there, and you mention it.
There are two ways round this.
Have someone outside the business send it, and ask them not to tell you when. You lose control of the timing, which is the point, and you get a result you did not influence.
Or have it done for you. That is why we run these free: the value of the number depends entirely on the recipient not knowing it is a test, and that is a condition you cannot meet on your own business.
Method two: measure from your own history
If you would rather work from real enquiries than a staged one, the data is already in your mailbox.
Search the enquiry address for the last month and pull the threads that started with somebody you had not dealt with before. Twenty is enough. For each one, note the arrival time of the first message and the send time of your first substantive reply.
Two decisions will change your answer, so make them deliberately and write them down.
Wall clock or working hours. An enquiry arriving at six on Friday evening and answered at nine on Monday morning is either 63 hours or about one working hour, depending on which you count. Both are true. The wall clock is what the customer experienced, so it is the number that predicts whether they waited for you. Working hours is what your team controls, so it is the number to manage against. Measure the wall clock first, because that is the one deciding whether you won the job.
Median, not mean. A handful of enquiries that sat for a week will drag an average somewhere unrecognisable. If nineteen enquiries were answered inside two hours and one took nine days, the average is around eleven hours and describes nothing that happened. The median is the middle value, and it tells you what a typical customer experienced. Note the worst case separately, because that one is a real customer too.
Reading the result
A number on its own is not much use, so it needs something to sit against.
The largest published audit of the question sent a real enquiry to 2,241 companies and timed what came back: 37% replied within the hour, 16% later the same day, 24% took more than a day, and 23% never replied at all. We go through what that research does and does not prove, and what a realistic target looks like when you are the one doing the work, in how fast you should reply to an enquiry.
Against that, the rough reading is this. Inside an hour during working time puts you ahead of most of the field. Same working day is respectable and is a reasonable target for a business where the person answering is also doing the work. More than a day and you are relying on the customer not having contacted anyone else. Never is more common than most people assume, and it is worth checking that your form actually delivers before assuming it cannot be you.
What to do with a number you do not like
Two moves, and they are separate problems.
Make the acknowledgement instant. Every enquiry gets a reply within seconds confirming it arrived, naming a real person, and saying when a proper answer will come. This is mechanical, it never has a busy Tuesday, and it stops the enquiry being at risk while you finish the job you are on. It is not the reply.
Make the reply happen the same day. This is the part that needs a person, and it is the part that wins the work. What goes in it matters as much as when it goes, and that is a separate question – we cover the four things a first reply has to do in what to say in a first reply to an enquiry.
If the reply is the thing that keeps slipping, the honest diagnosis is usually capacity rather than discipline. Nobody forgets to answer customers because they do not care. That gap is the half you can hand over: our lead follow-up service runs the acknowledgement, the reply and the chase as one system.
The version we run
We send one enquiry to your own enquiry address, time it from arrival to first human reply, and email you the result about a week later.
You write the enquiry yourself and approve the exact wording before anything is sent, so nothing reaches your team that you have not seen. It goes from a named person at a real address that accepts replies. There is no call attached to it and no obligation afterwards.
The reason it is free is that it is the most honest sales argument we have. Every other page on this site tells you that slow follow-up loses work. This one measures whether it is happening to you, and sometimes the answer is that it is not.
Have us run it on your business – there is a decent argument that you should not run this one on yourself, because you will know it is coming.
LoadOff handles lead follow-up and daily inbox work for UK small businesses. The acknowledgement is automatic; the reply is written by a person. Most clients are live within 48–72 hours. Get in touch if you want to talk it through.
Related reading: How fast should you reply to an enquiry? · Where enquiries get buried