How long your clients are really waiting for a reply

The clock starts earlier than you think

The most common mistake in measuring response time is starting the clock the moment someone at the company sees the inquiry. That's convenient, because from that point everything usually moves fast, but it's not the time the client actually experiences.

For the client, the clock starts the second they submitted the form, sent the message, or heard the ringtone. Everything that happens between that moment and your first contact is waiting time, regardless of whether anyone was watching the inbox.

This difference isn't theoretical. In most service businesses, it's the entire problem: handling the matter itself takes five minutes, but the inquiry sat there for sixteen hours beforehand.

Set one rule and stick to it throughout the measurement: the clock starts when the inquiry reached you, not when you noticed it.

What actually counts as a response

The second mistake is subtler, and tool vendors are happy to keep it alive. An automatic message saying 'thanks, we'll be in touch soon' is not a response. It's a receipt.

We call it a response only when the first contact actually moves things forward: a conversation, a concrete question about the details, a proposed time slot. If, after that contact, the client knows no more than before it, the clock is still running.

This distinction matters in practice, because an autoresponder is cheap and tempting to treat as closing the loop. It's worth having, but you need to know what it actually does: it buys you some patience, it doesn't shorten your response time.

Why the average lies

Say you had ten inquiries in a week. Nine handled within fifteen minutes, one forgotten for four days. The average comes out to around ten hours, and it describes neither the nine nor the one.

That's why, for response time, you should look at two numbers instead of one. The median, the middle value, tells you what the typical case looks like. The worst ten percent tells you what your failure looks like, and those are exactly the cases people tell their friends about.

If the median is good but the tail is long, you don't have a speed problem. You have a problem with inquiries falling through the cracks, and that requires a completely different fix.

Measuring thirty inquiries

You don't need any system for this. You need a spreadsheet and half an hour.

• Take your last thirty inquiries

  From every channel at once: phone calls, the form, messages. Thirty is enough to see a pattern, and few enough to do in one sitting.

• Record two timestamps

  When the inquiry arrived, and when the first contact that moved things forward happened. Hours, not just days, because at daily resolution the whole measurement loses its point.

• Calculate the difference and sort it

  The middle value is your median. The three worst cases out of thirty are roughly your worst ten percent.

• Decide what to do with nights and weekends

  You can count elapsed calendar time or only business hours. Both are fair, as long as you pick one and note it alongside the result. If you only count business hours, remember the client is still waiting in calendar time regardless.

Do this measurement yourself, or with one other person, but don't announce it to the team beforehand. An announced measurement changes behavior, and ends up measuring the one week everyone tried extra hard.

Three places where time gets lost

If the result surprised you, the cause almost always lies in one of three places.

• A channel nobody watches

  A message on a social media profile, a form landing in a shared inbox, a text to a number nobody actually uses. Inquiries from channels like these usually have the worst times in the whole set.

• Blurred ownership

  Three people saw the inquiry, and each assumed one of the others would handle it. This is the most common cause of individual, very long times, the tail that ruins the average.

• One attempt and silence

  You called back once, the client didn't pick up, and the case got closed in your head, but not in reality. Decide how many attempts you make and at what intervals, because without that rule everything depends on whoever happened to be handling it.

Three decisions that follow from the measurement

The result itself isn't the goal. The point is to use it to make one of three decisions, and which one depends on exactly where the problem lies.

• If the median is good but the tail is long, fix the process

  No tools needed here. You need one person responsible for each inquiry, a set number of contact attempts, and one place where every case is visible. This is a change that costs nothing.

• If the median is bad during business hours, it's a capacity problem

  The team can't keep up because they're out in the field or with a client. In that case, the real options are an extra person or a process that contacts the interested person after the form and books a time slot. Calculate both before you choose.

• If it's only bad outside business hours, that's a separate issue

  Don't try to fix this by changing your daytime process, because nothing there will change. That's a separate decision, and we cover it in its own guide.

If the measurement gives you a median in minutes and no long tail, that means response time isn't your problem. Close the topic and revisit it the next time your volume of inquiries grows.