Measuring how long things actually take: cycle time for small business

Why I started timing my own process

A few years into running things solo, I noticed something uncomfortable: I had no idea how long anything actually took. I knew roughly how many orders I filled or tickets I closed each week, but if you’d asked me “how long does it take from the moment someone orders to the moment it ships,” I would have guessed. And my guess would have been wrong, because I was remembering the fast ones and forgetting the slow ones.

That’s the gap cycle time is built to close. It’s not a fancy metric. It’s just an honest clock on a process, and for a one-person business it’s one of the few numbers that tells you where your own time and attention are actually going.

What cycle time actually measures

Cycle time is the elapsed time between the start of active work on something and its completion. Order placed to order shipped. Support ticket opened to support ticket resolved. Draft started to article published. You pick the start point and the end point, and you measure the gap for each individual item, not just the average across your whole business.

That last part matters. A lot of small business owners think they’re tracking cycle time when they’re really just tracking throughput, like “we shipped 40 orders this week.” Throughput tells you volume. Cycle time tells you speed per item, which is a different question and often a more useful one, because it’s the number your customer actually experiences.

It helps to separate cycle time from a close cousin: lead time. Lead time is the full clock from when a customer makes a request to when they get the result, including any time the request just sat in a queue before you touched it. Cycle time is narrower: it’s the time from when you actually started working on it to when it was done. If someone emails you Monday and you don’t open the email until Thursday, that three-day gap is lead time, not cycle time. Both are worth knowing, but they point to different problems. A long lead time with a short cycle time usually means things are piling up before you get to them. A long cycle time means the work itself is slow once you’re in it.

A concrete example from a small operation

Say you run a small print-on-demand shop by yourself. You could define cycle time as the gap between “order confirmed” and “tracking number generated.” If you log the timestamp for both events, even in a plain spreadsheet, you can calculate that gap for every order.

Once you have even a couple weeks of that data, patterns tend to show up that a gut feeling would have missed. Maybe orders placed on weekends sit longer because you’re not checking the queue. Maybe a specific product takes noticeably longer because the supplier for that one item is slower. You don’t need a dashboard to see this. You need the two timestamps and a sorted column.

The same idea applies past order fulfillment. A freelancer could track cycle time from “client sends brief” to “first draft delivered.” A local service business could track “inquiry received” to “quote sent.” A content creator could track “topic chosen” to “video published.” Anywhere you have a repeatable start and a repeatable end, you can measure the gap.

Median, not average

Here’s a mistake I made early on: I looked at the average cycle time and thought it told me something it didn’t. Averages get pulled around by outliers. If nine orders ship in a day and one gets stuck for two weeks because a supplier ran out of stock, the average will look worse than what most customers actually experienced, and it will hide the fact that you have one specific, fixable problem rather than a general speed issue.

The median, the middle value when you sort all your cycle times low to high, is usually more representative of what a typical customer goes through. Alongside the median, it’s worth looking at the slowest 10 or 20 percent of cases separately. Those outliers are often where the real story is. A distribution that’s mostly fast with a long tail tells you to go find out what’s different about the tail cases, not to panic about your average.

What cycle time can and can’t tell you

Cycle time tells you how long something took. It does not tell you why. If your cycle time for order fulfillment is a day longer than it used to be, the number itself won’t say whether that’s because a supplier changed, because you added a manual step, because you got busier, or because you changed how you’re logging the start time. You have to go look.

It also doesn’t tell you anything about quality. A fast cycle time on a support ticket isn’t automatically a win if the ticket got closed without actually resolving the customer’s problem, and you’ll see that show up later as a repeat ticket or a churn signal, not in the cycle time number itself. Speed and quality are separate questions, and treating a shrinking cycle time as proof that things are getting better, on its own, is a mistake I’d rather flag than let you walk into.

And it’s sensitive to how you define the start and end points. If you’re not consistent about what counts as “started” and what counts as “done,” you’re not comparing cycle times over time, you’re comparing definitions. Write your definition down once, in plain language, and don’t change it without noting when and why, or your trend line will lie to you without meaning to.

How to actually track it without new software

You don’t need project management software to start. Two timestamps per item and a spreadsheet formula are enough to get a real number. For each item, log the date and time it started and the date and time it finished. Subtract one from the other. Sort the results. Look at the median and the slowest handful.

If you’re already using a tool that timestamps things for you, an order system, a help desk inbox, a CRM, that data is probably sitting there unused. Most of these systems record a created date and a completed or closed date somewhere, even if there’s no cycle time report built in. Exporting that data and doing the subtraction yourself is often faster than trying to configure a built-in report to show you what you actually want.

The habit that matters more than the tool is doing this on a rhythm, weekly or every two weeks, rather than as a one-time audit. A single snapshot tells you where you stand. A repeated one tells you whether a change you made, a new process step, a different supplier, a decision to batch a task instead of doing it as it arrives, actually moved the number, or whether it was noise.

Using it without overreacting to it

Once you have a cycle time number, the temptation is to treat every wiggle as a signal. Resist that. A handful of orders is a small sample, and normal week to week variation will look like a trend if you’re staring at it hard enough. Give yourself enough data points, at least a couple weeks worth in most small operations, before you decide something changed.

When you do see a real, sustained shift, treat the number as a starting point for a question, not an answer. “Why did this get slower” is worth ten minutes of actually looking at the outlier cases by hand. The number tells you where to look. It doesn’t do the looking for you, and it’s not a substitute for knowing your own business well enough to recognize when something’s off before the spreadsheet catches up.

Measuring cycle time won’t make your business faster by itself. What it does is replace a guess with a fact, so that when you decide to change something about how you work, you can actually tell afterward whether it helped.

If you want more of this kind of plain-English breakdown of the numbers that actually run a small business, you can find the rest of what we cover here.

Get new guides and videos first — join the Telegram channel.