Cadence and speed

Going faster isn’t just about pedalling harder

Most businesses spend a not-insignificant chunk of time figuring out how to go faster: to do more in less time. That often gets framed as an issue of speed: how fast can you do things? But there’s more to productivity than just raw speed: cadence is important too.

Cadence literally means a rhythm or a beat. It’s used most obviously in music, but in all sorts of other fields too. Cyclists use it, for example, to refer to the rate of their pedalling: if they rotate the pedals 100 times each minute, then their cadence is 100 rpm. And cadence is used in business, too.

When developing the iPhone, for example, Apple operates on a yearly cadence. Ever since the iPhone 4S was released in October 2011, Apple have held an event each and every autumn in which they’ve released the latest iPhone model. This cadence affects the world outside of Apple (people plan their upgrades by it), but it mostly manifests internally. New features and hardware are planned years in advance, and the cadence gives urgency and predictability to planning.

The fashion world likewise operates on its own cadence, but six-monthly rather than yearly: everything revolves around the launch of spring-summer collections (roughly in the September of the previous year), and autumn-winter collections (roughly in February). Again, the predictable cadence sets the tone for everything, dictating when and how – and if – a particular product will launch.

Every business operates according to its own cadence, and how quick that cadence is is a huge factor in how productive businesses are.

That’s particularly true in knowledge work, where cadence often determines how long work takes. If you run projects with two-weekly check-ins, for example, it’s quite likely that you’ll find that Parkinson’s law applies: “work expands so as to fill the time available for its completion”. Each phase will, unsurprisingly, end up taking two weeks to complete.

But cadence isn’t purely synonymous with speed and productivity. For a start, it’s not possible to increase cadence beyond a certain point. If you switch your check-ins from fortnightly to weekly and then down to daily, where do you go from there? If you simply dictate that you’re going to do a project every week instead of a project every two weeks, will you be able to deliver that? If Apple tried to launch a new iPhone every six months rather than every year, would that be possible? At some point, you hit both limits of the physical reality of the underlying work, and limits of what the external market can sustain.

If cadence alone isn’t enough, how else do we increase speed and productivity? Manufacturing has had to grapple with these issues for far longer than knowledge businesses have, and gives us the terminology we need to have a nuanced view of speed. There are three concepts that are important:

Lead time
The time difference between the project being ordered/signed off and the project being delivered. If a customer orders something from you on 1 January and you deliver it on 10 January, the lead time is 9 days.
Cycle time
The time difference between starting the project and finishing it. If a customer orders something from you on 1 January, you start it on 5 January, and you deliver it on 10 January, the lead time is 9 days and the cycle time is 5 days.
Throughput time
The number of projects you complete in a specific period of time, or the number of phases if you split projects up into phases. If you complete 10 projects every 30 days, your throughput time is 3 days.

These three times are, of course, highly interrelated. If you find a way to reduce your throughput time, you can decrease your cycle time; improving both will likely mean removing bottlenecks, freeing up resource, and reducing overall lead time. But by splitting them out, you gain important nuance, and the ability to understand where exactly your problems are coming from.

A large difference between lead time and cycle time? That suggests that you have problems getting projects in and started, perhaps because you have too little resource or a process that’s not parallelised enough. A cycle time that’s greater than the sum of the throughput times of your phases? That suggests you have coordination issues, with delays happening between the stages of projects.

These three measures are vital, but they’re somewhat artificial; they’re divorced from external reality and from the business’s customers. How do you know whether your levels of speed and your cadence are good enough? To understand that, there’s one other measure that lean manufacturing gives us:

Takt time
A reactive measure of how fast you must work in order to deliver what’s demanded of you. If you have 50 projects to deliver in a month, and 10 staff who each work 22 days a month, then you have 220 person-days / 50 projects = 4.4 person-days to complete each project; that’s your takt time.

Takt time (from the German Taktzeit, or “interval time”) is a supremely useful measure. First, it allows you to spot if there are going to be production problems in advance. If your takt time is shorter than your average throughput time, then you know that you need either to find a way of reducing your throughput time, or to take on more resource.

And second, it allows you to set a regular cadence. If you know you need to complete one project every 4.5 days, then you need to build the cadence of the business – its regular meetings, its deadlines, its check-ins, its reporting – to that rhythm. Everything needs to build to and from it.

The lessons of manufacturing are that there are two ways to improve speed and productivity, each different but highly complementary:

  1. Improve your absolute speed, by concentrating on throughput time: find ways to do what you do in less time

  2. Getting your cadence right, by concentrating on takt time: flushing out any inefficient “bunching-up” of work, eliminating wasted time, and making sure you have the right amount of resource to meet demand

One is about speed, the other is about rhythm; to go fast, you’ve got to have both.