Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Metric

Cycle Time

The elapsed time from an accepted request to live work, read stage by stage to show where work waits; a process measure, not a target for individuals.
AnalyticsmetricUpdated

Formula

Date the work went live − Date the request was accepted

Unit: time · lower is better

Definition

Cycle time is the elapsed time for one item of marketing work, such as a campaign, an email, a landing page or a content piece, from an accepted request to live work. The hub on marketing process design defines it in those words and pairs it with rework as the two measures that show whether a process is getting better.

Date the work went live − Date the request was accepted(time)

  • Date the request was accepted is when the request passed triage and the team agreed to do it, not when someone first mentioned it in a meeting or a message. Counting from the first mention measures how long requests sit before intake, which is a different question.
  • Date the work went live is when the campaign launched, the email was sent or the page was published: the end of the route, not the end of the build.
  • The result is a duration, measured the same way each period.

A team reports it across many items, by type of work. With only a handful of items of a type, a range says more than one summary figure, because a single stalled item can drag an average far from the usual case.

The name is shared. In the growth loops framework, cycle time is how long one trip round a loop takes; this page covers only the operations measure, and neither is a sales cycle.

How to interpret it: by stage, waiting versus working

A total tells a team that work is slow. It does not say why. The useful reading splits the elapsed time into the stages of the workflow, such as intake, brief, build, review and launch preparation, and then splits each stage into time spent actively working on the item and time it spent waiting for something: a reviewer, an asset, a decision, a reply from the requester.

That split usually changes the conversation. The process hub observes that most elapsed time is waiting, most often in a review queue or for an asset. Faster work rarely moves the total; removing a wait does, so a team that answers late launches by asking its builders to hurry is working on the smaller part of the problem.

Three readings do most of the work:

  • By stage. Which stage holds items longest, and is it the same stage each month? A queue that keeps forming in front of one stage marks the step that limits the whole route.
  • By type of work. Each request type has its own route and its own normal duration, which is what any promise to requesters has to be built on.
  • Against the amount of work open. Waits lengthen when too much is in flight at once. The guide to limiting work in progress bases each stage's limit on the team's own records and uses the same history to tell a requester roughly how long the queue ahead of their item tends to take.

Where the work system cannot show where items wait, the tutorial on mapping a marketing process follows a small number of recent items by hand, marking each wait and loop back, before anything is changed.

What moves it

  • Waits at review. Reviewers who see work only when it reaches them create the longest queues, and a review column that stays full marks review as the step limiting the route.
  • Incomplete requests. Work accepted without its inputs stops mid-route to wait for them; required fields at intake move that wait to the front.
  • Rework. An item sent back from a gate repeats stages and rejoins their queues.
  • Too much open at once. More items in flight lengthen the wait before the stage that limits the route.
  • Late dependencies. Photography, product information and agency deliverables requested after the brief leave finished builds waiting for their parts.

What good looks like

There is no reliable benchmark for cycle time, and the site does not cite one. Durations depend on the type of work, the reviews it needs, the size of the team and how much it relies on others, so a figure from another organisation describes a different process. Judge it two ways. Against the team's own history: the same type of work, measured from the same start and end points, before and after a change to the route, ideally with the new stage run on one stream of work while the old route continues on another. And against what the plan needs: whether work goes live in time for the dates the calendar and the market set.

Direction and composition matter more than the level. A shorter cycle time with steady rework is a process improving; a shorter cycle time with rising rework or skipped gates is a number being protected.

Illustrative example — not a real result

A lighting retailer finds where its campaigns wait

Nindaard Lighting, an online lighting retailer, exported the stage changes for the 24 campaigns it launched in a half-year. From accepted request to live, the median was 22 working days. Building the emails, pages and ads took a median of 6 days. Of the other 16, 7 were spent in the queue for brand review and 5 waiting for photographs of new fittings. The marketing lead had planned to hire a freelance designer to speed up builds. Instead the team booked two fixed brand review slots a week and sent the photography request on the day each brief was signed, so the shots arrived before the build began. Over the next half-year the median fell to 14 working days, while build time stayed at 6.

Rework Rate is the measure the process hub reads beside cycle time; a cycle time that falls while rework rises usually means gates are being rushed. A SLA turns cycle time by request type into a turnaround requesters can plan around, set from the team's own records. A Request Intake route fixes the moment the clock starts, because a request counts as accepted only once it has passed triage. An Approval Matrix decides which reviews each piece of work passes, and so how many review queues sit on its route.

The marketing operations hub gives each of its disciplines a measure, cycle time and rework being the process's, and the guide to work management and marketing calendars lists cycle time by work type and stage among the reports a work system produces once every stage change is recorded.

Common mistakes

  • Making it a personal target. Gates get bypassed and items are marked done before they are, so the figure improves while launches carry the mistakes the gates existed to catch.
  • Starting the clock at the first mention. Time in inboxes before intake inflates the figure.
  • Reading only the total. Without the stage split, waiting looks like slow work.
  • Blending every type of work. A mix of quick and complex items moves the average whenever the mix changes.
  • Borrowing another team's figure. Turnarounds and limits come from the team's own records.
  • Changing several things at once. Nobody can then say which change moved the measure.

More terms to look up

Where this term is used

Related