Order cycle time is the elapsed time from an order being placed to it leaving your premises. Measuring the average is close to useless because the average is dominated by the routine orders that were never a problem. The number worth watching is the tail: the slowest five or ten per cent, which is where complaints and support load actually come from.
In more detail
This is the metric an internal SLA would be set against, and the reason to measure the distribution rather than the mean is that customers experience individual orders, not averages.
Breaking the total into segments is what makes it actionable. Paid to picked, picked to packed, packed to collected. A store with a slow overall figure usually has one slow segment, and the fix for each is different: staffing for the first, layout for the second, carrier scheduling for the third.
Measure in working hours, not calendar hours, or every Monday will look like a failure for a store that does not ship at weekends.
Where people get it wrong
- Reporting the average. It hides exactly the orders that cause problems.
- Measuring in calendar hours when you do not operate at weekends.
- Measuring only to label creation. A label is not a despatch, and the gap between them is where lost orders hide.
- Not segmenting. A single number tells you that you are slow, not where.
Common questions
How is order cycle time measured?
From order placed to order despatched, in working hours, ideally broken into paid to picked, picked to packed, and packed to collected.
Should I track the average or the worst cases?
The tail. The slowest five to ten per cent is where complaints and support load come from, and the average is dominated by orders that were never a problem.
Is label creation the same as despatch?
No, and treating them as equivalent hides a real failure mode. A label can be generated for a parcel that never physically leaves, which is where a surprising number of lost orders originate.