top of page
Writer's pictureFernando Cuenca

Identifying "periods of interest" using "Reference Sets"

"I found some historical lead time data. Should I just keep the most recent? last month or quarter, perhaps?"


There's nothing special about the last month, or the last quarter. Moreover, change doesn't follow the calendar, so month-to-month comparisons may not be that useful. You have to first understand if the process you're observing is going through a period of stability or if it's trending in some direction.


With that understanding, you can segment your data in one or more "reference sets", using significant events in the timeline of your business to separate them. Within a reference set, you might consider keeping only the most recent data, as long as you're sure it still represents the same process.

📊 Reference sets can be compared to each other. Comparing a calendar period of stability with another where the process is trending will be misleading. Within a period of stability you will still see variation, and comparing one month/week/quarter to the previous one doesn't really add more information


3 views0 comments

Recent Posts

See All

Commentaires


bottom of page