Kanban Metrics that Work23 Oct 2018

Kanban, Scrum, Scrumban - however you're managing your team's workflow, you're going to want to know how well the method works, or how far off the goal it still is. For that, you should be looking at the process metrics.
Some of the most common metrics measure: throughput - work completed over time, speed of response to
a request, tasks started vs completed count and team satisfaction - both measuring process sustainability, or cycle time efficiency - working time alone in relation to waiting & working time.

The most accurate way to achieve a metric relevant to your process would be you designing it yourself, around the desired outcome. So you'd look at what kind of information tells you whether the process works or not, and then proceed to gather that data for future analysis.
For example, if to you, the measure of an efficient process is the ratio between number of tasks done in a week and their difficulty - you're going to need data on number of tasks done each week, and of each their difficulty points - it's as easy as that.
Imagine the potential such an approach has - but while you do, also keep in mind the amount of work running custom metrics on a regular basis will consume. Not a problem if your team includes someone, whose main job it will become. A bit of an issue though, if it's just you and a few workers - in that case, developing own productivity metrics might be the last thing you'll have time for.

For smaller teams like that, it may make a lot more sense to use auto-generated metrics, like the ones offered by Kanban Tool. It will run the following analytics for you, without any input needed -

• Lead & Cycle Time, showing the length of time that tasks took to get from point A to B,
• a Time Report, listing exact hours that items were being actively worked on and by whom,
• a Cumulative Flow chart, visualizing done work, accumulated over time,
• and a Changelog - a searchable list of all changes made.

No matter what you do to learn about your process efficiency, there are a few things to keep in mind: don't focus on just one, very specific metric, as that can lead to tunnel vision and misrepresent the overall process health. Also, don't obsess about measuring the flow in general, the rule of thumb here could be: if you suspect something is going wrong, get measuring. If things seem to be OK, sticking to scheduled routine checks will be just enough to let you know of any flaws and improvement opportunities.