The article lists the most important Kanban metrics you should monitor to measure your team’s performance using Lean management. These are particularly helpful indicators that can help you avoid system failures, predict project completion time and apply other Lean methods.
Do you want to know how to effectively measure your throughput? Or keep an eye on lead time? And what about visibility – do teams have it or not? Well, these metrics will tell you exactly this! Let’s take a look at them one-by-one.
Lead Time (or Cycle Time) — How Long It Takes To Get A Done Task Out The Door. Lead time is probably the single most important metric in kanban (from now on I’ll simply use this word for both kanban and lean). It shows how long it takes to get a task out of the door.
Measure lead time by taking the interval between two events: The first one is when a work item enters your workflow, say being picked up from a team’s “Ready” column. The second event is – you guessed it right – a work item successfully completed.
One can use PERT or similar estimation techniques to figure out how much time is required for each step in the process, but these are typically quite inaccurate. Instead, why not ask people involved? They must know what they’re doing or at least should have an idea about their rough capacity (more on that later). Ask them to do all tasks manually using a simple spreadsheet tool like Google spreadsheet or a ticket system. Here’s how to fill in a simple spreadsheet:
As you can see, estimating is super-simple and takes around 5 minutes per task. We’re going to use times from this spreadsheet later on. The next step would be to distribute work items between processes (or sub-processes). For simplicity assume that each process has three steps; whatever the number is, it doesn’t really matter for this article, but I will mention one big downside of having many short steps. Assuming your team does not have dedicated QAs they may likely start working on the most important task first, which may result in building the wrong product! You need to make sure they don’t forget about future business value they add by verifying it functions normally.
– QAs need to do their part and add some quality marks on the tasks they work on. This is just a minor step but it ensures we keep our value in mind while working on unimportant tasks which would become important in the future. Marking a task as “ready for dev” doesn’t mean you’re actually finished with all checks and ready to deploy, it’s just an indicator that if you hit the ‘deploy’ button right now this task looks at its best! Remember to get us back on track, when needed – If you think one of your mates messed up or anything else may get in your way, seek help from management or product owner/creator, explain to them the current state of things and get their help in getting back on track.
Just remember to keep your value in mind while you’re working on unimportant tasks, because the next day they’ll be important to somebody else! Marking a task as “ready for dev” doesn’t mean you’re actually finished with all checks and ready to deploy – it just means if you hit the ‘deploy’ button right now this task looks at its best.
If someone messes up or anything else may get in our way, seek help from management or product owner/creator, explain to them the current state of things and get their help in getting back on track. At the end of each sprint we always ask ourselves: did we increase our value? Did we work towards our mission? Did we contribute to our long term vision?
Finally, when you work in a team don’t expect others to perform tasks you can do yourself – many hands make light work and getting people involved in different process stages will help everyone get a better understanding of how things work and why.
Don’t be afraid to speak out if the task doesn’t align with your personal values, it may even turn out that there is someone else who feels exactly the same way you do. When you come across problems or obstacles during development check previous sprint’s retrospective (s)hedule for similar cases and use them as examples when preparing your own report. When reporting any issue, make sure to provide all data needed for conclusion, this ensures no time is wasted on guessing what was causing the problem in the first place.
Always keep in mind that everything will take longer than expected, especially if there is no reliable data on a similar project. When someone asks you to do something that you are not clear about don’t be afraid to ask questions, people won’t get mad at you for wanting clarification on information provided or requirements set.
Don’t forget to document all relevant decisions made even if it seems like there was nothing special about them – just look back after a few sprints and you’ll probably notice a bunch of tasks which took way more time than assumed so remember this for future reference.