After my last article, I said I’d be writing about the extra features available in LeanKit that allow metrics to be derived. This will not be the only focus of this article. I’m including my recollections of running my Personal Kanban over the last few months. Lets first start with what could be some problems.
For the last 2-3 months I’ve been working on a project this has meant it has remained in the In Process column for this long. Many of the other tasks are very short – they stay In-Process for hours at most. Also when I’m reading a book this can stay In-Process for a long time. This is also like projects and as I mentioned in Part 2 and cards like these are supposed to be a place holder for projects.
Is this really a problem then? Maybe or Maybe Not. Cycle time has been increasing as shown below in the Cycle Time graph from LeanKit. I’ll talk about why this may be a problem, in so far as a Personal Kanban is concerned.
Is Task Size a Problem
Maybe, maybe not. It could be how you perceive it. For a Personal Kanban an average cycle time of 5 days might not concern you. Maybe it would come back down. Maybe it will find it’s average here and stay around the mark, reflective of the project work that is being done.
However, maybe an increasing cycle time could lead to the following types of dysfunctions:
1. Lack of Milestone Motivation
Finishing something creates a feeling of satisfaction. Leaving big jobs in process means this satisfaction does not occur. Feelings of ‘When will this be finished’ occur. It then can feel hard to keep going and requires an extra bit of self motivation to keep going.
This follows from the first point. Having a big job is more reason to avoid doing it, taking longer to get started. Feelings of guilt might arise. A uneasiness and even unhappiness can result. Can this effect quality. Perhaps, in my work has been validated to be of a very high quality. However, for me, it has created a peaks of valleys in the emotional roller coaster. One needs to wait to the end to get the feedback. I did feel this emotion and asked to have a colleague look at my work midway through to validate it somewhat. That was useful.
As mentioned in point 2, lack of feedback created feelings of whether the work was of a worthy nature. Dividing the work in to separate smaller tasks allows feedback to asked for on a regular cadence. I reasoned at the time that the project was a sole effort and therefore wouldn’t require such a rigor. I think that was an error. I think this is case for why Pair Programming is such a good practice. Positive feedback on a constant basis should be paid back in better quality and better overall cycle time.
Don’t Put Big Projects in a Personal Kanban
Another solution is not to have big projects in a Personal Kanban. Do they belong here? You wouldn’t put your work projects on a Personal Kanban would you. A project has more rigorous policies around ‘doneness’ Personal Tasks could be argued that they probably don’t have amount of rigor involved.
A project is a different Class of Service, mixing the two is probably ill-advised. You’d rather feel good about what you are doing and having what you are doing traceable to the type of work you’re doing could help to reduce uneasiness around throughput.
Beware the Holding Pen
I found that my Holding Pen column was getting a little over used. Long running tasks or putting off tasks is an indicator that the task is too big or just an excuse to procrastinate. Consider a WIP limit to force resolution one way or another.
I also recommend having a WIP limit on the Ready Column. I introduced one on this column to force me to do the work rather than have it for another dumping place for tasks, making it difficult to prioritise. The dumping place for all idea still exists and WIP limit on Ready reminds me not to use Ready for these ideas, however ‘urgent’ stuff does tend to find its way there and stay there for some time. Maybe this limit can be revised down.
Let’s talk about some of the other charts in LeanKit. These charts are available in the Team and/or Portfolio editions and not available in the Free Edition. They may or may not be useful in a Personal Kanban.
The most interesting chart for me is the Card Distribution By Type. I was a surprised at the amount of email I was sending at 15%. This is not be confused though with the amount of time being spent on that task or any other task. I tend to think my Project work is taking up more time than this 🙂
With the By Priority chart I find that I tend to allow most of my cards to be normal priority. Rarely is something so critical that it needs immediate attention. This may indicate that I’m not so rushed to get stuff done or a little lazy in setting the priority. Setting priority is a fleeting thing for me, only a second or two for consideration. It’s not a big deal.
The By Lane chart appears to show that I’m getting through work at a good rate, over 75% when you include Done and Retro lanes. It would be even higher if I include the Daily and Weekly habits which cycle through ToDo, In-Process and Done on a daily or weekly basis.
Cumulative Flow Diagrams
Probably the most important if not most used chart in a Kanban Value Stream. Here’s an early one. It looks quite nice. Seems to be getting through work quite well, this is before the big Projects entered the mix. Slight fattening of the In Process (green) band about 11 Feb 2014, only indicative of the size of the work but maybe also meaning that the work in there is too big – as I mentioned earlier this was a problem.
When viewing the chart without card size your get quite a different view as shown below.
In the next example you can see the full screen dialogue. The bottom of the screen shows the lanes that have been selected for the chart. This is important as I also have separate ‘small’ projects and Daily and Weekly habits which I didn’t want to include in the main kanban. It does allow you to select separate parts of the board and do separate analysis, like on the Projects.
These show allocation of work across the categories of Completed, Ready and In Process. Lanes are allocated to these types and lanes which aren’t allocated fall into the Unknown bucket.
Below is the diagram that shows the percentage and it shows not significant variation.
However this diagram by Queue Size shows a gentle increase indicating the larger items that were taken on.
However the next diagram calculates the same diagram but with card size included. Overall I don’t read a lot into these charts from my board.
To check that I haven’t missed anything I went in search of other examples from elsewhere. I found the documentation in LeanKit not providing a lot of clarity about how to use this diagram or not enough to make sense to me. I did find something on their blog. Quite interesting variations there. Completion Rates weren’t very uniform as my examples suggest. Looks like more complex work involved. Could it also mean work was in different phases of the value steam. Be interesting to compare the boards at these times.
There is another diagram called Process Control . It doesn’t seem to work for me. I revisit this blog entry when I work out what is wrong.
I find it enjoyable viewing work entering the board and progressing across. All my ideas get captured as well. I’ve fell into some traps but at least I can see them and I’m actively managing the situation to improve. I’ve added columns, moved them around and tinkered with WIP limits and added some where I felt they were needed. It’s getting better all the time.
Look forward to using this even more in other settings.