For firefighting situations I prefer using the Kanban framework. It is very informative, and generally keeps things under control.
If there are a lot of things to do, people tend to do context switching, which makes them less effective than they could be. Kanban limits the number of cards in a particular column, which does not protect developers from context switching. The number says nothing about the individuals, only about the team.
One method I find usable is to have a **limited number of Kanban avatars **(custom picture on a magnet, a sticker, coloured magnet etc.) for every colleague. If someone is working on a card, he or she must put one of his or her avatars on that particular card. One advantage of this method is that everybody can see who is working on what, and the number of possible context switches is limited. With two avatars, everyone works on maximum two items, and switching between these two is not that difficult to handle as switching between four or five. Finishing one card is still the highest priority.
If, for some reason, one card is blocked, the avatar stays there, and that colleague can put another avatar on another card, if he or she has any left. If not, an already started task shall be finished. If that is blocked as well, then the team shall solve the blocking situation.
Another interesting question is “Can we move items backwards if they are blocked for some reason?”. The usual, less agile answer is, “Yes, why not? At least we can work on something, because we’ll have a free space in a column”. There is some truth to that, if it solves the current situation and the team can continue to work. I see more disadvantages, though:
a possible impediment is not investigated
what happens with the priority?
a card may get forgotten
Instead of moving items back, use blocking signs. There is nothing new in using these signs,the point is to use a limited amount of them. If a card is blocked, place a big red sign on it, so that everyone can see it. Until the pressure is not too high and there are no dependency issues, the team can mark as many cards blocked as many signs they have. If they run out of signs, then it’s time for a short on the spot retrospective to find out what is going on, because the team is facing a bottleneck situation. With this approach
the cards can stay where they are
no problem will become hidden
the original Kanban number can fullfill its purpose
Try it out and give me feedback.
Updated: here are some real life pictures on avatars and stop signs:
A regular column, where the first task is blogs and we were working on the other items, either alone or in pairs:
Mind the red strikes on the left side of the cards. They mark the age of the items. It seems almost every item is 2 days old in that column.
This situation is quite rare, but in order to finish something almost everybody had to work on the same task:
comments powered by Disqus