Comment by wpietri | Hacker News Reader

Comment by wpietri | Hacker News Reader

Parent Full thread

I personally recommend physical boards when possible, especially for people just getting started. E.g., my startup-sized kanban board is here:

http://williampietri.com/writing/2015/the-big-board/

You can happily use, e.g., sticky notes on a wall by your desk or inside a cabinet door. If you really can't do that, software can be ok, but it's easier to develop bad habits with software. With a physical system, certain things are usefully hard.

When I'm starting someone out, my standard columns and WIP limits are:

  * to do
  * soon - max 5
  * on hold - max 3
  * working - max 3
  * done

The "to do" column is the backlog. "Soon" is the things that I'm doing next. When I start on something, it goes in "working". If I'm blocked on an external dependency, it goes to "on hold". Items never go back to "to do".

When I've finished an item and want to know what do to next, I'll first look in "working", then "on hold", then "soon". If all of those are empty, it's time to think about strategy for a few minutes; I'll look at everything in "to do" and promote a few to "soon".

Hopefully that's enough to get you started. The important to remember is that the board is supposed to mirror (and shape) how you think about the work. If somebody happened to randomly ask you "Hey, what are you working on now?" it should generally be what's in "working". If somebody asks what you're doing today, it should be the stuff that's in "working" and "soon".

If you have different categories or WIP limits in your head, that's fine, go ahead and represent that accurately on your board. But every month or so look at the board, think about how you work, and tinker. E.g., I might say something like, "Things felt really chaotic this month, so I'm going to try dropping the WIP limit on "working" next month.

Does that help?

Thanks. That helps.

Is there any place or book where I can read more about the "personal kanban" principles.

Basically, I don't want to overdo it either. Want to understand the principles before jumping in.

There's this one: https://www.amazon.com/Personal-Kanban-Mapping-Work-Navigati...

As cotsog says, there's a whole personal kanban literature. I personally haven't read any of it, but I hear good things.

One tip, though: Be careful of focusing too much on the theory.

In my experience, a lot of task management methods fail to take root because the person is supposed to follow the system. However, the system doesn't really fit the person, so what is begun in hope ends up gradually falling apart. Doing a kanban approach right is about paying close attention to what's actually going on and then tuning the system.

As an example, take WIP limits. When I start teams with kanban approaches in work settings, they often want to pick high WIP limits, because a) they currently are doing a bunch of things, b) it seems macho and achievement-oriented to have many balls in the air. But they also want to follow the system, which says "minimum WIP". So they will either pick a high limit, and so get no benefit, or pick a very low limit and get frustrated with all the chaos that a big process change causes.

The approach I instead suggest is to just lay out the board to match what's really going on. Inevitably, they have a lot of WIP, so that's what the kanban board should show. If that's 10 things, it's 10 things. They agree that's a lot, so we might set a WIP limit of 8. If that works, we'll drop it a little more. And then again. At some point, it will get uncomfortable, and that's when things get really interesting, because it's exposing a problem in the workflow. Do you change the WIP limit back? Or do you say, "Hmmm, how can we remove the blockage?"

There's no right answer, really. It's about paying attention to the details of the work and your experience. Kanban boards aren't really a solution. They mainly make your workflow visible so you can notice problems and fix them. The magic isn't in the board. It's in your willingness to iterate.

Reply on news.ycombinator.com