In the Agile Alliance LinkedIn group discussions, there is a discussion about 'how to adopt agile in my organization?' by Mark Lummus. This is a complex topic, with many things to say. Here is most of my first post (in that group) about this. There are many other things I might have said.
QUOTE
Mark,
Here are my opinions on this (based on much experience). And I am talking not so much to you (you probably have the experience to consider some of my remarks as obvious) as to a broader audience.
First: I think really doing lean-agile-scrum well requires lots of things. But I think I am agreeing with you to say: Start the change with Scrum.
Second: Go high and go low. Meaning: Get senior guys in support and get people in the teams to support.
2b: You need to build a team (or each one needs to be a real team), and it needs to be a team with a winning spirit. Scrum helps with this, but there are many other factors that contribute.
Third: Use some metrics to show that you are making progress. Do not let them insist on perfect metrics (never exist) and make sure people use common-sense in interpreting the metrics (common-sense is not so common).
Fourth: Being aggressive about removing impediments is key. In the team, by the ScrumMaster, by the managers. Hard to predict from afar where your biggest impediments will lie: people, rigor in doing Scrum (too much ScrumBut), engineering practices (see XP!), the flow of business info into the team (eg, weak PO), or what.
Fifth: Build off the success of the team(s). To me, success is more business success (as in: the customer is really happy) rather than technical success (we did what we said we would do X months ago). I don't know what you all do today, but I can just about guarantee, after 3 Sprints, some degree of relative success. Build on that.
Fifth: The 'change management' around good lean-agile-scrum is quite significant. It is never-ending. People think they understand lean-agile-scrum, but I think the paradigm shift is actually hard. Typically firms need coaching for the teams (and surrounding managers). We have seen a team have a full-time coach for 2 months. And I have seen a team get a coach for 3 days out of each 2-week Sprint, for a bunch of Sprints. Almost always, when I see a firm that chooses not to get coaching, it smells to me like being 'penny wise and pound foolish'. But I am sure there are also some cases of strong success without coaching...just not often, percentage-wise. (I am a coach; you might call me biased.)
Sixth: Technical debt and better technical practices will be very important, sooner or later. Start attacking them soon.
True success with Scrum should be measured as 5x-10x better than your current baseline. Some firms do this quite well and often, but most never get there. Far, far too many settle for minor success (eg, a 50% improvement). I think the secret sauce is not so much the practices of Scrum (the dance steps), but rather more people 'getting' the underlying values and principles (the music). The dance steps become truly powerful when in sync with the music. Without the values and principles, a 'daily scrum', for example, can be almost a waste of time.
What would others put as the top 6 or 7 things for 'Mark' to consider?
UNQUOTE
More on this topic in a bit.
QUOTE
Mark,
Here are my opinions on this (based on much experience). And I am talking not so much to you (you probably have the experience to consider some of my remarks as obvious) as to a broader audience.
First: I think really doing lean-agile-scrum well requires lots of things. But I think I am agreeing with you to say: Start the change with Scrum.
Second: Go high and go low. Meaning: Get senior guys in support and get people in the teams to support.
2b: You need to build a team (or each one needs to be a real team), and it needs to be a team with a winning spirit. Scrum helps with this, but there are many other factors that contribute.
Third: Use some metrics to show that you are making progress. Do not let them insist on perfect metrics (never exist) and make sure people use common-sense in interpreting the metrics (common-sense is not so common).
Fourth: Being aggressive about removing impediments is key. In the team, by the ScrumMaster, by the managers. Hard to predict from afar where your biggest impediments will lie: people, rigor in doing Scrum (too much ScrumBut), engineering practices (see XP!), the flow of business info into the team (eg, weak PO), or what.
Fifth: Build off the success of the team(s). To me, success is more business success (as in: the customer is really happy) rather than technical success (we did what we said we would do X months ago). I don't know what you all do today, but I can just about guarantee, after 3 Sprints, some degree of relative success. Build on that.
Fifth: The 'change management' around good lean-agile-scrum is quite significant. It is never-ending. People think they understand lean-agile-scrum, but I think the paradigm shift is actually hard. Typically firms need coaching for the teams (and surrounding managers). We have seen a team have a full-time coach for 2 months. And I have seen a team get a coach for 3 days out of each 2-week Sprint, for a bunch of Sprints. Almost always, when I see a firm that chooses not to get coaching, it smells to me like being 'penny wise and pound foolish'. But I am sure there are also some cases of strong success without coaching...just not often, percentage-wise. (I am a coach; you might call me biased.)
Sixth: Technical debt and better technical practices will be very important, sooner or later. Start attacking them soon.
True success with Scrum should be measured as 5x-10x better than your current baseline. Some firms do this quite well and often, but most never get there. Far, far too many settle for minor success (eg, a 50% improvement). I think the secret sauce is not so much the practices of Scrum (the dance steps), but rather more people 'getting' the underlying values and principles (the music). The dance steps become truly powerful when in sync with the music. Without the values and principles, a 'daily scrum', for example, can be almost a waste of time.
What would others put as the top 6 or 7 things for 'Mark' to consider?
UNQUOTE
More on this topic in a bit.
No comments:
Post a Comment