So,

The good old Blitz Planning never fails! I have started a new engagement recently, and although I have been doing my best, some days, as always, are better than others. Today was better than others. I was asked to facilitate a retrospective and a planning session. I think I did alright. Time will tell, which in software land means, I will find out within days!

Well, my colleague Nancy mentioned that I managed to make the group a lot happier! Oh my, I wish I could take credit for that. Alas, self-organising teams moods don't shift because of a retrospective facilitator (even one as charming and funny as my self).

The orange ticks are the before retrospective emotions. Thus, seven miserable, six mediums and one happy. The blue are the post retrospective emotions. Thus, two miserable, eight mediums, and three happy! Those astute readers will see there is a vote missing, my colleague Callum was at the dentist poor fellow, and then he was high on Novocain for the planning! (That was a joke. You can't get Novocain in England.)

What is interesting is that moods change at all during a retrospective. This is key, a large mood shift could indicate that there was a lot of information hidden. Or, in the case of a positive shift, people sometimes just feel a hell of a lot better after been heard.

After the retrospective we did a planning session. This was not my choice, I never ever subject a team to back to back meets like that! What cruelty. We are in the final stages of the project and this is where a plan with dependencies and time lines can really help. This is a pretty simple practice. You line the current iteration's work up on the board as normal. Then you take the rest of the backlog and you create a chain of dependencies. This allows the project manager to have a detailed release plan.

By the way, for the XPers, you can't plan with no dependencies in mind... you see, that only works with generalised teams, i.e. no specialists. So, for a non-perfect project you have to get smarter with your planning. The planning game is too naive for most teams. It's the ideal solution for an ideal world.

One technique I like to use at this part of the project is one week iterations. This is really a laugh. The left most items feed into a new iteration plan and the team essential eat the release plan. If the team isn't eating quickly enough then something may be wrong.

A release plan with felt marker to highlight dependencies.