Saturday, November 23, 2013

The Holiday Stratagem

What do the Thanksgiving and Christmas holidays do to your teams tempo and cadence?

For most US teams this holiday period from mid November to after the first of January is hectic and disruptive in various ways.  This calls for a Holiday Stratagem.  A trick to allow the team to have some semblance of continuity and flow during these times.
The Sontaran Stratagem - Doctor Who

To discuss this stratagem let's first define some terms:

Tempo - the rate of workdays to calendar days
Cadence - the beat of the sprint events to fall upon the same calendar day of the week
Sprint duration -  in work days (in this example I'll use 10 work days)
Sprint length - the length of calendar days between two sprints (14 days for the example 2 week sprint)

What do you value more?  The ability to deliver more work in the short term -or- the ability to predict long term the capability of the team to deliver that work?  Or perhaps something else, like teaching a new team the trade-offs in making decisions and helping them to learn from reflecting upon these decisions.

This is one area that I find lots of variability in agile coaches answers when the team just wants an answer to the question - what do we do?  I love this variability.  I enjoy the ensuing debate over the better answer and delving into the reasons why for a specific case.  And this is definitely one domain that is not simple - nor is there a best practice.  So reader beware - there will be no one best practice, no correct way - here is the land of trade-offs.  And it is an interesting land well balanced between the trade-offs such that the replications of the decision are virtually uninteresting.  I think that makes a great play space for the team to practice decision making.

Let's use an example and then discuss the options.  Team Alpha-Dog is on a 2 week sprint schedule with the beginning on Monday (Sprint Planning) and the end of sprint activities (Demo & Retro) on the Friday of the following week.  Along comes the 2013 holiday season with the company's holiday plan for Thanksgiving days off on Nov. 28th & 29th.  In the sprint planning on Monday, Nov 25th the team has a decision.  Do they keep the sprint duration constant (at 10 working days) and there by break cadence - or - do they break tempo and keep the sprint length consistent at 14 days?


Option A - Illustrated
Option A:  Sprint duration constant at 10 working days.  Sprint begins Nov 25th and runs to Demo on Tuesday, Dec 10th.  The cadence of the team and stakeholders will change.

Option B:  Sprint cadence constant at 14 calendar days.  Sprint begins Nov 25th and runs to Demo on Friday, Dec 6th.  The duration changes to 8 working days - impacting planning and subsequent velocity calculations.

Option C: Sprint duration and cadence are broken.  Sprint begins Nov 25th and runs to Demo on Friday, Dec 13th.  A sprint duration of 13 work days to try to reestablish the Monday/Friday cadence - just on an alternate rhythm.

I'm sure someone will offer another option, I'm just going to stop with these three and quickly rule out option C as the worst of all.  It's always nice to have an option everyone can agree upon - even if it is the worst - at least we all agree.

One may want to do a bit of mid-term planning before making the decision.  What will happen with the Christmas and New Year's holiday if we continue with the decision principles we use now; does this leave us in a cadence we wish for the coming year?



If the team desires to switch it's cadence or duration - there is no better time than now.  Have you met the 13+2 Sprint Cadence?



If the team chooses option A then planning is easy, the average velocity calculation is consistent (no need for an asterisks in the log book); however the team ends up changing the sprint cadence and this may wreck havoc on stakeholder's schedules.  The impact to this decision typically ripples outside the team more than inside the team.  It's a good test of the servant leadership paradigm at your company.  Do the leaders server the team or does the team serve the leaders?

If the team chooses option B then planning is a touch harder involving fractional math (I've found this surprisingly difficult for some,  capacity this unique sprint is 8/10 times the typical sprints velocity).  This sprint's velocity may be thrown out of the set for future statistical calculations.  Stakeholder's may be happy because they don't have to rearrange their busy calendars.  But the externality impact to the team may be hidden, it is the compression of the sprint that will throw them off, they will be off tempo.

Projecting these decision into the mid-range future and through to the middle of January may help you make a decision on principles.  But don't expect these principles to hold true when it comes to the following year's calendars.  These are not true principles but just situational guide lines.  It is a good time to practice team consensus making tools like the fist of five.

Have a happy holiday!




Post a Comment