The Dog Food Diet: Sprint 3

Overcrowded_ferry_boat_on_Meghna_River,_BangladeshWhen the home team wins a game they’ve played poorly, the announcers often describe it as “not an artistic success”.  That’s how I feel about this project right now.  Like the overloaded river ferry on the right, we’re getting somewhere, but it ain’t pretty.  We shipped some water in Sprint 3.

Sprint Goal: The physical element being used for its designated task and one instance of the new embedded technology running within the physical element.

Sprint Planning: The product backlog is in good shape.  We come to Sprint Planning having achieved a velocity of 50 points in Sprint 2.  So what should we commit to in Sprint 3?  If you guessed 50 points, you’re right (sort of, see retrospective below).  We over-committed seriously in Sprint 2 despite having a known velocity of 18.  So what did we commit to in Sprint 3?  If you guessed “a little less than 50 points” you’re right.

Sprint Review: We showed an instance of [a thing] being built in the physical element
but did not get an instance of the new embedded technology running our skeleton app mounted in it.  The sprint goal was not met.  We achieved a velocity of 32 points.  More on that later.


Sprint Retrospective:  Anyway … we’re happy with our backlogs, which is nice, but you don’t have to be an agile savant to smell the #scrumfail in the Sprint 3 burndown chart.  We missed the sprint goal and for the third sprint in a row we achieved less than we committed to.  In this case we committed to 45 but only achieved 32.  The 45 point commitment was (marginally) defensible because we hit 50 points in Sprint 2.  But it didn’t work.  This brings us to a great sprint planning tool – the rolling average.

What would have happened if, instead of using the 50 points from Sprint 2 (the simplistic version of yesterday’s weather we should have used in Sprint 2 planning), we’d used the average of Sprint 1 and 2?  Sprint 1 – 18 points, Sprint 2 – 50 points, the average – 34 points.  Almost exactly what we got in Sprint 3.  Lesson learned.  In fact, what we will use going forward is the rolling average of the last 3 sprints.

We also backslid by adding 20 points of scope mid-sprint.  Looking back at that added scope we had a big development story, a development research story and a business research story.  The two research stories got done, the development one didn’t but they were all prioritization mistakes by our PO because they didn’t need to be done right then.

A Brief Rumination On The Beauty Of Points:  We found ourselves, at one point in this sprint, blocked on the top stories we wanted to work on.  We sat staring at the sprint backlog and somebody finally said:

What can we do to score some points today?

It took awhile, and trying to start another couple of stories that turned out to be blocked, but we found a winner and went to town.

Now, there’s nothing noble about that – as “highly motivated self-starters” we should know what to do and be doing it already.  That said, it occurred to me that there was something important exposed by this episode.  Namely, that in a non-point-based system, we would have picked something to work on, and just worked.  In a point-based system, we found the highest priority thing we could take to done and did it.

“Doing work” as we would have done in a waterfall system, creates sunk cost, that’s all.  Taking a user story to done creates business value.  And that’s what point-based accounting incentivizes – taking user stories to done.

“Scoring points” is not the way I typically think about “getting work to done”, but it’s not wrong.  If it helps you then use it.

How did our kaizen from Sprint 2 work out?  Our kaizen from sprint 2 was to keep our sprint backlog stories to 8 points and under.  In fact, we took two 8s, started but didn’t finish one, and didn’t start the other.  A wash.

Our kaizen for next sprint?  The consensus is that our velocity without interrupt-chaos would be somewhere around 40, so our kaizen for next sprint is obvious – push back on adding scope mid-sprint.

Notes on sprint length: The fact that we run 2 week sprints accounts almost entirely for the 20 points of added scope.  If the next sprint started next week instead of a week and a half from when they were added we probably could have resisted the urge to throw them into this sprint.  This is an argument for one week sprints.




About JR
Software guy, startup guy, non-fiction glutton, south shore inhabitant

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: