Team Practices Group/FY2016Q3 Work tracking

Review TPG FY2016Q3 Work Tracking reportEdit

What went wellEdit

(in terms of tracking work and finishing it?. Our original goal was to do retro on both the work and the work tracking, but actual discussion focused on retrospecting the work tracking.)

  • Phab tags allowed us to categorize work
  • Is "what did TPG work on" useful?
    • AR: useful for planning next quarter.  Best if available before quarter planning.
    • KL: This gross level of seeing stuff is potentially valuable, maybe after a couple of quarters of tuning this is. How often would we look at it? Could look at it any time.
    • KS: value is debateable but cost seems low enough to keep experimenting.  for other teams, would be a lot more useful.  KL: why is value debateable?
    • GG: agree that cost is low; don't think it's reflective of the time we spent.

What went poorlyEdit

  • JA: never resolved definitional conflict of translating quarterly "plan" to Phabricator.  E.g., we had 3 quarterly goals, which broke into 6 milestones; we put all 9 items into Phabricator, which was confusing.
  • We finished some things after end of Q, so we fell behind on paperwork
  • Parent Tagging in Phlogiston/Phabricator is very lossy
  • Cannot currently tell if we are working on high priority stuff vs lower priority stuff (by category)
  • High overhead using Phlogiston to help track and plan TPG vs embedded workload? (personal workboards in addition to TPG etc)

General observationsEdit

  • "completed in Phab" means that there are some resolved tasks and that there are no open tasks
  • Could never accurately predict when we'd be 'finished' because lots of work defined ongoing (is this a bad thing? we never had consensus about whether or not we should)
  • Work from essential functions seems to grow faster/more than other work - what does that mean, does it matter?
  • Gut feel that we only tracked ~50% of our work, but obviously cant be sure since... it wasn't tracked
  • Estimation
    • GG: We aren't actually measuring time. (because we're not doing estimates)
    • Using task counts instead of estimated work means it's less informative
    • Could use placeholder tasks with points to estimate our embedded activities
  • JA: should we look mid-quarter?
    • Why? - to refine data collection, to learn from it.
    • KL: to keep a finger on the pulse - if there was every a point where we were doing zero goal work, that would be good to know.
      • AR: e.g., if that happened we would talk to our stakeholders,
  • KSː if we really do this next quarter, would like to see Uncategorized at Zero.
  • KL: this could forecast completion and make sure we don't fall behind, but not sure how big of a problem that is for TPG right now, with our weekly check-ins.
  • KS: biggest thing for me is to look back and say, wow, it took 33% of our work to do our goals.  We are a ways away from being able to do that, not sure what I would do with it if we had it, but that would be most interesting.

Open Q'sEdit

  • If you create a tracking task, and have a bunch of children, does closing the tracking task count, or should you only count when the children are finished?
  • Should we change categorization SOP for Q4?
    • GG: if we track Qgoal vs essential (which is current SOP), makes it easier to prep for Quarterly Review
    • KL: if we had separate subtags for different quarterly goals we would get ease of Phabocity - being able e.g. to look at the board for light engagement, helpful to product owner/work manager.

Next stepsEdit

Joel to write up notes and follow up by email