Growth in Retrospect

• QA and Sprints were the biggest time sink

  • Eliminated the presence of a QA team, as the features were smaller and easily testable by another engineer (don’t worry the features were still tested by another engineer)
  • The engineering teams were competing to secure QA resources.
  • The solution was right in front of us to test them better through automated tests and run through them by another engineer in the team.

• We transitioned to Kanban and said final byes to Sprints

  • This gave life to weekly shipping cadences
  • Adopted a v1, v2, v3 approach for bigger features by breaking them into smaller chunks
  • Removed subtasks from a ticket pragmatically if they were deemed to be a time sink and we were missing our weekly cadence → then shipped it as a fast follow

bye bye sprints . hi kanban . bye bye sprints . hi kanban . bye bye sprints . hi kanbannn hellooo

• Only one meeting per week

  • No sprint planning, tech planning, or other unnecessary meetings
  • Conducting all activities asynchronously and through RFCs
  • Strong product leadership helped us close decisions quickly during our weekly meet
  • When an engineer or designer felt uncertain, a huddle was initiated in our Slack channel to provide immediate assistance.

• Open communications (via public slack channels) and was visible to the seniors

  • When we raise a concern - we have multiple eyes that would trigger great discussion leading to great solutions
  • This meant we were open to receiving suggestions for improvements from the senior team (product/engineering)

• The HEALTHY PUSH - that comes from the lead to meet deadlines

  • I would always scan the board and try to push/jump in to speed up things.
  • If I see a ticket taking too long, I would flag it and would offer help in any way – this puts healthy pressure on the engineer.

• Always being agile

  • I’d look at our Kanban board (under this week’s feature list column): and think how much can we ship today?
  • If one of the engineers is working on something but we won't make the weekly cadence deadline, I’d ensure they would help someone else's ticket, so it can be shipped instead.
    • Someone working on their own thing doesn't add immediate value for this week – it's better off contribute toward team shipping cadence.