FY21-Q4 Development Department OKRs

Reduce the pain for customers ARR Objective => 51.25%

• Key Result: Fix 10 of 16 Fulfillment pain points => 50%
• Key Result: Complete Migration of HAML component instances to Pajamas => 52.4%

Improve our product use Product Objective => 73.3%

• Key Result: Address transient bugs and improve performance and quality tooling => 80%
• Key Result: Demos and project management on Tiering Changes => 100%
• Key Result: Increase Narrow MR Rate from 10 to 11 => 40%

Strengthen the team - Team Objective => 100%

• Key Result: Continue to expand use of the architecture discipline => 100%
• Key Result: Performance reviews completed in-budget by Jan 31, 2021 => 100%

Retrospective

Good

• HAML migration - As it was before, the community involvement was high. Hosting this topic at the hackathon really helped.
• HAML Migration - A portion of these issues were fixed outside this effort as general feature / tech debt work.
• Transient bugs - Great teamwork across Dev section and Quality
• Transient bugs - Looking at roster and participation in working group we are split evenly based on gender. This wasn't intentional, but still is an interesting result.
• Demos
  1. We completed the EoA project and so far early feedback is that it has been a success.
  2. We did several demos
  3. The team was creative in solving problems and working quickly.
• MR Rate - We hit higher productivity marks than last year every month during the quarter (0.5 to 1.5)
• Architecture
  1. The architecture workflow helped us to unblock progress on multiple initiatives that had not been going well before
  2. We made progress and eventually concluded the Cloud Native Build Logs initiative to unblock Kubernetes migration
  3. We got a lot of valuable input from PMs and EMs that are getting involved in the process of writing blueprints early
• Performance and Compensation
  1. We were on time.
  2. We were in budget.
  3. We created a communication message.

Bad

• 10 of 16 Fulfillment pain points - Because of competing priorities, we were unable to complete as much as we aspired to.
• HAML Migration - We didn't focus on larger components. In the last retro we wanted to try this, but we didn't.
• Transient bugs - we blew through our goals of 7 transient bugs and fixed 14. We weren't aspirational enough.
• Demos
  1. We didn't use a scorecard for demos
  2. We finished up a week later from our original goal
• MR Rate - Because of vacations we didn't hit close to our goals for the quarter.
• Architecture
  1. The architecture practice is still something new to team members and we hear a lot of questions about how it can help
  2. We saw two blueprints being stuck in writing for longer than we expected
  3. Assigning DRIs for a blueprint execution is still sometimes challenging because of the scope of work
  4. It sometimes takes a some effort and time to find an architecture coach who is willing help with a blueprint
• Performance and Compensation
  1. We didn't track this explicitly, so our 100% is based on guesswork.
  2. We had too much back and forth in review process for performance evaluations.
  3. We had too much back and forth in compensation review.
  4. We tried to give EMs more control, but early feedback is they didn't feel it mattered much.
  5. Our guidance within team was lower than company guidance because of developments unique situation.

Try

• (Many) Think about making realistic goals based on competing priorities.
• HAML Migration
  1. Categorizing the remaining issues by stage / group. We may be able to get them scheduled in as regular product work. I've spoke with @stanhu about unleashing Tanukibot on these issues to aid in this effort.
  2. Triaging these issues. We've done light triaging periodically for buttons and alerts, but there's likely some issues that can be closed now due to them being fixed as part of regular product work.
  3. HYPE! We've done a great job in the last two quarters generating interest with the larger community, we should push on doing the same within GitLab
• Transient bugs - Consider asking for a retro of this working group. Was it better/worse than others.
• Demos - Scorecards for demos in future projects
• MR Rate
  1. Lower goals for Q1 to be more in line with standard objectives.
  2. Work with a couple of teams specifically on productivity goals.
• Architecture
  1. Schedule an AMA about the Architecture Practice at GitLab in FY22-Q1
  2. Document the role of an architecture evolution coach in getting a blueprint written an approved in a timely manner
  3. Encourage technical leaders to get more involved in the architecture practice
• Performance and Compensation - Followup retro discussion with total rewards to affect change.

Cache Refresh: 2020-02-08