GitLab Next

Objective (IACV): Assist IACV and reduce Opex => 98%

• Key Result: Deliver top IACV Drivers => 85%
  • Development: Gitaly HA GA eventual consistency => 100%
  • Development: Achieve Disaster Recovery Viable Maturity => 69.5%
• Key Result: Reduce cost => 100%
  • Quality: Reduce GitLab CI monthly cost from $60,600 to $42,420 => 100%
  • Infrastructure: GitLab.com finishes the quarter on plan, and forecasting being on plan for all of FY21 => 100%
• Key Result: Development: Raise MR rate from 11 to 13 => 110%

Objective (product): Use more of the product and improve site performance => 75%

• Key Result: Dogfooding => 79%
  • Quality: Performance load testing => 100%
  • Quality: Selectively running tests => 95%
  • Infrastructure: Migrate all public dashboards to GitLab monitoring => 60%
  • Support: Dogfood by contributing one TBD feature to ServiceDesk => 50%
  • Security: Secure Scanning => 90%
• Key Result: Improve performance => 70%
  • UX: Create new, prioritized recommendations for Merge Requests => 100%
  • Development: Apply suggestions on improving MR Experience => 80%
  • Development: SpeedIndex <1000 => 30%

Objective (product): Improve efficiency through product documentation => 70%

• Key Result: 50% of Support Engineering solved tickets have docs associated => 97.5%
  • UX: On average, 50% of SE solved tickets have docs associated => 100%
  • Support: On average, 50% of solved tickets by an SE has docs associated => 90%
• Key Result: Support contributes to docs => 93%
  • Support: Deliver 20 docs by June 30 => 100%
  • Support: Deliver 10 docs by July 31 => 90%
  • UX (TW): All new reference architecture topics are live and have setup procedures relevant to their stated user sizes => 89%
• Key Result: Contribute feature to GitLab to support OKRs in GitLab => 20%, determined feature was not needed

Retrospection

Good

• I got to get hands-on with our GDK for the first time in a while
• Our Pass-thru KR experiment seems to have worked and helped lead to focus and high achievement
• Our shared objective experiment seems to have work and led to great collaboration (e.g. UX and Support on docs)
• See linked Department-level OKRs for detailed retrospection on pass-thru OKRs

Bad

• I procrastinated my "add an OKR-related feature to GitLab" until too late in the quarter. I then determined caching would make it problematic to implement. Also Stan highlighted that dual syntaxes for issues / MRs might be confusing to users. But I did determine that including pass-thu OKRs as "related issues" shows their fresh titles, so if the => scores are there, it makes it easy to score. So I'd call the original idea OBE (overtaken by events)
• See linked Department-level OKRs for detailed retrospection on pass-thru OKRs

Try

• Don't procrastinate :)