GitLab Data follows a program to help align all teams towards the results that will help GitLab grow.
Watch our Planning Drumbeat Walkthrough Video on GitLab Unfiltered.
The Data Team Planning Drumbeat is a pre-set sequence that happens every quarter. The Planning Drumbeat adheres to GitLab's financial year/quarter schedule and Quarterly Objective and Key Results for prioritization and alignment. The Data Team Planning calendar is as follows:
Milestone naming convention; Milestone FYxx-Qxx-MSxx. I.e. FY22-Q02-MS01
LT = Leadership Team (
DT = Data team (Product Fusion Team, GTM Fusion Team or Data Platform Team)
The overall ambition is to structure why, when and what we are doing.
This top-down approach is leading, but not fully limited to the way we work.
The goal of the Planning Drumbeat is to improve our ability to plan and estimate work through better understanding of our velocity. During the milestone planning process, we assign issue points, which are a good measure of consistency, as milestone over milestone should share an average. Then we pull into the milestone based on when we think we can commit to the work being completed. Then issues are prioritised according to our standing priorities.
This approach has many benefits, including:
The Planning drumbeat and the results it helps us deliver is a collaborative team-effort
The data team works in two-week intervals, called milestones. 1 milestone per quarter will take three weeks long, because a quarter contains 13 weeks. The milestone that takes 3 weeks is the milestone that covers a major holiday or if the majority of the team is on vacation or at Contribute. Milestones start on Wednesdays and end on Tuesdays. This discourages last-minute merging on Fridays and allows the team to have milestone planning meetings at the top of the milestone.
Milestone planning should take into consideration:
The timeline for milestone planning is as follows:
|Day||Current Milestone||Next Milestone|
|0 - 1st Wednesday||Milestone Start
|7 - 1st Tuesday||Midpoint
Any issues that are at risk of slipping from the milestone must be raised by the assignee
|10 - 2nd Friday||The last day to submit MRs for review
MRs must include documentation and testing to be ready to merge
No MRs are to be merged on Fridays, or on Thursday in the case of Family and Friends Day.
|Milestone is roughly final
Milestone Planner verifies issue priority and team capacity for next milestone.
|13 - 2nd Monday||Last day of Milestone
Ready MRs can be merged
|14 - 2nd Tuesday||Meeting Day
All unfinished issues either need to be removed from milestones or rolled to the next
Sync-meeting to perform retro perspective on the current milestone and align/start on the next milestone according to the created milestone planning. All unfinished issues either need to be removed from milestones or rolled to the next
Data Team OKRs aspire to align with Business Technology OKRs, Finance Division OKRs, and CEO OKRs, thereby aligning with the OKRs of the Divisions we support. The Data Team also creates OKRs for Data Platform infrastructure development and these OKRs may not always map to immediate term Business Partner OKRs. Overall, OKRs constitute 60-75% of the Data Team's Quarterly Capacity, with Production Operations as the only established higher priority.
Data Team OKRs are managed with GitLab Plan using Epics and Issues using a standard naming scheme and structure and this approach:
After OKRs have been drafted and shared across the Data Team, they are shared with Business Partner stakeholders for feeedback. Once the draft process is complete, three summarized Objectives and KRs are published into the Business Technology Group Project for sharing with the broader Business Technology Team, Finance Team, and company. These summarized Objectives and KRs attempt to capture the top OKRs in support of business results.
Data Team Objectives, Newsletters, and Report CardsEpic.
DRAFT: FYxx-Qx Data Team Objectives and Planning Drumbeat
DRAFT: FY23-Q1 Data Team Objectives and Planning Drumbeat
FYxx-Qx Data Objective x: <name of objective>
FY23-Q1 Data Objective 1: Accelerate Growth and Product with Usage Analytics and Tooling
FYxx-Qx Data Objective X, KR Y: <name of key result>
FY23-Q1 Data Objective 1, KR 1: Deliver a New Adoption Performance Dashboard to Product with 10 Users Trained
During the OKR Review, the Data Leadership Team reviews the current status of the current FQ KRs. The results of the OKRs are evaluated and propagated to next FQ OKRs where applicable.
We use a T-Shirt sizing approach for quickly sizing the work required to deliver new issues, epics, and longer-term initiatives. The T-Shirt sizing approach is designed to support work breakdowns towards the creation of detailed work plans, but also intended to provide a sufficient level of detail to manage overall scope.
|Size||Dedicated Person Time||Weight (issue points)||Examples|
|XS||< 1/2 Day||1||Update existing handbook page. #data research/response. New Trusted Data Test.|
|S||< 1 Day||2-3||New handbook page; typical triage issue; New dashboard on top of existing models.|
|M||1 Week||5-8||New dashboard requiring new models. New data source with Stitch or Fivetran.|
|L||2-3 Weeks||13||New fact table implementation & testing. Full XMAU solution.|
|XL||1-2 Months||26||New Data Pump to new system. New Data Source with complex source API.|
|XXL||2-4 Months||52+||New Dimensional Model subject area with New Data Sources.|
Work breakdowns are always developed as a part of the Quarterly OKR Planning Drumbeat, but can also be leveraged to help scope and plan new initiatives, infrastructure projects, and similar multi-person or multi-week projects. The outcome of the work breakdown is a detailed description of the work to be performed, deliverables and responsibilities, and a high-level timeline.
Work Breakdowns consider the following inputs:
The work breakdown is team effort and everyone is encouraged to contribute.