Data Team

The GitLab Enterprise Data Team is responsible for empowering every GitLab team member to contribute to the data program and generate business value from our data assets.

Welcome to the Data Team Handbook

Our Vision is to Contribute to GitLab’s journey of becoming the leading AllOps platform by responsibly harnessing the power of data.
In pursuit of our vision, we will focus on 4 outcomes:

Drive company results by building trusted, reliable, and innovative data products and insights when and where needed.
Minimize time from question to insight to action, enabling team members to move faster by implementing efficient processes and enabling self-service analytics.
Develop and secure our data into a uniform, trusted asset through data protection & privacy, iterating on processes, people, and platforms.
Enable every team member to contribute to initiatives responsibly and with trust, building a powerful data-driven culture.

Read our Direction page to learn what we are doing to improve data at GitLab.
Our Principles inform how we accomplish our mission.
Watch our Data Recruiting Video to learn about the growing Data Program.

Would you like to contribute? Recommend an improvement, visit Slack #data, watch a Data Team video. We want to hear from you!

How Data Works at GitLab

The collective set of people, projects, and initiatives focused on advancing the state of data at GitLab is called the GitLab Data Program. GitLab has two primary distinct groups within the Data Program who use data to drive insights and business decisions. These groups are complementary to one another and are focused on specific areas to drive a deeper understanding of trends in the business. The two teams are the (central) Enterprise Data Team and, separately, Function Analytics Teams located in Sales, Marketing, Product, Engineering or Finance. Watch the Data Recruiting Video to hear from some of the teams involved and what they are working on.

The Data Team reports into Business Technology and is the Center of Excellence for enterprise insights & analytics (not operational), data science, data platform & infrastructure, BI technologies, master data, data governance and data quality. The Data Team is also responsible for the enterprise data strategy, building enterprise-wide data models, providing Self-Service Data capabilities, maintaining the data platform, developing Data Pumps, and monitoring and measuring Data Quality. The Data Team is responsible for data that is defined and accessed on a regular basis by GitLab team members from the Snowflake Enterprise Data Warehouse. The Data Team builds data infrastructure to power approximately 80% of the data that is accessed on a regular basis. The Data Team also provides a Data Science center of excellence to launch new advanced analytics initiatives and provide guidance to other GitLab team members.
Function Analytics Teams reside and report into their respective divisions and departments. These teams perform specific analysis for business activities and workflows that take place within the function. These teams perform ad-hoc analysis and develop dashboards based on the urgency and importance of the analysis required, following the Data Development approach. The most important and repeatable analysis will be powered by the centralized Trusted Data Model managed by the central Data Team. Function Analytics Teams also build function-specific/ad-hoc data models and business insights models to solve for urgent and operational needs, not requiring trusted data features. Function Analytics Teams work closely with the Data Team in a variety of ways: expand GitLab’s overall analytics capabilities, extend the Data Catalog, provide requirements for new Trusted Data models and dashboards, validate metrics, and help drive prioritization of work asked of the Data Team. When data gaps are found in our business processes and source systems, the team members will provide requirements to product management, sales ops, marketing ops, and others to ensure the source systems capture correct data.

Data Program Teams

The GitLab Data Program includes teams focused in the following areas:

How Data Teams Work Together

On a normal operational basis, the Data Team and Function Analyst teams work in a “Hub & Spoke” model, with the Data Team serving as the “Hub” and Center of Excellence for analytics, analytics technology, operations, and infrastructure, while the “Spokes” represent each Division or Departments Function analysts. Function analysts develop deep subject matter expertise in their specific area and leverage the Data Team when needed. From time to time, the Data Team provides limited development support for GitLab Departments that do not yet have dedicated Function Analysts or those teams which do have dedicated Function Analysts, but might need additional support. The teams collaborate through Slack Data Channels, the GitLab Data Project, and ad-hoc meetings.

classDiagram
    Sales <|-- Data
        Sales : +Sales Analyst
    Marketing <|-- Data
        Marketing : +Marketing Analyst
    Product <|-- Data
        Product : +Product Analyst
    Engineering <|-- Data
        Engineering : +Operations Analyst
        Engineering : +Infrastructure Analyst
    Finance <|-- Data
        Finance : +Financial Analyst
    People <|-- Data
        People : +People Analyst
    Data : +Data Analyst
    Data : +Analytics Engineer
    Data : +Data Engineer
    Data : +Data Scientist

The Data Platform & Architecture Team

The Data Platform Team & Architecture Team is part of the Enterprise Data Team and focuses on building and maintaing secure, efficent, and reliable data systems data infrastructure. The Data Platform & Architecture Team is both a development team and an operations/site reliability team. The team supports all Data Pods with available, reliable, and scalable data compute, processing, and storage. Platform components include the Data Warehouse, New Data Sources, Data Pumps, Data Security, and related new data technology. The Data Platform team also drives the Data Management processes. The Data Platform Team is composed of Data Engineers.

Analytics Engineering Team

The Analytics Engineering Team** transforms raw data into clean, structure and usable formats for data decision-making. The Analytics Engineering team also drives Enterprise Data Program and supports the wider data community. The team focuses on inventorying, integrating, maintaining, and governing the data at an Enterprise level. This includes collaborating with the business units and data teams in establishing and facilitating commonly accepted guidelines around Enterprise data along with building enterprise-wide data models, supporting Self-Service BI and Analytical capabilities by providing Data Enablement and required training to the Users on Enterprise Data Models.

The Enterprise Insights & Data Science Team

The Enterprise Insights & Data Science Team utilize analytics and Machine Learning (ML) for insights into customer behavior and company performance. The Enterprise Insights & Data Science team focuses on delivering a complete view of the customer (Customer 360), predict customers that are likely to buy, expand or churn, develop models to predit the long-term value of customers, create detailed customer profiles, and deliver insights on company performance. The Team acts as a Center of Excellence for predictive analytics and supports other teams in their data science endeavours by developing tooling, processes, and best practices for data science and machine learning. List of the current projects can be found in the Data Science handbook page.

Data Job Families

The job families are designed to support all of the routine activities expected of a Data Team. In FY22 we are introducing two new job families, Data Scientist and Analytics Engineer.

How We Measure Impact

Our impact will be measured against 4 dimensions (these metrics will adjust as our data maturity increases and our focus areas change):

Data Platform Stability

Infrastructure Cost vs Plan: This performance indicator tracks the financial position of the actual cost vs the planned costs for the data infrastructure (warehouse, ETL pipelines, etc.).
Data Uptime: This performance indicator measures the % of time a data pipeline was providing data without reported incidents. This indicator is currently measured based on Monte-Carlo data, according to the configured (automatic) monitors on any given table in the raw data layer.

Data Quality & Governance

% completion of the data validation and data cleansing roadmap
Governance & Quality data assessment scores

Data Adoption

Data Monthly Active Users (DMAU): DMAU Measures the direct usage of the Data Platform by GitLab Team Members based on usage of the primary analysis tools we provide: Snowflake, Tableau, and Sisense. Over time we will include additional tools such as Jupyter and Data Studio, as well as usage of data pumped into EApps such as Marketo (PQLs), Gainsight (Usage Data), and Salesforce (Propensity Scores). The DMAU worksheet stores the code and historical stats and a visualization of these numbers can be found in the Data Monthly Active Users report.
- Data Monthly Active Users (DMAU) = Unique Sisense Users + Unique Snowflake Users + Unique Tableau Users in a given month
- Quarterly Data Monthly Active Users (Q-DMAU) = Unique Sisense Users + Unique Snowflake Users + Unique Tableau Users across all months in a quarter OR sum(months in quarter)/
- Note: Users of Sisense, Tableau and Snowflake might be double counted if they access multiple systems. We do not count distinct users across the tools.
Tableau adoption health: % of licensed Tableau users still leveraging Sisense
Data Maturity Score: measured annually, evaluates our current data maturity against 8 data capabilities: 1. Strategy & Approach 2. Culture & leadership 3. Metrics & KPIs 4. Organization & Skills 5. Architecture & Integration 6. Governance & Quality 7. Deployment & Usage 8. Technology & Operations
Number of certified Tableau dashboards
% views of certified dashboards

Revenue/Efficiency Impact

First we have the evaluation criteria known as Dollar Value of our Results as calculated by the Data Value Calculator. We can use the Data Team Value Calculator to calculate the dollar value of the initiatives we contribute to and the issues we complete. Additionally we want to shift to a more aspirational measurement which is to measure the ARR impact or efficiency gain from each of our data products. Our data science models will be measured in the following ways:

Propensity to Expand (PtE) and Purchase (PtP) - We will evaluate two metrics: 1) Incremental revenue impact 2) # of leads generated that are not currently in the sales funnel
Propensity to Churn (PtC) - We will evaluate two metrics: 1) # of high propensity to churn customers that didn’t churn 2) Incremental revenue impact

How To Connect With Us

Primary #Data Slack Channel

Issue tracker GitLab Unfiltered Data Team Playlist What time is it for folks on the data team?

Data Slack Channels

#data is the primary channel for all of GitLab’s data and analysis conversations. This is where folks from other teams can link to their issues, ask for help, direction, and get general feedback from members of the Data Team.
#data-daily is where the Data Team tracks day-to-day productivity, blockers, and fun. Powered by Geekbot, it’s our asynchronous version of a daily stand-up, and helps keep everyone on the Data Team aligned and informed.
#data-lounge is for links to interesting articles, podcasts, blog posts, etc. A good space for casual data conversations that don’t necessarily relate to GitLab. Also used for intrateam discussion for the Data Team.
#data-engineering is where the GitLab Data Platform team collaborates.
#bt-data-science is where the GitLab Data Science team collaborates.
#business-technology is where the Data Team coordinates with Business Technology in order to support scaling, and where all Business Technology-related conversations occur.
#analytics-pipelines is where slack logs for dbt runs and monte carlo analysis are output and is for analytics engineers to maintain. The DRI for tracking and triaging issues from this channel is shown here.
#data-triage is an activity feed of opened and closed issues and MR in the data team project.
#data-pipelines is where alerts from the ELT pipelines / FiveTran/ Monte Carlo RAW layer anomalies published and is for data engineers to maintain. The DRI for tracking and triaging issues from this channel is shown here.

You can also tag subsets of the Data Team using:

@datateam - this notifies the entire Data Team
@data-engineers - this notifies just the Data Engineers
@data-analysts - this notifies just the Data Analysts
@analytics-engineers - this notifies just the Analytics Engineers

Except for rare cases, conversations with folks from other teams should take place in #data, and possibly the fusion team channels when appropriate. Posts to other channels that go against this guidance should be responded to with a redirection to the #data channel, and a link to this handbook section to make it clear what the different channels are for.

GitLab Groups and Projects

The Data Team primarily uses these groups and projects on GitLab:

GitLab Data is the main group for the GitLab Data Team.
GitLab Data Team is the primary project for the GitLab Data Team.

You can tag the Data Team in GitLab using:

@gitlab-data - this notifies the entire Data Team
@gitlab-data/engineers - this notifies just the Data Engineers
@gitlab-data/analysts - this notifies just the Data Analysts

Team, Operations, and Technical Guides

TECH GUIDES	INFRASTRUCTURE	DATA TEAM
SQL Style Guide	High Level Diagram	How We Work
dbt Guide	System Data Flows	Team Organization
Python Guide	Data Sources	Calendar
Airflow & Kubernetes	Snowplow	Triage
Docker	Permifrost	Merge Requests
Data CI Jobs	DataSiren	Planning Drumbeat
Rstudio Guide	Trusted Data	Data Science Team
Jupyter Guide		Data Management
Meltano Guide
Experimentation Best Practices
Data Onboarding
Learning Library
Tableau Guide
Tableau Style Guide

Data Team Handbook Structure

Data Catalog

The Data Catalog page indexes Analytics Dashboards, Workflows, and Terms.

Data Development

This page defines the Data Development lifecycle

Data Quality

MVC for a Data Quality Program at GitLab

Data Team - How We Work

GitLab Data Team Workflow

Data Team Data Management Page

The Data Management Page covers the content around managing, securing, and governing the Enterprise Data Platform and related activities.

Data Team Direction

This page contains forward-looking content and may not accurately reflect current-state or planned feature sets or capabilities. Strategy As an important step towards achieving our mission, meeting our responsibilities, and helping GitLab become a successful public company, we are creating an Enterprise Data Platform (EDP), a single unified data and analytics stack, along with a broad suite of Data Programs such as Self-Serve Data and Data Quality. The EDP will power GitLab’s KPIs, cross-functional reporting and analysis, and in general, allow all team members to make better decisions with trusted data.

View page source - Edit this page - please contribute.