Self-Service Data Team at GitLab

The Self-Service Data Team is responsible for leveraging data to optimize for the self-service customer experience and drive nARR growth via sales efficiency. Data insights from this team feed: sales visibility, self-service fulfillment features, and growth/marketing experiments. The Self-Service Data Team also aims to create data tools to help with efficiency, prioritization, and decision making.

Welcome to the Self-Service Data Team Handbook

Team

Name GitLab Handle Title
Max Fleisher @mfleisher Mgr, Self-Service & Online Sales Data
Sara Gladchun @sglad Sr. Analyst, Self-Service & Online Sales Data
Ed Lu @eclu94 Sr. Analyst, Global Digital SMB Operations

Who We Work With

Self Service Team - We partner with the overall Self Service team to provide data insight around the self service customers (currently AMER SMB FO and Pooled accounts) to drive nARR and growth through sales efficiency and strategy.

Central Data Team - We work with the central data team by staying involved in cross functional data initiatives, collaborating where possible, and providing feedback on data models and the data that live in Snowflake. We also work with the data science team by staying up to date on their projects and models and incorporate many of their predictive outputs into our analyses and triggers,

Product Analytics - We work with product analytics by staying up to date on what major projects they are working on and by leveraging many of their models in our own data work.

Fulfillment - We work with fulfillment to provide data around self service fulfillment features and feature requests.

Sales - We work with the Low Touch Sales team to provide data insights, data tools, and sales visibility to the AEs to increase efficiency and make the most up to date data available for quick response times and targeted outreach. We also provide forecasting models for the Low Touch sales teams (FO and Pooled teams).

Marketing Analytics - We partner with Marketing Analytics to provide data around FO Funnels as well as targeted digital outreach to the Pooled Account customers.

Resources

General

Resource About
Data Request Issue Template Template that should be used for ad-hoc data questions and requests
Data Hub All of our data assets and resources in one place

OKRs

Quarterly Work: Weekly Progress

Quarter Max Sara
FY23-Q2 Issue Issue
FY23-Q3 Issue Issue
FY23-Q4 Issue Issue
FY24-Q1 Issue Issue
FY24-Q2 Issue Issue
FY24-Q3 Issue Issue

Working with us

Purpose: Outline how the broader Self-Service team can engage the Self-Service Data Squad (Max, Sara)

Goal: Minimize dependencies/blockers to insights while providing transparent engagement model

Disclaimer: Not all data questions will be able to be answered. Ultimately, taking time to answer ad-hoc questions means less time on projects (aka the zero-sum capacity problem). That is not to say that ad-hoc questions are not important; however, we do have “boulder” level projects in flight have been prioritized via the OKR process, which we also need to make progress on.

How to submit your ad-hoc data request or question:

  1. Have you tried to answer this question leveraging existing resources (e.g. data hub, SFDC)?

    • If no, please try to answer your question using these existing resources.
    • If yes, but you’re still unable to answer your question, go to question 2.
  2. Using the Data Question Intake issue template in our project, please:

  • Fill out all items under the “Filled out by Requestor” section
  • Add the “Self-Service Data” and “Self-Service Data Ad Hoc” labels
  • If business stopping: tag Max in Slack (ideally in self-service_public) with link to issue.

Data Definitions

Purpose: To ensure that we are all speaking the same data language, we have created clear metrics that align with our agreed business definitions.

General Definitions:

  • Low Touch - Consists of any AMER SMB FO or Pooled Account opportunity
  • Pooled Model - handbook page
  • Pooled Cases - SFDC cases that are created for pooled accounts (Pooled cases will have record_type_id = ‘0128X000001pPRkQAM’)
  • Tasks - SFDC tasks
  • WW FO Count = All New - First Order deals
  • WW Web Growth nARR = Any non-First Order nARR transacted through Web (Web Portal Purchase = T)
  • LT all-in nARR = Any nARR owned by the FO and Pooled AEs
  • LT FO Count = Number of First Orders owned by AMER SMB FO AEs (not limited to Pooled accts)
  • Web Growth = All non-FO transactions through Web Portal
  • LT all-in = Everything closed by the AMER SMB FO and Pooled AE teams

Retention, Renewal, and Churn Definitions:

  • NET_RETENTION = ARR in the Pooled account set in a given month / ARR in that account set 1 year prior
  • NET_LOGO_RETENTION = What % of account set from a year prior are still customers
  • Pooled Renewal and Churn Rates: Shows the components of nARR renewal outcomes, either Uplift, Contraction, or Churn, unit is % of ATR ARR monthly/quarterly
  • Pooled Renewal and Churn nARR: Same as above but the actual nARR totals
  • Pooled Acct Set Growth Rates: Shows how the overall Pooled account CARR changes monthly/quarterly, broken in the TRX type components (“Growth” excludes Renewals)
  • Pooled Acct Set Growth nARR: Same as above but the actual nARR totals

SSOT Queries

SSOT data is necessary in order to have confidence in our metrics, have repeatable and replicable reporting, and for our data team to work more efficiently. We have created a GitLab repo to house our SSOT SQL queries for both our foundational base queries and for ad hoc analyses.

This allows us to keep a record of queries used for foundational projects like our dashboards and for one off analyses that may need to be repeated, tweaked, or modified in the future. Dashboard queries are also housed in Sisense as snippets in order for the data team to work more efficiently within the BI tool. We are currently updating these queries to work within Tableau as well.

The current workflow for creating or updating snippets and SSOT queries is the following:

  1. Create or update the query and ensure it produces the desired results and accurate data
  2. Update the query in the SSOT Queries directory and commit to a new branch
  3. Create an MR and tag another Data Team member as a reviewer
  4. The reviewer will review the code and merge the MR to update the query
  5. Once MR is approved and merged, the original author will either update the snippet in Sisense (original author must also update the date in the comments with the most recent date updated)

Tableau Data Source Workflow (WIP - New Data Source):

  1. Create query using Snowflake worksheet to produce desired output
  2. Create new Data Source in Tableau Desktop from Snowflake
  3. Copy query into SSOT Repo to save work and track changes
  4. Copy query as Custom SQL into Tableau and set to Extract
  5. Run Extract and check data integrity and accuracy
  6. Commit any necessary changes to the Repo
  7. Once the query is finalized: - Publish Data Source to Server - CRITICAL - Data Source must be published to the Development -> Sales -> SAFE folder - Embed Snowflake credentials to allow for refresh

Updating a published Tableau Data Source:

  1. Log into Tableau Server
  2. Locate the Data Source in My Content or in the folder above
  3. Select Edit Data Source
  4. Use Snowflake worksheet to test any changes to the stored query
  5. Commit changes to the Repo
  6. Copy new query into the Data Source Custom SQL
  7. Run the extract
  8. Publish the updated Data Source

Current SSOT Queries (Updated Quarterly)

  • ARR Mart Price Quantity
  • Churn Renewal Rates (Low Touch)
  • Churn Renewal Rates (Monthly)
  • Facts from Price Points
  • FO ASP Price Points
  • FO Bucket (Low Touch)
  • Monthly Subscription ARR snapshot (Low Touch)
  • Opportunities with Prior Year
  • Pooled Account Snapshot
  • Pooled Case Triggers
  • Pooled Cases
  • Pooled Overage List
  • Renewal Price Points
  • Retention Rates (Low Touch)
  • Self Service Usage (Low Touch)
  • Touch Level Data on Opportunities (WIP)
  • TO BE ADDED: Tableau Queries

Ad Hoc Analysis (updated Quarterly)

  • Account Tiering (deprecated)
  • Credit Card Failure and Payment Method Analysis
  • Case Creation Automation Queries
Last modified December 6, 2024: Update file self-service-data.md (1d1a712e)