Vertabelo helps Uber employees get the most out of their data

The Challenge

How can you teach users to identify trends and pull data themselves?

Thousands of Uber employees use data in their daily work. Operation teams, as well as marketing and promotions, need it for planning and decision-making processes.

Data is at the heart of the decision making process at Uber. People across various roles and locales need access to accurate and reliable business metrics.
Lyle Adams, Analytics Manager

Even though there are internal Business Intelligence tools available, the most up-to-date, precise, and granular data can be accessed programmatically. Uber’s Data team was flooded with requests from various organizations and individuals on the ways to access the relevant datasets.

Uber tried educating users through video conferences, learning groups, and how-to guides. However, these solutions were insufficient on their own. Uber needed a way to educate its global workforce at scale.

Road to the Solution

Rise and fall of the traditional learning process

The traditional approach with lectures and presentations proved to be ineffective. SQL training became an obligatory part of onboarding, but with the number of employees growing exponentially, it was taking more and more time.

PowerPoint presentations and in person sessions didn’t scale since it would take months to educate our global operations team. We needed a solution that allowed us to reach as many users as possible and educate users at their own pace.
Lyle Adams, Analytics Manager

Also, Uber’s internal data architecture was constantly evolving. Continuous revisions and updates were costly and time-consuming; thus, central coordination of the SQL education process in several countries became almost impossible.

Uber wanted an e-learning solution that could scale with its business. To that end, it contacted the Vertabelo Team in hopes of a potential partnership.

The Solution

Tailored e-learning platform based on Vertabelo Academy

To meet Uber’s needs, Vertabelo created a dedicated e-learning platform based on its cutting-edge Vertabelo Academy. Starting with an SQL Basics course, participants gain knowledge and SQL skills and can immediately test them on a specially designed dataset that resembles Uber’s actual databases.

All of the course tasks are strictly connected to Uber’s domain and educate users on how to identify specific Uber metrics using SQL. Vertabelo and Uber partnered to create a completely custom course that teaches users the basics of SQL syntax while giving them an overview of important business metrics. Thanks to this course, employees can learn the specifics of Uber’s databases with their actual structure, dictionaries, etc. –always in their most current version.

Teaching SQL using daily problems and real-life examples improved user engagement, which resulted in higher user retention and more impactful training.

Data consumers at Uber face not only challenges of being able to query, but also understanding how to use our many databases. With the course created around Uber specific tables, even users who are savvy with coding gain insight in the different ways the key tables work with each other. For those who are learning SQL, they get a hands-on experience in our database without the intimidation of our database.
Mallory Wang, Data Analyst

Moreover, Vertabelo Academy allows central management of SQL training, which solves many problems with coordination and updating the course content in different countries. The e-learning platform is available 24/7 regardless of time zones. It also allows Uber to check the progress of all course participants and help them if assistance is needed.

Results

The SQL introductory course was just the beginning

Learning SQL and business domain at the same time

Solving daily problems for people across various roles

Worldwide training management made easy

We compared total execution for data at Uber during a 2-week period for new employees who finished the introductory course with total executions during a comparable 2-week period for new employees who never took the course. The data shows that those who completed the course queried 20 times more than those who didn’t take the course. This means that the SQL Learning Platform gives users the confidence to query data, which is certainly the first step in using data at Uber.

Further leveraging this powerful tool, future growth includes creating courses for other parts of Uber’s data, such as Eats, Driver Operations, and more.
Lyle Adams, Analytics Manager