July 2015 marked the kickoff of Indeed University (IU), our inaugural 12-week summer program to teach Indeed’s development culture to new Indeedians. Over 50 new Indeed software engineers from our Tokyo, San Francisco, Seattle, and Austin offices took part in the program, held at our Austin headquarters.
Indeed University onboarded new hires by acquainting them with Indeed’s culture, technology, and software development philosophy. Our three main goals were:
- Accelerate the onboarding of new college grads
- Provide leadership opportunities for current employees
- Prototype new ideas for Indeed’s business
With these goals in mind, we set out to test a new means of onboarding Indeed’s software engineers.
Onboarding
We like to pair new hires with mentors, who help with anything from dev environment setup to version control to A/B test definition. If a new hire has a question about something beyond their mentor’s expertise, the mentor helps them find the right expert. We encourage every new engineer to get a change into production in their first week, and the mentor supports them as they navigate that learning experience.
From Fall 2014 through Spring 2015, Indeed hired 51 new graduates from Computer Science and Engineering programs in the United States and Japan. With this growth, we could not sustain an onboarding process based on 1:1 mentorship.
But Indeed University did not lose the emphasis on taking ownership and getting code into production. Within the first two weeks of the program, every participant had completed their “gong project,” in which they defined and implemented an A/B test on Indeed. Then they dove into a hands-on curriculum, attended informative talks, and formed teams to build new products.
IUers collaborating in their main work space
Gong project
We take pride in our ability to test everything from simple changes to new experiences. For IU, we embraced this philosophy and offered a blank canvas for our new hires to implement an A/B test in their first week.
The A/B tests aimed to help job seekers find jobs faster by emphasizing helpful filters, such as salary.
Screenshot of an Indeed University A/B test that encouraged job seekers to filter by salary
It’s an Indeed tradition for an engineer to ring the gong after their first line of code goes live on Indeed. At IU, everyone gathered around when a test was live. The IU participant (IUer) explained what A/B test they implemented and then rang the gong. Usually with great gusto.
Ringing the gong
Designing and implementing these tests provided an opportunity for these new Indeedians to bring fresh perspectives to Indeed’s core products. What better way for a new hire to learn about data-driven development than to form a hypothesis, design a test, and analyze the results? IUers got a first-hand experience with our culture of data-driven development. It was exciting to see new ideas come out of these A/B tests.
Curriculum
After completing the gong project, IUers launched into the curriculum – a self-paced course designed by seasoned Indeed engineers. Exercises followed each section, making the curriculum more hands-on.
The curriculum began with engineering basics and covered the tools and technologies we use at Indeed to build webapps and services. Then it covered logging (using logrepo) and data analysis (using Imhotep).
Talks
IU talks introduced our product development philosophy and some interesting market opportunities. We wanted participants to think about new products Indeed could build.
Paul D’Arcy presents on How People Look for Jobs (Photo: Hannah Spellman)
Tech talks and product deep dives provided more specific details about infrastructure and technology. Unlike typical startups, the IU teams had immediate access to Indeed data and users, as well as tools for deployment, testing, and authentication.
Social events
Social activities helped IUers get to know one another. They developed lasting connections before many of them headed off to Seattle, San Francisco, and Tokyo.
We planned Friday happy hours that were complete with freshly baked cookies from Tiff’s Treats, local brews, a slew of board games, and Wii U. We also took our new hires to our favorite local restaurants, including Franklin Barbecue, Torchy’s Tacos, and Michi Ramen. We went to a Round Rock Express baseball game. We also did standup paddleboarding, go-karting, and paintball. Last, but not least, we visited the Austin Panic Room.
Standup Paddle Board on Town Lake in Austin
Enjoying Franklin Barbecue
Developing new leaders
Twelve emerging leaders from the engineering, product, and online marketing organizations joined Indeed University to teach best practices and our style of iterative, data-driven development. These leads gained experience by directly managing 4-5 new hires. They also each advised up to 3 product teams.
The experience gave the leads their first taste of engineering management, including weekly 1-on-1s and quarterly evaluations. The 1:4 ratio allowed leads to develop relationships with the new hires. IUers talked with their leads about product and technology challenges as well as more mundane concerns like finding an apartment.
As product team advisors, leads challenged themselves to teach rather than tell. They encouraged teams to plan product iterations, prioritize issues, design tests, and analyze results.
In other words, leads taught participants how to think like Indeed engineers. They encouraged teams to be independent but unafraid to ask questions, to take risks and use data to measure outcomes, and to take ownership of their products.
New products
We encourage all engineers to imagine product changes that focus on our mission: helping people get jobs. IUers brainstormed new product ideas for Indeed. We held three brainstorming sessions. After each session, groups pitched solutions to problems they had identified.
Brainstorming wrapped up with a final round of pitches to Indeed’s senior leadership team, Indeed University leads, and other interested Indeedians.
Brainstorming in the IU lounge
Following brainstorming, IUers formed teams based on their interests and spent the remaining 9 weeks building working products. They served as their own product managers, team leads, designers, marketers, and testers.
At Indeed, we believe we must explore as many product ideas as we can, as quickly as we can. IU immersed participants in this culture of engineering velocity.
Before product development began, every team researched their market. Teams created Google Surveys, called Indeed employers, and ventured to local shopping malls to speak with retail managers and employees directly.
These conversations challenged some assumptions, helped drive initial product direction, and caused one early product pivot.
We challenged each team to build the minimal viable product (MVP) that would allow them to validate their idea. How would they demonstrate value? One word: data. Teams needed to collect data in order to confirm their product’s value.
Indeed University teams had a few “unfair advantages.” For one, they had Indeed’s data at their fingertips. They could use this data to do preliminary research or to tailor their design to a specific audience. Second, each team had a generous online marketing budget for traffic acquisition. Teams tested showing ads on Google, Facebook, and Indeed job search, to see where they could best connect with potential users.
Teams presented their work at weekly product reviews with Indeed executives and IU leads. In these weekly reviews, teams answered the following questions:
- What did you do this week?
- What data did you collect?
- What did you learn from this data?
- What are you doing based on the data?
In these meetings, we encouraged teams to start small, validate with data, and iterate. They learned that building successful software is about more than software architecture.
The product reviews encouraged discussion of A/B test results and opportunities for testing assumptions. Many teams experienced high bounce rates on their landing pages, and they ran A/B tests to test their hypotheses about these bouncing users. One team tested a new call to action on their landing page. Another tested delaying the sign-in requirement, allowing job seekers to experience the product before creating an account.
In the course of three months of Indeed University, the new hires built eleven new products:
- A search engine for college students to find jobs
- A data trends exploration tool for HR professionals
- A site that helps high school students choose a college major based on jobs that interest them
- A product that uses social sentiment analysis to give job seekers unfiltered comments about company reputation
- A hiring tool that helps employers manage active candidate statuses
- A product that asks job seekers a series of questions in order to match them to jobs based on their interests
- An application that allows employees to track progress toward their goals
- A mobile app for finding local retail and food service jobs
- An automated phone screening solution to help employers efficiently evaluate candidates
- A gig marketplace for job seekers to find small jobs in their area
- A site that lets people visually explore career paths based on a current or desired position
Graduation
At the Indeed University Graduation Party, each team introduced the product they built and presented their learnings. Five products continued beyond IU for product validation through further development, testing, and iteration.
An IU team talks about what they learned at the Graduation Party
After IU, the participants joined their new teams in Tokyo, Seattle, San Francisco, and Austin. They take with them the connections, skills, and knowledge they gained during the program. They are on their way to having a real impact at Indeed.
Whether you are a student or an industry veteran, we are looking for talented engineers to help with our mission of helping people get jobs. To learn more, check out our open positions.