How Booz Allen Hamilton is Winning the War on Talent
At Booz Allen Hamilton, a management and technology consulting firm, we help our clients innovatively harness big data to drive their business forward. In recent years, our world has experienced an explosion of data. 2.5 quintillion data points are created daily, and 90% of the world’s data has been created in the last two years. The rise of algorithms and artificial intelligence have contributed to an increase in big data. This has, in turn, resulted in the whole new field of data science to create, analyze, and process data as well as draw actionable insights from it. In a nutshell, data science is the art of turning data into actions for your business.
The Booz Allen Hamilton Data Science 5K Challenge
At Booz Allen Hamilton, we strive to be game-changers in the data science field. We want to innovate and change the conversation around data to help our clients use data in ways they’ve never used it before. That’s why we set a goal to employ 5,000 data scientists. We saw this as a challenge to both hire and train our existing analysts to be data scientists.
We’re located just outside Washington, DC, where there’s a huge demand for this skill set. Looking at the overall demand and supply of data scientists, we saw that the “build and buy” combination was necessary. In other words, we couldn’t just rely on hiring data scientists externally, we had to also train people with data science skills internally. To meet this goal, our learning & development (L&D) team set out to create a personalized learning program at scale for our globally distributed firm.
I recently covered this topic in the “Winning the War on Talent: Scaling Personalized Learning” webinar hosted by Udemy for Business and Degreed. In this post, I’ll be sharing some highlights from the webinar. If you’d like to watch the entire presentation, you can access the on-demand recording here.
With a lean L&D team, we had to think carefully about how we designed every aspect of the program and still provide a great learner experience. We saw our roles as “learning experience architects.” We wanted to build the right ecosystem using curated content, technologies, and platforms to offer a great learner experience.
To equip our data scientists with the right skill set, we first piloted and then scaled our program from 125 to 1200 employees. In L&D, we often are constrained by limited resources and we need to think in terms of operational excellence. Scaling learning is at the core of achieving this goal. Here are 4 ways we scaled personalized learning at Booz Allen Hamilton to meet our data science talent challenge.
4 ways Booz Allen Hamilton scaled personalized learning
1. Online assessments to tailor learning
Our data science curriculum is rigorous and includes topics such as Python, developer tools, and advanced mathematics. To bring everyone up to the same proficiency level, we offered an initial voluntary skills assessment. Based on how an employee performed on the assessment, we assigned pre-work online courses on Udemy for Business. We also made the completion of these online courses mandatory in order to join the 60-hour course.
2. Personalized online learning pathways
Next, we worked with our subject matter experts to curate personalized online learning pathways. We use Degreed as our learning experience platform and Udemy for Business as our content library. We assigned learning paths consisting of Udemy for Business online courses to help build key data science skills. Our online learning paths are used by more experienced analysts who are interested in joining the data science cohort as well as more junior employees who want to start building their skills. Our employees also like to discover courses on their own and often take advantage of Udemy for Business’ personalized course recommendations. Find out more about how you can create personalized learning paths for your employees on Udemy for Business.
Personalize learning at scale
Find out how you can personalize learning at scale with Udemy for Business.
3. Blended learning model focuses on hands-on projects in the classroom
Our blended learning model means that learners have the opportunity to engage with online courses as well as work on hands-on projects in the classroom. Throughout the course, mini-projects give them the opportunity to practice new skills. The program culminates in a capstone project. This final capstone project allows learners to demonstrate mastery to both classmates and the leadership team through in-person presentations. For example, employees might pull open-source data sets and prove a hypothesis — demonstrating they can use data in a live setting.
Having the leadership team observe these presentations connects them to employees as well as highlights the work of our L&D team. When leaders see employees in action showing off their new skills, it has amazing results.
4. Mentor circles guide the learning journey
To guide employees throughout the learning process, mentors hold learners accountable while also acting as coaches to support them. Mentors help employees navigate the challenging content as well as offer advice on career development at Booz Allen Hamilton. In order to make the mentorship program scalable, we use Zoom and Slack to connect mentors and mentees. We also leverage our program’s alumni to serve as mentors for newer cohorts of learners. These mentor circles are one of the highlights for our employees.
Proving how we’re moving the needle
Now the program is underway, we’re looking at a number of factors to see how we’re moving the needle. We’re considering both leading indicators (e.g. engagement and consumption) as well as lagging indicators (e.g. impact to the business).
For leading indicators, we’re tracking learning consumption in the form of course enrollments and completions via Udemy for Business. To determine which marketing channels are effective, we also create trackable links and embed them in our marketing newsletters on Degreed.
For lagging indicators, we have a number of Key Performance Indicators (KPIs) that we’re following. We want to measure success beyond simply hitting our goal of reaching 5,000 data scientists. These include surveying participants 90 days post-training to see if they have applied their new skills (Level 3 Kirkpatrick). We’re also looking at retention, attrition, and employee billing rates of our graduates versus the entire company workforce.
We’re about halfway to reaching our goal of 5000 data scientists, but the results from our pilot were promising. 93.5% of graduates scored proficient or better in their data science skills.
In closing, I’d like to mention that this program is applicable to any priority skill for your organization. We are now applying this formula to other skill areas like AI, data engineering, etc. To find out more about data engineering skills required to manage your big data needs, download the eBook: Choosing the Right Database for Your Enterprise. In addition to continuing to run the program and measuring its success, we’re also thinking about how we can take any element and apply it to other areas. That’s the beauty of taking a scalable approach to learning.