Senior software developers have an unusually high bar when judging a company’s candidate experience. They’re, by far, today’s most toughest roles to fill with ample job opportunities before them. The technical manager/director role, for instance, is the hardest IT job to fill at 107 days, according to Burning Glass.
Senior developers have little motivation to move forward with companies that are lacking in providing a world-class, positive experience for them in the job application process.
The reality is that many companies struggle with this piece. But Badoo…the social-focused social network….has been wildly successful overall, earning enormous growth in their user base to more than 380 million, with a whopping 400,000 new users added daily. Its senior technical team, of course, is pivotal to sustain this type of growth.
So, when we came ran into Badoo’s Head of Talent Acquisition, Brad Richards, we were excited to learn more about their problem – achieving a rapid hiring process for senior developers en masse, while retaining the integrity of Badoo’s high technical bar. Brad gave a fascinating talk on how he achieved this world-class senior engineering recruiting process, which centered on improving the screening process.
Here’s an overview of what Brad and his team did:
What was the initial screening process like for senior developers?
Brad: We’ve always relied on take home coding projects to assess our candidates technical skills over other samples of work. This started to become an issue because we primarily hire senior engineers.
With families or other commitments, we realized they were limited on time when trying to complete our 3-hour projects. We started thinking if we were going to change our technical screening process we had to take into consideration all the unique problems senior engineers face like… security restrictions at their current jobs that don’t allow them to send samples of code, or being people-managers for so long that their coding skills might be a little rusty. We want our senior engineers to be up-to-date with the latest tools.
So how did you change the framework?
Brad: Initially we were crafting coding tests that were graded manually. We were pretty confident in this method, but it just wasn’t scalable. It created a major bottleneck, requiring multiple email correspondences, difficult time constraints for both developers and candidates, and lack of secure environment.
Plus, we realized on the candidate side, this was a big ask. It might make them feel like they needed to give a response within 24-hours.
We needed to figure out a new, less time-consuming and streamlined process to evaluate high volumes of technical candidates fast. This was a difficult challenge given that our hiring managers were attached to their method of screening.
We researched several different solutions to boost efficiency in technical hiring while creating a strong candidate experience. The ones we came across were HackerRank, DevSkiller, and Codility.
How did you ultimately make the decision?
Brad: We took into account a lot of factors when evaluating each platform, but in the end HackerRank offered a number of key differentiators. One of the big ones was helping us overcome our manual assessment bottleneck. Offering automated grading was critical given we had a finite amount of time in the week of interviewing people, and a high target number of candidates that we wanted to extend offers to by the end of the year.
Reducing time from 3 hours to 1.5 hours for technical assessments created a compounding difference in time savings.
From a recruiting perspective, the seamless experience with Lever ATS integration was the real clincher. Unlike competitors that we researched, the integration allows us to send tests without leaving the ATS, the report scores are returned back to us in an evaluation format so it’s the same as our candidate feedback report. And if we want a more detailed look at how the candidate has approached the challenge we can do that easily via the HackerRank platform as well.
In order to get hiring manager buy in, we needed a platform that would allow customizable challenges so they would feel comfortable changing their testing methods. This feature has given our hiring managers the confidence to test various aspects from database administration to core computer science competencies; they have the option to compile tests in any way that also allows customizable questions.
Timing also played a key role in choosing the right platform. With the test expiration link feature, candidates would have the ability to complete the task whenever they’d like as opposed to our fixed time before.
Another differentiator was CodePair. Having the live ability to assess remote candidates programming ability in a real-world environment was so valuable. It also eliminated untracked Skype conversations and gave us more ability to review and measure success.
In the end it also boiled down to what appealed most to developers. The HackerRank brand and its community has a currency with developers. Developers like to test themselves and are familiar with it, not only as a brand but also the interface. When looking at two competing options, this helped us make a final decision.
How successful have you been since optimizing your framework?
Brad: We transformed our recruiting process to be more streamlined for our recruiters,
developers, and candidates. Using HackerRank assessments gives us a more secure environment that candidates cannot plagiarize and we look more professional from their perspective.
We’re now able to screen technical skills at scale. In a recent one day hiring event where we pre-qualified candidates using HackerRank, out of 40 developers who passed the test, only 1 candidate was a clear mismatch and we extended 9 offers. Our candidate journey went from 8 hours down to 5.5, and this includes the 2 hour coding test.
Hire, hire, hire has been our motto and HackerRank has been an important part of accelerating our ability to do so. HackerRank reduced time from start to hire by removing the bottleneck of finding time to take the three hour test, and grade them manually. This reduced developers’ time and gave us more time to consider a lot more candidates to help us achieve our ambitious hiring plans.