Why HackerRank and other coding tests are ageist
Given his background in financial services technology, you might think that Marc Adler would have no problem walking into another job in the sector. Indeed, he's already got a new role as a chief architect lined up. But after 10 years in financial services technology, including two as head of front office development at Citadel Securities and three as the chief equities architect at Citigroup, Adler – who’s currently on a short sabbatical – came up against something unexpected when he tested the market: the newly ubiquitous coding test.
“Suddenly, banks, hedge funds and proprietary trading companies have started screening candidates using these tests provided by HackerRank and other coding sites,” says Adler. “It’s fine if you’re a masters student who’s just graduated in computer science, but if you’ve got 20-25 years’ experience working in the sector and haven’t spent your time building complex algorithms, they’re a lot less relevant.”
Adler has interviewed hundreds, if not thousands, of candidates in his time. He says Hackerrank tests have replaced the traditional “whiteboard screening” for technologists. Although he applauds the tests' use for candidates leaving university, he says the automated process also has its limitations.
“With a whiteboard, you could pose a typical problem and even if someone didn’t the get the code exactly right, you could see their thought processes,” says Adler. “With Hackerrank, they’ll ask you to do the test before they even meet you: it establishes a standardized performance hurdle.”
The problem, says Adler, is that the new performance benchmark set by the coding tests favours university leavers for whom writing esoteric algorithms is second nature. If you’ve spent decades working at the coal face, this kind of thing is a lot less relevant.
“When you’ve got a history of having built a high frequency trading system for Citadel, Deutsche Bank, Morgan Stanley, you’ll know a lot about latency and order flow – but you might not know how to find a maximal duplicated substring within a palindrome on a HackerRank-type test. Does that really matter?” Adler asks.
Most damningly, he says the questions posed by HackerRank and similar tests aren’t necessarily relevant to the work actually done in finance technology roles: “They’re the sort of problems you’re posed at university. - There’s a focus on data structures and algorithms, but this is less relevant when your success as a developer is in building complex enterprise systems.”
For experienced programmers like Adler, the new tests aren’t insurmountable – all that’s required is an injection of time to learn the new Hackerrank tricks. This in itself can be a challenge, though. Neophyte programmers who aren’t in full time jobs might have plenty of time to discuss Hackerrank solutions in forums. When you’re already in full time work, it’s not so easy.
“Unless they take a significant amount of time out of their day jobs to really prepare for these tests, it’s hard for experienced developers in the finance industry to be ready for this new process,” says Adler. “- All the more so if they’ve climbed up the management chain and have additional responsibilities."
If banks use HackerRank and similar tests as a screen for all candidates, Adler therefore argues they'll discourage candidates with real world experience of working with their systems from applying. This may be their intention, but it doesn't sound like a good idea.
Contact: sbutcher@efinancialcareers.com
