At the recent CIOFEST CIO conference, senior IT leaders from consultants to High St chains shared their fear over rising data scientist salaries, retention but biggest of all, where the next wave of such valuable employees will come from.
While getting more and more committed to Artificial Intelligence (AI) and machine learning, UK-based CIOs are starting to be concerned at the war for the data science talent they need to make these ambitious uses of their extensive corporate data work.
In one startling but far from unusual case, the UK arm of major consultancy EY recently lost a 26-year-old data scientist to an organization his boss had “never heard of” for a starting salary of £170,000—or €205,000. The dearth of candidates is also leading to novel approaches to recruitment of this increasingly sought-after specialism, including intensive internal training and interest in cultivating starters from non-traditional backgrounds, including the neurotypical community.
This was one of the most striking conclusions from a 90-minute live panel discussion of a number of prominent UK Chief Information Officers at last week’s ‘CIOFEST’, the 2022 global user conference of CIO membership organization CIONET.
“We’re struggling to retain our best developers and our best infrastructure people, but I think all of us are struggling to keep good people in roles at affordable salaries,” said one, David Jack, CTO at data services firm Dunnhumby. “When your people are being picked off at 50%, 100% more money than you’re paying them, it’s hard. And that’s our experience in India, the UK, Germany, North America, elsewhere: the market’s seeing extraordinary salary differentials going to new skills all over.”
The CIO who lost his promising employee for that huge salary–Mark Powell, Partner for Data and Analytics at EY UK& Ireland—agreed with the trend. “There is a broader question here, and it’s a UK Plc question: we have not got enough people coming through the system to fulfil the need we all have for data scientists, data engineers, and architects. Her Majesty’s Government has a real problem, for example, as it’s not seen as the most attractive place to work, so it’s really struggling to get data scientists. I just don’t think the British education system is really geared up provide anything like the kind of volume we’re going to need as a country to do this.”
Sanjay Patel, Group CIO at food giant Tate & Lyle, agreed. “We have to get a supply chain of human beings have these capabilities filled who we have to start growing at a young age. But I don’t think it’s just down to government or education or academia to do this: as business leaders, we also have to take accountability here.”
What to do until that supply chain gets created? Powell says there is a temporary work around: never let your data scientists get bored. “You need to find them really cool stuff to do, and not just basic analytics, but hard stuff, something that’s challenging,” he said. In EY’s case, important pandemic projects helping the UK government had inculcated a real “sense of mission stuff,” for example. “It was interesting to see how the scientists got excited when their work has meaning, and they were doing stuff that was technically difficult. The more projects you offer that rather than just kind of mess around with analytics tools and in base analytics, the more data scientists get excited. But trying to feed them that kind of work is a never-ending battle.” As another CIO noted, the last thing you want to do is have this precious set of employees on “boring, boring, segmentations.”
Is home-brewing your own data scientists a solution?
But obviously, the group agree, such tactics are either very hard to deploy or only deliver short-term solutions–hence the growing debate at UK CIO level about how to build a more sustainable pipeline of future data scientists.
“This has to be about actually hiring and developing a different class of data scientists,” offered Powell. “We sponsor multiple PhDs and master’s across the world, just trying to grab the high performers while they’re still embryonic. But ultimately, we’re all going to need to look to attract people into this out of pure science degrees like Maths and Physics, and show them that this is a very, very valuable career path that is even going to become more and more so.”
However, not everyone believes such ostensibly non-data oriented but numerate candidates might be the only source to try and tap. “Where we started to see some traction is bringing people in more at the bottom of our organization,” said Phil Jordan, Group CIO at major British supermarket brand Sainsbury’s.
Jordan cautioned his peers that it looks like no enterprise will be able to meet its recruitment needs from the current hyper-competitive market, so harnessing the talent organizations have elsewhere and turn them into data insight specialists is proving an answer for Sainsbury’s.
“And actually, we’ve had a lot of success, particularly for data engineering roles, by going out to our stores and looking for people on a restart basis,” he said. “We look for people who’ve got the right cognitive ability, bring them in and give them a very, very focused 16-week data science bootcamp, and then making a choice about it bringing them on as an entry level data engineer.
“We find that we’re getting a lot of success with that. I also took a marketing exec and made her our chief data officer, because what she does is make every outcome a business outcome and so stops us drowning in the data lake.”
Jordan was just one of the CIOs who believe that the neurodiverse community is also “another rich seam of talent in this space”, as is EY’s Powell: “We’re looking at people who wouldn’t traditionally come to our organization and looking at how can we enable them to be comfortable here and leverage their great skills. We think there’s a lot we can do in this space.”
However, the group agreed that while challenging, creative strategies to start solving the global data scientist gap need to be explored. As Jordan declared, “I think it’s not all bleak–but that we do have to look in different places for this workforce of the future.”
Read more on government and automation in the CIONET-report The Automation Governance Playbook: Practical Advice for Managing Your Automation Roadmap