When senior executives are sitting on the bench during a period of gardening leave, the temptation is just to find some casual consultancy gigs before the next big job comes up. Few think about ways to disrupt the industry they have just left, even fewer do something about those ideas, especially if it would involve considerable personal financial risk.
Tim Smeaton and Simon Walker, managing partners of Kubrick, are an exception. They had just spent 16 successful years in their last venture, Hydrogen Group, where Tim as CEO and Simon as COO grew the business from a UK start-up to a leading global talent specialist. Hydrogen was listed in 2007 and, at its height, had 450 staff and traded with more than 72 countries. “When we left Hydrogen, most people assumed we would have relaxed. However, having a very lengthy non-compete gives you a lot of time to think, reflect and assess the market,” Smeaton told DataIQ during an exclusive interview at the company’s Tobacco Dock headquarters in London’s Docklands.
During conversations with Simon and long-term friend Nick Allen, now a partner in Kubrick, it became clear that all of them recognised the growing gap around big data, data analytics and data sciences skills and yearned for a better way to address it. “The common thread was that we all wanted to do something in the data world where there was a real issue around resources, especially for data engineering and analytics,” he said.
“We also recognised that organisations don’t train people in the way they used to - we don’t see the appetite for that. Graduate recruitment schemes are good for general management, but not for specialist skills. That allowed us to define something that would be more effective in meeting their people needs.” Smeaton said.
The outcome of those conversations was a disruptive idea for an entirely new business model. Graduates from science, technology, engineering and mathematics subjects, as well as other relevant courses, would be hired directly by the consultancy. It would then train them for the specific needs of big data and data science operations in the commercial world before hiring them out as skilled associates. For two years, Kubrick alumni would be on exclusive contract, after which the client is able to employ the consultant directly with no fees.
The trio researched the idea among their target market where it gained unanimous approval and acceptance as an innovation with the potential to make a real difference. Walker recalled: “We talked to them about our cause - bringing more people into the data industry, especially junior professionals. If we took the risk, hired graduates or junior professionals, trained them and placed them into organisations, wouldn’t that have value for them?”
This research phase had two benefits. Firstly, it opened doors, such as getting invitations to speak at CIO Forums and the like. Secondly, it pre-qualified organisations with an interest in hiring from Kubrick once the scheme started to generate qualified advisors.
It also uncovered another dimension to talent acquisition in this space - a lack of diversity, whether by age, gender or race. Allen pointed out that most organisations, especially in financial services, are very top heavy with white male senior executives and vice presidents. “They need a new workforce to transform their business and do the things which it has never done before,” he said.
That was the spark for Data Inc.lusive, an initiative set up by Kubrick’s female data engineers to attract more women into data. Ola Phillips, big data engineer said: “When I was in my final year at university, it was only because a female role model convinced me to try a career in data, if I didn’t have her, I would not be at Kubrick. At Data Inc.lusive, we are telling female undergraduates how awesome a career in data could be and making sure we can build their confidence.” To date, its graduate intake has been 37% female, compared to an industy average of 16%.
Before it could get to the point of injecting diversity into the talent stream, Kubrick had to overcome three major challenges to get the business up and running. The first of these was how, exactly, it was going to train the graduates it hired to make them technically strong and business-ready.
It brought in advisors with both academic and industry credentials - Peter Hanlon, global CTO, director of big data and monetisation at Telefónica, Dr John Sandiford. head of data science technology at BP, Alice Cooper, global head of FI technology at BNP Paribas, and Avinder Mudhar, chief of staff, global technology at Barclays. They worked with industry data expert and trainer Lawrence Freeman to research and write the entire Kubrick syllabus from scratch over five months. None of its content was sourced from existing training schemes, such as online training courses, and there is no pre-existing collateral culled from YouTube and the like. “We knew to gain our client’s confidence, we had to ensure our staff were exceptional, we therefore knew a major part of making that happen was having class leading training. Significant upfront investment went into this and it will always reamin a key focus for us,” explained Walker.
The modular curriculum steps graduates through key stages and subjects, starting with professional skills, then technology skills before graduates focus on specialist areas, like data engineering, analytics or data science. At the end of the training programme, graduates take independent qualifications, such as Cloudera’s CCA Spark and Hadoop Developer exam, to add to their CV.
Ensuring these juniopr professionals learn real-world business skills is an essential component and part of what differentiates the programme. “It is not just theory - they are given practical problems to solve,” explained Smeaton. “Clients can run projects with us for free. That might typically be a proof of concept using anonymised data. The outputs are then taken to the client - good ones can even get adopted. The more live briefs we use in training, the better equipped our graduates will be. It also improves our project.”
These PoCs are not simplistic - they can involve recommendation engines, multiple data sources, complex algorithms. The process also introduces Kubrick into potential clients, such as the challenger bank which provided an early brief for the project. Clients can also benefit from much cheaper development of potential solutions compared to working with other business partners, such as management consultancies.
Current academic curricula have typically yet to catch up with the commercial need and tend to focus on technical learning. “We want our consultants to understand the real-world context in which those technical skills are needed and utilised,” Allen continued. “We are very cost-effective, but we expect our consultants to add immediate value to our clients and not to have their hand held.”
The second challenge for Kubrick was to hire its first set of junior professionals, given the low visibility which the data industry has within university employability programmes. Its solution was for the three founders to do the milk round in person. “Our strategy was to make our company standout and be as appealing as possible to STEM graduates. We attempted, wherever possible, to tell a complelling story about how the intelligent use of data has the possibility to change the world. We had just over 400 applications to filiter, technically test, interview and assess, before we chose our first group of 15.” Explained Walker
Kubrick now receives close to 700 applications for each intake of 15. Initially, the business opened conversations with candidates to assess their communications skills as well as their qualifications. At this stage, 23.5% are considered worth pursuing. But this approach also risked screening out potential stars because any CV and personality-driven selection process holds inherent biases. To tackle this, Kubrick is looking at setting an online challenge as the first stage to reveal people who can solve problems. Said Smeaton: “It means we are able to select people who we might not otherwise look at because we no longer have tunnel vision.”
In the second stage, candidates are given technical tests using real-world scenarios. A typical example might be to ask candidates to work out whether the UK economy is improving using social media data. Those who spring straight into action and simply scrape tweets using #economy will fail. What the company is looking for are individuals who start by questioning the question. What is “improving” and even whether social media is the right data source. “We look for inquistitive and energetic minds,” noted Smeaton.
The third challenge for Kubrick is to get to profit given the high cost base of the business. The first cohort of 15 graduates was hired on 1st July 2016 and Kubrick is now onto its third wave. With the third cohort in training, the first two waves are now entering profitable consultancy engagements which start to return on the founders’ investment.
The business is entirely self-funded - a considerable risk for the founders, given the £26,000 cost of the initial four months for each graduate it hires, trains and employs. “That is both good and bad. Good because it is a high barrier to entry for any competitiors. Bad because we need a lot of operating capital,” he said. But Smeaton is confident. “Most clients who give our consultants a project to work on want to take them on as staff.”
The business is also picking up very positive word of mouth - it was introduced to DataIQ via an end-user who had taken several of its consultants into a project. Nobody in this industry needs reminding of just how hard it is to recruit new talent. By getting on to the front foot and creating “oven-ready” practitioners, Kubrick may just have found the breakthrough solution.