Max Blumberg, Ph.d., Blumberg Partnership Founder, Visiting Professor at Leeds University Business School

Upon completing his Ph.D., Max Blumberg launched the Blumberg Partnership, a top 50 analytics consultancy which delivers analytics and machine-learning solutions to clients that include Nestlé, Hilton Hotels & Resorts, Lloyds Register, the BBC, Fujitsu, Barclays Corporate and Friends Provident.

With truly unique expertise and a practical grounding in strategy, finance, marketing, economics, computer science, mathematics and HR, he developed a full-systems approach that can be applied to diagnose and solve complex challenges arising from operating models, organisational structures, and job design. He was a speaker at the recent HR Experience conference in Belgrade, organised by Infostud, where he spoke on talent acquisition, people analytics and the challenges faced by the HR industry.

The overall impression is that the talent pool today plays an increasing role in a company’s success and that the talent base is scarce. Based on your experience, would you say that this is so, and do companies have an adequate response to such challenges?

– That’s an interesting question. Depending on the industry, wherever there is a scarcity of talent, someone will try to automate this work. Therefore, I think talent has an important role in jobs that cannot (yet) be automated. For example, the work performed by STEM workers is not easy to automate, and hence it is widely recognised that STEM workers are an increasingly scarce resource (until technology discovers a way to automate their tasks). In contrast, however, many routine tasks like clerical work, administrative work – and even some aspects of medical practise – are to some extent already being automated. So, for companies, certain kinds of talent – particularly automatable talent – are not considered particularly important in the eyes of many companies.

How can companies recruit employees with rare skills that cannot be automated?

There are many approaches to acquiring scarce skills, but in the end, it all comes down to simple supply and demand: if there is a shortage of the skills you require – like STEM skills – you can expect to pay a lot of money to acquire those resources. So, for instance, some companies build communities of these skills so that potential employees feel an affinity with the employer brand even before they join the company, or after they have left. For example, some consultancies invite STEM students to join communities even before they graduate.

I also think that the government should be encouraging more people to take on STEM education and careers – particularly from poorly represented groups based on gender and ethnicity etc., but that’s a long-term project. Another approach for acquiring rare skills is to outsource them or use contractors – but only on the condition that the skills you’re outsourcing are not among your key strategic capabilities (which you should never ever outsource).

The bottom line is that, if you cannot outsource or automate, you’ll need to work hard and invest significantly to attract scarce talent. Companies that excel at this are, for example, Facebook and Google: many STEM workers – whether they admit it or not – would love to work in these kinds of innovative environments.

Can digitisation and automation be used to manage teams?

– Companies tend to replace people with automation wherever they can, and this goes for managers as well, but the skills relevant for good management – like emotional intelligence, collaboration and so on – are probably among the last areas that will be automated.

Be very careful about using technology to replace certain tasks usually performed by human managers. For example, say you have the choice to invest a million dollars in management soft skills training versus building an attrition model to predict whether employees are likely to leave your company. In this instance, the ROI for manager training is likely to be much higher, because while the attrition model will only fix attrition, the manager training is likely to improve retention, productivity, engagement, innovation etc. It has a whole range of positive consequences, whereas the attrition model will only address a single issue.

Be very careful about using technology to replace certain tasks usually performed by human managers

When it comes to people analytics and AI, what are some areas where these emerging technologies could help HR?

– Before we answer this, we need to be careful about how we define ‘People Analytics’. If you have a thousand job applications to process and simply automate these, this is not people analytics – it is just simple task automation. For me, people analytics is not just about simple automation, reporting, pretty graphs etc., rather it is about making usable statistical models to make useful predictions.

What about AI? Someone once asked me: how can we best use AI in our company? My answer was a question: what’s the most important prediction that your managers could make so that, if they get it right, your company would make millions? For example, if you work in a gold exchange and you could predict the price of gold, that would be amazing, right? If you believe that predicting which employees would be the best to make your business profitable, that’s a great job for AI in the field of people analytics.

There’s a lot of buzzes around all the things we’ve discussed, but in reality how many companies are actually eager to really embrace scientific methods in HR?

– I would say that a lot of analysis in HR has been very basic up until now. I mean, some companies have had some good results, but very few are using scientific people analytics to make predictions that make them those millions. And I think many senior managers are looking at their people analytics spend and asking “shouldn’t we be getting more for our investment?”. As a result, business consultants like myself, with deep statistical experience, are suddenly getting a lot of calls requesting help in leveraging their people analytics investments. But I’m very curious about why these requests are suddenly coming now. I personally believe it is because people analytics vendors have made promises they cannot keep and HR bosses are suddenly having to justify their investments. And that’s why they are now starting to look for more sophisticated solutions.

What’s your vision of a company in the future? – I think the company of the future will be managed by a very small, senior executive group, and then below them, there will probably be a lot of automation, outsourcing and gig work going on. This top team needs to be very innovative and very good at working in small teams; it’s whole different psychology; it’s the way that boards typically need to learn to operate.

I also think that we are moving from the age of people analytics into the age of resource analytics, where you are looking to determine the best balance of resource that you’re going to need across the organisation, what percentage will be automated, what will be outsourced and so on. In other words, workforce analytics will be replaced by the more general “resource analytics”.

The way companies are surveying workforces and collecting data about them sometimes seems to be at the edge of invading someone’s privacy, or even over the edge. Are they sometimes being too intrusive? Where’s the limit?

– I think it’s about transparency. If you’re going to use your employees’ data, you must let them know. But I also have to say that, as an employee, it is naïve to think that your data is not being used by your employer.

More generally, I think it is unfair is when companies make money on the basis of an employee or consumer data – selling it to advertisers – without sharing any of that profit with the people whose data they used.

Finally, I’m very keen on the idea of value equity for employees. What I mean is that if there is a profit improvement of 20 per cent for the investors of a company, then there should also be a 20 per cent increase in employee benefits. Why should one group win at the expense of another?