Tech Talent: Where Meerkats Meet Machine Learning

Hajo van Beijma, Founder of Atilax, is joining us as a speaker at World Summit AI on 12-13 October 2022 in Amsterdam. He will be talking about The Great Resignation and Tech's Talent Drought: Reimagining How and Why We Work. Below, you can read about the future of Machine Learning and Hajo's journey towards it.

Tech companies are all focussed on the same talent pools and while salaries in Europe are getting closer to the silicon valley way of remuneration a lot of the searches for senior talent is outsourced to executive search firms. The recruitment firms that are winning the bids for best tech job searches are mostly firms who are proud of their enormous network but is the size of the network still the most important factor for success in helping grow tech companies and scale-ups and what can we learn from the animal world?

Coined by Robert Metcalfe, the inventor of Ethernet, Metcalfe's Law is often used to describe how tech companies have come to dominate the current business landscape. The law states that the value of a network grows by the square of the size of the network. So, for example, if a network has 1000 machines, its value would be 1,000,000.

It's a neat concept, but does it actually work in practice? Unfortunately, the simple answer is no, not most of the time. And in 2022, it's time to radically rethink how we approach networks.

The crux of the issue here is around the word "value" and how value works in reality across different types of networks. When Metcalfe originally put forward his law, he was focused on phones and fax machines - tangible tech with wires. Tightly designed hardware and software (nodes) fulfil a specific purpose. The value of these nodes lies in the functionality they offer, and this functionality is fixed and unchanging. Each node has inherent value, so scaling these nodes up adds more value¹.

Although "value" is a relatively vague concept in itself, it becomes even trickier when considering networks of people. Although Metcalfe later expanded his theory to include people, the connections between users, and how people collaborate and influence each other, this is where the law starts to fall down. Why? Because networks of people work very differently.

When considering social networks, it becomes clear how additional users can negatively affect value - extra noise or "low quality" content can drown out the intended communication²³.

So what does this mean for how we approach networks of people going forward? Well, Metcalfe's Law might not provide all the answers, but the animal kingdom might. By studying the mathematics of animal populations like Meerkats, we can gain a unique insight into how the life cycles of networks work.

Population size, density, availability of resources, and overall fitness are critical factors in the health of Meerkat populations. So, for example, Metcalfe's Law would suggest that having more meerkats in a group would increase value (enable the population to thrive) exponentially.

However, this is only true up to a certain point in reality. Small meerkat populations are vulnerable to predators, but too crowded groups deplete the natural environment and eventually cause population decline. In other words, meerkat populations are at their most healthy (or provide the most value) when they exist in the "Goldilocks zone" - the point where population and value are high but population-depleting factors are low.

The Future of Headhunting for Scale-Ups

So, how does all this talk of meerkats relate to scale-ups? First, companies entering a rapid growth period need to attract high-quality talent to accelerate their success. And they pull that talent from the available networks of people. The value of these networks lies in high-performing individuals and their participation in the network.

Today, much of recruitment is still done manually, and as a result, it's subject to individual biases and human error. Try to not only focus on the size of your network or the network of your preferred executive search firm but look for smart and powerful links inside networks and between different networks. This work is difficult to leave up to human beings alone and there is a better way - recruitment in combination with machine learning (recurrent neural networks). Successful long-term talent strategies require a more scientific approach to headhunting, act as a meerkat - An intelligent, social animal that looks beyond the obvious (candidates) and is focused on working in teams.