Who will be in demand in five years: data scientists or welders? Vitaly Altukhov, director of development and research at Profilum, explains how to predict demand for professions using predictive analytics.
The shortage of personnel is one of the main problems of the modern labor market – partly a consequence of its rapid development: in the last decade alone, 30% of in-demand job skills have been replaced by others, and with the advent of generative artificial intelligence, this process can accelerate significantly.
Requirements for specialists and the professions themselves are constantly being transformed, and the education system does not have time to adapt to the changing demands of the market.
The cycle of introducing changes in classical higher and secondary specialized education takes several years, and during this time trends and technologies manage to change again. In such a situation, the role of predictive labor market analytics becomes especially important.
Why predictive analytics is needed
Unlike standard analytics, which assesses the state of the labor market at the current moment, predictive analytics predicts what will happen to supply and demand in different professions and fields of activity in the coming years, which competencies and specialties will be in demand and which will lose relevance.
What is it for? The first and most obvious thing is that people will be able to more rationally choose the direction of study, assess their prospects in their current profession, and make decisions about retraining. But this is only one of the effects of predictive analytics. Its more global capabilities make it possible to change the labor market and education system taking into account future trends.
Education. Knowing what competencies and professions will be in demand in a few years, educational institutions will be able to expand and contract programs, make adjustments to them, and change content to give students the skills that will be in demand by the time they graduate.
State. Understanding trends in the region’s economy gives authorities the opportunity to adjust strategy, open new universities and colleges, create jobs, and formulate programs to attract certain specialists.
Business. Based on predictive analytics, a company can predict the shortage/surplus of specialists and adapt its HR strategy to these inputs.
Development of predictive analytics
Classic labor market analytics, which assesses its current state, began to develop quite a long time ago. Perhaps one of the most striking examples is the American project O*NET, created in the 1990s, which conducts regular studies of the professional landscape throughout the United States and in individual states.
The first data-driven solutions that analyze large volumes of data based on machine learning began to appear at the turn of the century, when, on the one hand, sufficient amounts of information had accumulated, and on the other, it became possible to computerize them.
This is how such structures as the American The Burning Glass Institute , the Singaporean SkillsFuture , the Australian Faethm and others arose.
, in 2023, a message appeared about plans to develop a system for universities that will allow them to analyze the personnel situation by specialty and make forecasts about the development of the labor market for five years in advance.
It is expected that this will help reduce the gap between changing market demands and the competencies that students receive.
IT solutions for predictive analytics based on big data and machine learning already exist: we have been developing one of them at Profilum for several years based on the Data Lab direction.
How it works
The main source of information for many predictive analytics systems is job sites and other resources where vacancies and employment advertisements are posted. Algorithms collect data on published vacancies, resumes, and their content:
salary levels, skill requirements, and requests for competencies, and then process this information, combining similar vacancies, groups of professions, and skills. As a result of the analysis, the algorithms produce clear conclusions about changes in supply and demand in the labor market.
For example, the solutions of The Burning Glass Institute, a global benchmark whose research results are used at the World Economic Forum, work on this principle.
The organization collects open data on the labor market on websites and hubs that provide information about the demand for specialists and resumes of applicants and passes it through machine analysis.
The output is an understanding of what competencies are in demand and how this will change over time. In addition, The Burning Glass does many projects for the educational sector and helps adapt programs to changing market demands.
At Profilum we also use information from work sites for a certain period. Over the years of analytics, historical data has been accumulated, which gives an idea of certain long-term trends within a particular profession, region or country.
For example, in the last few years it is clearly visible how the profession of a credit analyst is being washed out of the market, whose functions are gradually being taken over by automated systems.
And since the mid-2010s, there has been a clear trend towards the growing popularity of a new role – marketplace specialist. Based on such dynamics, you can build a forecast for the development of a niche for up to three years, and by adding additional data – much further
Not just AI
But it is still impossible to build a forecast based solely on the dynamics of recent years. Supply and demand in the labor market depend on many factors: the emergence of new universities and colleges, technological trends, development strategy of the country and region, legislative initiatives, population migration, including internal, and much more.
If a region plans to develop a new field, it will soon need a large number of workers and engineers. Another example: after last year’s presidential decree on additional measures to ensure information security, the demand for information security specialists increased.
And the development of AI may soon reduce the demand for line specialists in a number of areas with highly algorithmic work.
Therefore, analysts look for information on technological, demographic, and economic trends and apply it to a forecast made using machine learning – this is what we do, in particular. And this new complete picture requires another expert “human” analysis.
Perhaps, with the development of AI, algorithms will be able to take on part of this work, but it is still difficult to judge what the role of machine learning will be in principle in any analytical processes in the coming years.
Most likely, we will use artificial intelligence as a tool to help collect and process information – but the key role in making final decisions and building predictive models will remain with humans.
Forecasts in the VUCA-BANI world
But here we face another problem. The crises since 2020 have demonstrated how fragile and unreliable any forecasts can be under the influence of external factors. Let’s remember the first year of COVID-19, which temporarily “canceled” all offline professions and brought medical and digital specialties to the forefront.
The events of 2022 also had a strong impact on the labor market. How is it even possible to make labor market forecasts in today’s unstable VUCA-BANI world?
In general, focusing on past trends is similar to driving a car using the rearview mirror. If the road begins to turn, we can predict that the turn will continue, but if the situation ahead changes dramatically, the reflection in the mirror will not help us much. Therefore, trends in the labor market, especially those formulated by artificial intelligence, must be treated carefully. And a “black swan” like the one we saw in 2020 or 2022 can be difficult to predict even for experts.
But on the other hand, global trends, as a rule, do not change at the snap of a finger. If there is a need in the market for a large number of certain specialists, it is unlikely that they will suddenly cease to be needed.
The structure of their competencies may simply change: for example, during Covid, teachers and tutors have not lost their relevance, it’s just that the skill of remote work has become important for them.
Global trends, such as digitalization, are quite stable. The demand for digital professions was actively growing even before COVID-19 – the pandemic simply gave this trend a powerful impetus. In other words, the labor market maintains stable trends, although some of its elements may change under the influence of events.
Therefore, it is important to make forecasts in the same way as it is important to constantly enrich them with new factors in connection with the changing situation. In Russia, the role of such developments is especially important given the severe shortage of personnel in many areas:
they will reduce the imbalance of supply and demand and ensure that the competencies of specialists match the needs of the economy.
