How AI-based strategic workforce planning can enable the „skill shift“ in digital transformation

Much ink and digital storage has already been used to explore, describe or even evaluate the disruption of our work through digital transformation. I, too, have published on this „skill shift“ several times, such as in this article more than three years ago.

Elaborated strategic workforce planning in response to „Skill Shift”

Against this background, more and more companies are intensifying their efforts in various aspects. Almost every company is now pushing to become more agile. Finally, agility means the ability to cope with any changes, including digital transformation (Krapf 2016).

Another topic that is increasingly gaining momentum is strategic workforce planning (SWP). This is because SWP is the logical answer to the question of which competencies organizations will need in the future if digital transformation – and especially automation through AI – radically changes many jobs.

Do agile organisations still need an SWP?

In a previous article, I raised doubts as to whether the traditional SWP is the right approach for agile organisations. The problem with traditional SWP is that they are too slow, too inaccurate and too complicated. It is a human solution that does not do justice to the complexity of the problem.

I have therefore advocated radically questioning the traditional SWP. Such a radical solution would be the complete abolition of the SWP. As an alternative to traditional central driven SWP, a decentralised, continuous development of individual employees would be encouraged instead. This quasi absolute self-organisation in competence development can succeed in agile organisations that are strongly purpose-driven and thus have a generally accepted development direction (keyword „swarm intelligence“).

 The traditional SWP is „stuck in the middle“

However, the absolute decentralisation of self-organised competence development may be a radical step that is not suitable for all organisations. It is a sustainable solution in comparison to the traditional SGP, which cannot centrally control the skill shift. However, decentralisation is not the only solution.

An alternative to decentralized self-organization is the radical improvement of the traditional SWP through artificial intelligence. Here the paradigm of central control is (for the time being) maintained. In contrast to the traditional SWP, however, the calculations are improved or automated in such a way so that the complexity of the skill shift can (approximately) be handled by complex algorithms.

In most large companies, however, neither one nor the other variant is lived by. The traditional SWP is „stuck in the middle“. It is not elaborated enough to correctly depict the skill shift. And there is no established culture where employees develop the needed competencies on their own.

stuck in the middel SWP_EN.png


How AI can take the SWP to a new level

Some time ago Linda Vos showed me how artificial intelligence (AI) is getting better and better in order to calculate the „Skill Shift“ in the SWP. Linda is a data analyst at PwC and specialized in SWP. As she is much more competent in this topic than I am, I link here to a recorded webinar on this topic and limit myself to a few lines to outline her basic idea.

Clusters by competencies not function

Linda’s idea is to no longer to think in terms of functional clusters, but rather to break down each function into all relevant competencies. Machine Learning, for example, can be used to discover jobs with similar competencies that are not obviously related at first glance. For example, the ballet dancer and the skateboarder, both of whom are incredibly skilled in staying in balance.

Data on competences

In order to know the relevant actual competencies, an initial investment is of course required. It must be clear which competencies are particularly distinct in which functions. This analysis can be simplified by assuming a normal distribution of the individual competencies for each function. This means that there is no need to actually analyze to what extent the function holders do justice to the function-relevant competences. Rather, it is assumed that the expected competences are fulfilled on average. Individual – also function-independent – competences are not considered in this approach.

In a further development, however, additional data sources could also be used to take individual competence data into account and thus improve the calculations. Such data could, for example, come from LinkedIn, from the employee profile, from previous job references or courses completed at Coursera, etc. But not least because of data protection, we are probably already moving further into the future with this variant.

Calculation of different scenarios

Once the algorithm has learned which competencies are available and how they can be grouped, target competencies are determined. AI support can be used to calculate different scenarios in order to find out which competence gaps exist in the entire organization. In contrast to traditional SWP, more elaborate data can be used, such as the probability of automation of certain competencies or other endogenous or exogenous influencing factors.

Gap analysis and measure determination

From the elaborate, AI-supported calculations, an analysis of the current state of the existing competencies and a list of the necessary target competencies according to the various scenarios are created. The AI-based SWP can then be used to calculate the smallest possible gap. For example, AI can show how existing competencies can be regrouped in such a way that the development effort is minimized by the skill shift.

Full decentralization or AI-based centralized SWP – what now?

Admittedly, the two solutions are diametrically opposed at first glance. On one side is the decentralized solution that gives employees full responsibility in development. On the other side the AI-supported, centralized „ultra-control“.

However, this comparison does not have to be seen as a contradiction. Rather, the AI-based results could be used to increase transparency for employees and thus promote their self-development. Similar to how I used my training data in the past to find out with which pulse, with which step length, with which speed etc. I ran in order to be able to specifically improve myself in later trainings. The AI-supported findings can thus be used to promote self-development accordingly.

swp KI und agil_EN.png


How do we achieve this utopia?

Now it is utopian for most companies to set up a SWP in the near future, in which self-organized employees drive their own development on the basis of an AI-supported SWP. But like everywhere else, the same applies here: Think Big, Act Small. Even Mount Everest is not climbed in one day. Rather, it takes years of training, months of acclimatization and several partial ascents until, with a little luck, the summit can be climbed.

Applied to our example, this means that companies take this admittedly visionary SWP as a long-term goal and then take their first small steps towards it.

On the one hand by promoting the self-learning competence of employees and the development of an agile culture.

On the other hand, by generating a first data base that can be used for first analysis. Hence, step by step a planning tool will be develop that can generate a real value added.



Kommentar verfassen

Trage deine Daten unten ein oder klicke ein Icon um dich einzuloggen:

Du kommentierst mit Deinem Abmelden /  Ändern )

Google Foto

Du kommentierst mit Deinem Google-Konto. Abmelden /  Ändern )


Du kommentierst mit Deinem Twitter-Konto. Abmelden /  Ändern )


Du kommentierst mit Deinem Facebook-Konto. Abmelden /  Ändern )

Verbinde mit %s