The true value of HR Analytics lies in new Data Science approaches capable of complementing HR methodologies. Today, new approaches, new skills, and new Data Science methods are essential — but above all, it’s about thinking with data in mind.
HR departments must make a logical leap in choosing the right data, formulating the right questions, and making — even strategic — data-driven decisions.
For years, HR departments have relied on well-established processes, methodologies, and measurement tools. Perhaps for this very reason, they often struggle to see the true potential of Data Science within HR.
Focus on the individual
Traditional People Management processes and related metrics all respond to the same question: what value can a person bring to the company?
Potential assessments, skills, performance, talent, succession planning focus on the growth of the individual within an organizational and business context assumed to be stable.
This is no longer the case, because the competitive context has changed.
Today, the customer is always connected, service components represent an increasingly significant part of the value proposition, and competitors may come from entirely different industries.
However, the innovation and transformation efforts many companies are making often translate into the recruitment of profiles related to new professions, which are then placed into an organization that retains unchanged working methods and models of performance and potential, rather than creating a workplace that provides the autonomy needed for people to generate innovation.
The companies that are proving most competitive today are those that ask people to act with greater discretion in achieving results. As a result, roles, responsibilities, and autonomy must also be redefined. The organization must become truly adaptive. New organizational models place teams and networks—not just individuals—at the center.
Today’s predictive indicators for personal performance and development are outdated. In summary, the data set of current People Management processes offers a "kaleidoscopic" view of the individual, often not useful for supporting Human Capital decisions with strategic impact.
Siloed thinking
HR Departments continue to leverage the methodological set and evaluation tools that have stood the test of time, providing reports that are certainly more appealing and easier to interpret today than in the past.
However, these reports only take HR data into account. On the other hand, CEOs and company Boards have always asked HR Departments simply this: "producing personnel reports".
The potential of HR Analytics goes beyond simple "personnel" dashboards. As in any Artificial Intelligence project, the richness and diversity of the data set are crucial factors in identifying new patterns, explaining the reasons behind positive/negative outcomes of certain actions, and especially predicting which factors will drive successful decisions, actions, and behaviors in future scenarios.
The true value of HR Analytics lies in the ability to identify relationships and algorithms capable of revealing previously unseen meanings and phenomena, and supporting HR Management decisions through predictions.
Just think of the topic of "Employee Experience". Going beyond trends, the ability to design a successful work experience requires a personalized understanding of employees across many aspects of their professional and personal lives.
Not only that. One cannot stop at taking a snapshot, but must "shoot a film". Metaphorically speaking, it is not enough to enrich employee data by going beyond professional and demographic information, but it is necessary to understand, for example, how certain workplace design choices (often digital), reward systems, and welfare programs can positively impact performance, sense of belonging, etc.
It is, therefore, crucial that HR Management knows and understands the methodologies of Artificial Intelligence and Machine Learning.
Only in this way is it possible to consciously leverage business (even external) and social data to build HR Analytics capable of revealing new patterns in the relationship between people and business impact.
Data availability and data quality
This is a key point in any Artificial Intelligence project, not just in the HR field.
As for data availability, even when the HR Department overcomes siloed thinking, it faces two problems:
- the data exists, but it is partial and scattered across different databases
- the data is not comparable because it was collected using different methodologies and timelines, is partial and/or unreliable
The first problem can be considered simple. That is, it is well identified and has available solutions. Automated Data Preparation processes, or in extreme cases more costly human intervention, are available and widely tested.
The second issue, however, is more complex. Often the data set is incomplete. Just think of performance or potential evaluation data, which differs from person to person depending on their career paths. Here too, Data Mining functions can help us, but with the aim of
The problem of data quality has long been evident when it comes to performance evaluation. We have often witnessed HR stepping in to urge Managers to use the evaluation scale more objectively and less "protectively" or "politically". These situations have led to a less reliable data set.
These problems are easily addressable, as demonstrated by the effectiveness of solutions already tested through Data Science approaches and methodologies.