The world of people analytics: An HR professional’s perspective

Meet Madalina Banuta, people analytics senior manager at Schneider Electric. Madalina has a diverse and extensive background in various roles related to people management, analytics, and quality/business intelligence. Her experience at Schneider Electric focuses on people analytics and employee lifecycle management.

In this interview, we discuss various aspects of people analytics, including its definition, key performance indicators, data sources, data privacy measures, ethical concerns, required career skills, and the future role of people analytics in organizations. Madalina emphasized the evolving shift from retrospective analysis to predictive and prescriptive insights while highlighting the importance of data governance and privacy.

People analytics (or HR/workforce analytics) teams collect and leverage all employee-related data to optimize any decision-making related to the workforce of an organization. The sphere’s core would constitute the data you gather on your workforce.

The second layer to your sphere: fit-for-purpose tools and tech stack! You need tools that enable your data collection and analysis strategy, with options for user-led design, enhanced user responsiveness and interactivity, iron-clad cybersecurity, and continuous innovation.

The partnership with the business side is the third layer that completes your sphere! People analytics teams work closely with:

• Talent acquisition
• Performance management
• Learning and development
• Wellbeing
• DEI
• Employee engagement.

And business process drivers within the HR ecosystem to ensure a consultative partnership on all data-enabled decision-making processes.

We have two big categories of KPIs when it comes to people analytics. The biggest and core of all activity in people analytics is the business-facing one, meaning the KPIs that are led by the business side and defined by the centers of excellence by the business stakeholders.

In people analytics, what we do is that we provide the necessary data, the right visualizations, and the right tools for people to understand and consume these KPIs and support them in figuring out how I am versus target:

• How has our performance evolved compared to the previous month, quarter, and year?
• Are we progressing in the desired direction and at the appropriate pace?

This becomes particularly crucial when we have ambitious long-term targets spanning 3-5 years. We should continually assess whether our current trajectory will enable us to achieve these targets or if significant adjustments are required.

Within this sphere, we encompass all KPIs related to our workforce, including diversity and inclusion metrics, demographic splits by gender and age groups, seniority profiles, employee well-being, attrition rates, talent acquisition effectiveness.

Additionally, we have another set of KPIs that are inward-facing. Our people analytics team largely defines these to gauge whether we deliver the right tools, content, and analytics to our stakeholders. These KPIs focus on utilizing our tools, including adoption rates and the number of registered users. For instance, if we create a dashboard for our HR community, we want to ensure its usage and understand who is using it and why some individuals may not be utilizing it.

We have two main categories of tools in our toolkit. First, we have foundational products. These are usually stable dashboards or self-service tools fed data from corporate-level tools. These corporate-level tools are essential for tracking employees throughout their journey, from the initial employment data entry to all the changes that occur during their employment. Most of these tools are integrated, ensuring the data remains consistent across different platforms. However, each HR domain has its specific requirements.

For example

A recruiter’s input tools differ from those used by someone in learning and development. Our role in People Analytics is to gather data from these various HR profiles’ input tools and consolidate it into a comprehensive data lake. We make sense of this data on the output side through appropriate visualizations and filtering.

Additionally, we explore large datasets and employ machine-learning techniques. Some of the algorithms and language processing models we use are more generic and not explicitly created for our data. We adapt them to our specific needs and apply them to our internal data. We know that many companies are increasingly exploring tools like OpenAI and ChatGPT. However, we approach these possibilities cautiously, as we are responsible for safeguarding our employees’ data.

Our goal is to strike the right balance between protecting employee data and utilizing the best tools available. We have successfully replicated certain aspects of OpenAI using our internal tools while ensuring data privacy and security.

There are two main aspects of our approach in terms of data privacy and security in response to the rapid digitalization of the workforce and labor market:

• Non-Disclosure Agreements (NDAs). Regardless of the extent of access, everyone with access to HR data must sign a non-disclosure agreement before gaining access. These NDAs are comprehensive and cover various scenarios. They outline strict guidelines on how data can be shared, with whom, and in what format. If someone needs to deviate from these restrictions, the NDAs provide instructions on who to contact and what approvals are required. It ensures that data governance is clear for all employees using HR data.
• Compliance-Related Courses. Anyone interacting with employee data in any capacity should complete mandatory compliance-related courses annually. These courses cover GDPR for European countries, global cybersecurity, data privacy, and security with specific HR use cases. This ongoing training ensures that all HR personnel are well-informed about what data they can share, with whom, and what constitutes private, restricted, or confidential data.

We also pay attention to data usage transparency. We provide employees with clear visibility into their data. A comprehensive disclaimer explains how the data will be utilized, analyzed, and processed whenever data is input into any company platform, such as surveys or profile updates.

Many surveys on employee engagement and satisfaction offer the option to contribute anonymously, allowing employees to provide feedback without disclosing personal data. While certain basic data points are legally required for all employees and are automatically captured, employees can decide what additional data they want to share. It includes creating profiles on internal recruitment platforms, linking external profiles like LinkedIn, or sharing skills assessment results.

As for the analytics level, we only analyze data voluntarily shared by employees. We do not access or analyze data that employees have chosen to keep private. This approach balances leveraging employee data for business insights and respecting individual privacy and data security concerns.

• How accurate and complete is the data we’re basing our analysis on?
• Are we interpreting what we see correctly?
• Are we allowing our personal or historical data-perpetuated biases to influence our analysis, or how do we build our algorithms?
• Are we sharing the analysis only with the people who are legally and functionally warranted to be able to access it?
• Do our employees know and have given their consent on how their data is going to be used?
• Are we creating self-fulfilling prophecies that may negatively impact employee morale or job security?

These are all valid concerns we keep at the top of our minds whenever we set out on any analytical project, regardless of size and scope. We have strong data governance principles and mechanisms that considerably reduce the risks and potential fallbacks. We are also mindful of striking the right balance between safeguarding all these aspects vs. limiting our analytical capabilities beyond the point of bringing any value.

For example
You can frequently use data aggregation to make sense of the data at a macro-level and be able to make valuable recommendations while at the same time protecting individual-level confidentiality of the data.

In our people analytics team, we benefit from having a relatively large team with specialized roles. Here’s a breakdown of the key roles and the skills required:

• Technical specialists. Our team is focused on testing, creating small prototypes, and exploring specific platforms and technologies on the cutting edge of predictive analytics and forecasting. These roles demand highly technical skills, including expertise in the platforms and tools they use.
• Data consultants. Another group closely interacts with the business side and serves as data consultants. They are responsible for guiding how data should be used and helping make sense of data in the context of business decisions, strategies, and organizational changes.
• Product owners. This group plays a dual role, combining technical expertise with soft skills like community management, data visualization, and storytelling. They balance understanding user needs and translating them into practical products, dashboards, and tools.

The common thread among these diverse roles is the ability to tell a story with data. It’s not just about presenting raw data points but making sense of the data within a business and evolutionary context. Concrete skills that prove valuable include:

• Strong presentation skills. The ability to effectively communicate data insights through presentations, whether in PowerPoint or other tools, is crucial. Presentation skills help convey the message and highlight key points for decision-makers.
• Technical skills. Depending on the role within the team, technical skills may vary. Some roles may require proficiency in Excel and Tableau, while others may demand expertise in Python, machine learning, and advanced analytics.
• Stakeholder interaction and empathy. Understanding the needs of the business and end-users is vital. It involves strong persona management—identifying who will use the product, their goals & challenges, and how data analysis can enhance efficiency.
• User-centric approach. Engaging with end-users through interviews and conversations to understand their challenges and needs. This empathy translates into effective data product design and continuous improvement.

Automation and advanced tools are taking over routine tasks. This shift allows professionals to focus on the role’s more imaginative and consultative aspects. Algorithms and automation handle tasks that could be more exciting and valuable to humans, leaving room for human professionals to envision possibilities, aid in scenario planning, and engage as partners in strategic conversations rather than mere executors of instructions. This shift emphasizes the unique value that human insight and expertise bring to people analytics.