Creating a culture of data-driven HR rests on the successful use of analytics, which, in turn, rests on superior...
change management strategy and execution.
Change management is critical to the success and adoption of new software. Different types of software implementations have both change management commonalities and specific needs. That is true of analytics in general, but HR analytics -- which is increasingly being referred to as people analytics -- has even more specific approaches that need to be considered, planned and executed to enable effective adoption and use.
HR analytics tools are typically used by a variety of users -- including executives, HR professionals, HR business partners and managers -- each of whom have different needs. And with such a variety of users and needs, change management serves as a strong base to ensure the right communication, training and education can be provided to enable successful design and adoption of HR analytics technology.
Data-driven HR and analytics' role
Analytics is still fairly new in HR, and the HR department has had a slow uptake of analytics since the technology has become widely available. While the HR department is a treasure trove of data, HR leaders and their staff haven't always understood the value of analytics and how it can help HR's value proposition, as well as help deliver on their goals and objectives. Many still don't. Typically, when an organization implements analytics technology, HR isn't one of the focal functions. Instead, those functions are much more likely to be finance, sales, stock or materials.
That means the HR department typically doesn't contain individuals with analytics experience, and likewise, analytics experts don't typically have much or any HR experience or understand what HR needs from an analytical perspective. Determining how to bridge this gap should form the backbone of your change management approach. The training and education aspect of your change management plan has to have a strong focus on enabling HR to become experts in analytics, as well as educating your analytics experts on how they can support HR with building and interpreting analytics that can support their business objectives.
In addition, senior executives, managers and even the CEO can use HR analytics data to make better informed decisions about aspects of the business, such as employee turnover and retention, employee training -- for example, costs, compliance, correlation with performance and productivity -- and succession planning and will need training specific to their needs.
Data-driven HR rests on nuance
Interpreting analytics and the data that it is derived from can make or break whether your HR analytics have any positive impact on your business decisions. Without the ability to understand what analytics is really telling you, you will most likely misinterpret and use that misinterpretation to make critical and possibly costly business decisions. Analytics is only as good as the methodologies, frameworks and models that it is built with.
Also, importantly, so is the underlying data. Additional change management around data accuracy, auditing and maintenance should be planned to ensure the source is fit for purpose for your analytical needs.
The implementation and understanding of analytical methodologies need to be communicated, and trainings need to be created. Likewise, a process for the ongoing modification of methodologies as analytics evolves or changes is needed and must be communicated to the HR analytics stakeholders.
Getting schooled in data-driven HR
Training and education are key components of your change management plan. Training is needed to ensure that users are able to accurately interpret the analytics available to them -- particularly predictive analytics -- so that both correlation and causation can be identified (one on its own doesn't necessarily prove what the analytics might be showing). The biases, worldviews and preconceptions people bring to analytics are a critical factor in the success or failure of analytics and can lead to misreadings, which, in turn, lead to inaccurate decision-making, with both fiscal and engagement ramifications for an organization.
So, who should provide this training? Typically, the implementation partner will provide some level of software training, either train-the-trainer training or direct user training for the different user groups. In some instances, both types of training may be provided. Depending on the in-house expertise, analytics experts may be available to provide training on the specific methodologies and models employed and on how to use those to correctly interpret the analytics and how to apply that interpretation to decision-making. The implementation partner may offer this type of training instead, in conjunction with the in-house analytics experts, or as complementary training.
Training and education also need to be top of mind as business needs inevitably change.
The ongoing journey of data-driven HR
As with most software implementations, going live with HR analytics isn't the end of gathering and implementing business needs. As users employ the software and utilize the analytics, new requirements will emerge that weren't considered earlier. Additionally, ongoing business changes will mean modifications and enhancements to the existing analytics. Sometimes, new analytics are added, and other times, analytics calculations or logic needs to be changed. Whatever changes are made, these need to be communicated to users with the right messaging for users to understand why the changes were made and what benefits they bring. In essence, change management doesn't stop when the technology goes live.
Predicting the future
Another layer of complexity in the learning curve for end users is predictive analytics.
Predictive analytics is a form of advanced analytics that uses both new and historical data to forecast future activity, behavior and trends. Essentially, predictive analytics -- as described -- creates a set of analytics that interprets past data to predict future trends. As such, predictive analytics is only forecasted trends and is not, by any means, accurate predictions or reflections of the future. Noting this, a degree of caution has to be taken when using and interpreting predictive analytics results.
While an understanding is needed to correctly interpret "regular" analytics, a whole new methodology is needed to understand predictive analytics. The models used to generate predictive analytics play a crucial role in determining how predictive analytics is calculated and what it means. The usefulness of predictive analytics varies based on the input measures and the underlying methodology, as well as users understanding what the data is telling them. A degree of change management is needed to ensure users understand how predictive analytics works, how it is influenced by different input measures and how the methodologies define the outputs.
The difficulty with predictive analytics is knowing when to trust the data to guide you and when to allow it to be one source of influence in an overall decision. This brings a degree of difference to the change management approach and a similarly strong use of training as the delivery method of the change.
Why creating a data-driven HR culture is so tough
HR analytics is a unique part of HR, if it can be considered HR at all. The overlap with analytical methodologies and practices makes HR analytics an area that requires a high degree of training and a change in mindset. Everyday HR professionals might be highly familiar with reporting, but analytics is a different beast, despite the similarities seen on the surface. The key factors relating to change management for HR analytics can be summarized as such:
- HR typically doesn't have analytics experience, and in-house analytics experts might not have an HR background.
- Users need to learn analytical methodologies for interpreting analytical data correctly, making sure the right decisions are made with the right interpretation.
- HR doesn't always understand the value of analytics; adoption of technology has been slow.
- Predictive analytics can add another layer of complexity in the learning curve.
- Changes to analytical calculations or logic need to be communicated; a process of communication needs to be in place.