Skip to content
Workforce Analytics: A Model For Predictive & Prescriptive Capability

Workforce Analytics: A Model For Predictive & Prescriptive Capability

Technology enables organizations to collect more data about their workforce than ever before—which makes it more important than ever that we don’t simply collect data just because we can, but that we have an idea of why we are collecting it, what we need to collect, and how it can be used in a positive way for both the organization and the individual employee. Workforce analytics providers, HR personnel, and leaders have an opportunity and obligation to not just measure, research, and deliver insights, but to deliver meaningful business impact.

In this presentation, Brad Wilson, from Perceptyx’s Professional Services team, explains how a mature people analytics practice takes workforce analytics to the next level. Instead of just using data to understand past events, employee predictive analytics uses correlations, trends, and patterns to predict future events—and more importantly, to influence future events and maximize the probability of desired outcomes. Employee responses to questions on census engagement surveys, complementary pulse surveys, and lifecycle surveys can be correlated and compared through cross-survey analytics to tell more compelling stories, deliver more meaningful impact, and empower leaders to make more informed data-based decisions with greater speed and confidence. This article is a summary of the presentation; if you would rather watch than read, click on the video above.

Using Workforce Analytics To Increase Engagement

Our core focus in asking for feedback through employee surveys is: identifying specific elements of the employee experience that have disproportionately positive impacts on our desired people and business outcomes.

Managers are the most direct point of contact between employees and the organization, and they have the responsibility of maintaining and improving employee performance— and retaining employees. To be successful, managers must find a balance between these two concepts.

If managers push too hard on performance, employees get “burned out,” feel unappreciated, and are more likely to leave the organization. But if retention becomes the focus, managers may shy away from the sometimes difficult but often necessary conversations around performance and accountability. When this happens, it breeds a culture of complacency which is demotivating, especially for high performers, and overall performance will decline hurting both the organization and its people.

Employee engagement analytics is the key to enabling both employee performance and retention.

When employees are engaged with the organization and their work, they exhibit increased performance through the application of discretionary effort (i.e., they are willing to go above and beyond). This is the type of environment where top performers thrive, and creating it helps keep them on board. When we create an environment where we can keep our best people and enable them to do their best work, it increases the likelihood of success and progress for both the individual and the organization. That increased success leads to the anticipation of further success, which encourages the further application of discretionary effort. These factors feeding back into one another become an altruistic cycle that builds on itself over time.

Strategic employee surveys, followed by actions informed by the data, are the tools we can use to align limited resources to influence those areas that will have the greatest positive impact on this employee engagement model. Effective people analytics is about identifying those areas that have a disproportionate effect on people, culture, and business outcomes.

Often, leaders and people analytics and HR professionals forget the fact that our own perspective is biased and does not necessarily reflect the experience of people who work for the organization. Strategic planning of survey questions can help us avoid those blind spots to uncover what we really need to know. This approach helps us avoid putting together well-intended—but ultimately misguided and ineffective—strategies for improving engagement. We can use survey data to inform a “mini-max” strategy, which minimizes the negative impact of engagement barriers and maximizes engagement drivers like purpose, growth, culture, and empowerment.

Once we are on the right track with a strategic survey program that reveals the barriers to engagement—and we are addressing those barriers—we can begin to build toward a more mature people analytics practice with predictive and prescriptive capabilities.

Learn more about how to increase engagement in your organization with our free guide, Employee Engagement: Redefined.

A Model For Workforce Analytics Maturity

A fully-realized workforce analytics program with predictive capabilities doesn’t come together overnight. Organizations build their capacity for collecting and working with data over time. These are the stages organizations can expect to go through as they move toward achieving full predictive analytics predictive analytics capabilities:

  1. Reporting and metrics: This stage is past-oriented—organizations look backwards and use the data to explain what happened and why.
  2. Intermediate: In this stage, there is improved data integrity, data integration, and automated reporting.
  3. Advanced: An advanced workforce analytics program enables and includes trend tracking, deeper understanding of the impacts of decisions and actions, the development of special research questions for surveys, and beginning to align analytics to strategic goals.
  4. Strategic: At this stage, analytics work is in alignment with operational metrics and people strategies.
  5. Predictive: Predictive workforce monitoring allows organizations to perform modeling and risk analysis, and have the capability to rate the quality of decisions.
  6. Prescriptive: Full maturity of the workforce analytics program allows confidence in modeling and prediction, and can prescribe behaviors that should be followed to achieve good outcomes and identify where to focus resources. When the organization’s analytics program reaches the prescriptive stage, it has the ability to make proactive decisions and head off problems before they develop.

The goal of all people analytics work is to get to that fully mature, prescriptive program. Our goal ultimately is not to just understand what happened and why, but to be able to influence what will happen in the future. In this sense, predictive analytics isn’t the future: It’s the present, because it gives us the ability to shape the future with the work we do today.

Critical Terms & Concepts For Designing An Analytics Model

Two different factors will affect any particular analytics model. The first of these is the unit of analysis, a term to describe what you’re looking at in the data. It might be an entire business or a workgroup within a business, but for us at Perceptyx, it’s the individual employee. That informs our data structure.

Having a small unit of analysis makes it easier to group data to create larger groups with similar characteristics or responses, and drill deeper to see what other similarities may exist and what the implications may be. We don’t report data at the individual level because we want to always avoid identifying individual employees, but we are able to link that data with demographic data to generate an employee profile.

The other factor is alignment in all the elements of our people analytics work—our research questions, the data collection methodology, and our sample. The management dilemma, which is ultimately what we’re trying to solve, depends on being purposeful in our selection of all of these elements. Our workforce analytics strategy also must align with the organization’s strategic priorities, whether that means addressing high turnover, improving organizational culture, or pursuing other objectives.

Aligning Pulsing Strategies With Purpose

Pulse surveys are useful for doing “temperature checks” on engagement, as well as measuring employee sentiments about events in the employee experience and organizational change. Pulses can be designed in a variety of ways and used for a variety of purposes. Organizations may choose to use one or all of the following pulse strategies:

  1. Time-bound pulse strategy: The frequency of these pulse surveys are determined by the calendar. They occur on a schedule: weekly, monthly, or quarterly, and can include census surveys.
  2. Employee-bound pulse strategy: The timing of these surveys is determined by the individual employee; these pulses are tied to significant events in the employee experience such as onboarding, promotion, or exit from the company.
  3. Event-bound pulse strategy: These surveys are tied to events affecting the organization: mergers and acquisitions, reorganizations, or other big organizational changes. Event-bound pulses are particularly valuable for helping to influence employee perceptions about changes by engaging them in the conversation and giving them a sense of perceived self-determination—and for informing communications from leaders.

The type of pulse strategy will typically also determine which employees are surveyed. Engagement pulses may cover all employees, or a random sample.

Using Cross-Survey Analytics To Gain Additional Insights

Comparing responses across different surveys unlocks additional insights, as it allows employee perceptions to be tracked over time and correlated to additional data points. Through the use of cross-survey analytics, any survey can serve as a predictive analytics survey.

As we noted in a previous article, to study attrition, we can examine employee responses across different surveys. By creating a post-hoc demographic of employees who have left the organization, we can look back and see what their responses were on engagement surveys six months earlier, or even go back and see what they said in their onboarding survey.

The post-hoc demographic is also useful for validating survey items. In our work with companies, we have found that the “intent to stay” question—“I intend to remain with the company for the next 12 months”—has on average 30 points better predictive reliability than other questions designed to elicit the same information, such as “I rarely think of looking for another job.” The limitation of these types of questions is that they treat all voluntary attrition the same.

Insights from one client’s data showed that the number one reason employees cited for leaving was dissatisfaction with the relationship with their manager; 17% of respondents gave this response. Another 13% cited dissatisfaction with the job; 12% had family or personal reasons for leaving, and 8% cited a lack of career opportunities.

Looking back at employee responses on the organization’s census survey, we could see links. Employees who cited their relationship with their manager as their reason for leaving scored lower on items related to respect from manager, support for development, and perception that their manager kept commitments. Employees who left because of dissatisfaction with their job scored lower on items related to systems and processes, communication, and cooperation. Those who cited a lack of career opportunity as the reason for leaving scored lower on performance feedback, support, and recognition, compared with employees who remained with the organization.

When we linked the client’s exit and census data, the big takeaways were that employees who cited frustration with their manager as their reason for leaving had indicated issues with regular feedback, keeping commitments, and recognizing employee accomplishments on surveys months earlier. With this insight, we were able to look through the data to identify manager actions that were linked to increased retention of talent; those actions were integrated into manager training and core competencies developed by the company. We also were able to identify attrition hotspots in the global organization. By developing profiles and attributing recent census data, we were able to identify locations with a higher risk for voluntary attrition, and where intervention should be focused to minimize attrition. This allowed the company to focus its resources on the areas of biggest opportunity for reducing turnover.

Another client focused on employees’ perceptions of inclusion, and integrated engagement survey data with their onboarding survey data. They learned that there were specific early experiences where, if employees were in agreement on a handful of items at 60 days, they were more engaged and connected, and identified more strongly with the organization on subsequent engagement surveys. One of the elements was related to product knowledge and the organization’s key differentiators, which had not been shared in the onboarding process for all employees. After seeing the link between this onboarding experience and later engagement, the client began incorporating it into the onboarding for all employees. It became a strong predictor of long-term retention—but only because the organization was able to link onboarding with engagement pulse data as part of broader listening strategy. Because it was possible to see the correlation, the organization was able to positively influence the employee experience.

Building Engagement With Predictive Analytics

Ultimately, predictive analytics is about identifying and influencing organizational behavior. As such, it is a long-term process that involves several stages. Companies cannot buy engagement with parties, ping-pong tables, or snow cones. Many organizations, when trying to influence engagement, look for a cheap and easy out; this is not only transparent to employees, it is ineffective. A toxic work environment with abusive management can’t be overcome by serving employees free snow cones once a week. True engagement comes from looking for the consistent elements of culture and experience that influence an individual’s perceptions and the way they interact with the organization. (Tweet this!) It’s about focusing on the issues that really matter most to employees in their experience with the organization.

The only way to identify those issues is through employee feedback. But asking for feedback is just the beginning; to really effect a positive change, organizations must be prepared to take action using what they’ve learned, and not only act, but communicate to employees that they’ve been heard and the organization is taking specific actions to address employee concerns. When companies take this approach, and align it so that it reinforces core organizational values, engagement becomes not something that we do to employees—but something we do with them.

Seeing The Way Forward

The Perceptyx platform gives you the flexibility to adapt your listening strategy to rapidly changing real time events—and exert positive influence to attain desired outcomes. Combined with support from our analytics experts, our platform can help you keep your finger on the pulse of your people’s needs, so you can provide the support they need throughout the employee experience. Get in touch to see how we can help your organization navigate and enhance the employee experience.

Subscribe to our blog

Opt-in for our weekly recap and never miss a post.

People Insights Platform

Drive Change, Deliver Impact


Employee surveys to illuminate the employee experience

Learn more about Ask product


Crowdsourced insights to engage your people on the topics that matter most

Learn more about Dialogue product


Lifecycle surveys and always-on listening to keep pace with your people

Learn more about Sense product


360 feedback and Intelligent Coaching to improve manager effectiveness

Learn more about Cultivate product

Getting started is easy

Advance from data to insights to focused action