The Skills Data Playbook from the Pros: 6 Ways to Find Hidden Talent and Upskill Faster:

The technological disruptions impacting businesses over the past few years are arguably more impactful than ever before. In fact, 56% of business leaders say that addressing AI- and tech-driven skill shifts is business-critical. Yet, only 15% feel that their workforce planning capabilities are sufficient to tackle these challenges, according to Visier’s Strategic Workforce Planning in the AI Era report.

While our report goes into far more detail, the obvious takeaway here is that poor workforce planning has steep consequences. Skills mismatches and worker shortages are eroding productivity, according to 52% of companies. In addition, 54% point to missed opportunities and project delays, and 37% say that recruitment and training costs are rising. 

During a panel discussion at the fall Global People Analytics Summit, leading workforce planning experts shared how their organizations are using workforce analytics to identify development opportunities, surface hidden talent, and conduct qualitative analysis at scale.

Skills intelligence platforms show promise—with some caveats

Organizations are beginning to deploy skills intelligence platforms, but with mixed results depending on use cases. As Lei Pan, PhD, Global Head of People Analytics at FrieslandCampina, explained during the panel, individually-focused applications like recommending internal jobs to employees, suggesting internal candidates to hiring managers, or matching people to project opportunities, work remarkably well even with accuracy rates of 50-60%.

She explains: "No one is going to be very upset if you get ten job recommendations and only five or six make sense.” These impacts, she says, are a game changer.

However, when organizations try to use skills data to make strategic decisions like determining whether to buy talent from the market or upskill existing employees, the level of accuracy required is much more demanding, Pan says. In these cases, additional work needs to be done to increase accuracy.

Our data supports this, noting that while 44% of organizations use AI primarily for identifying skills gaps and demand analysis, only 31% leverage predictive forecasting and a mere 17% apply scenario modeling. 

There's significant untapped potential in using AI for forward-looking talent development planning.

6 ways to surface hidden talent

Every organization has hidden talent whose skills and expertise are overlooked for a variety of reasons. This is where workforce analytics can deliver big benefits, identifying the invisible contributors whose impacts can deliver clear value. 

For instance, Pan described a network analysis project where the company simply asked employees: "Where do you get help if you have a difficulty?" The results were illuminating, identifying hidden influencers that might otherwise have gone undiscovered—employees who weren’t the most vocal or visible, but who were critical to getting work done. 

People analytics, says Pan, “brings the less visible people to the table."

Here are a few ways the panelists are using analytics to make the invisible visible: 

Look beyond traditional performance metrics

Despite the potential for these insights, access to productivity and output data remains surprisingly limited. In fact, 71% of leaders lack appropriate data on performance metrics. It's critical to look at performance consistency over time, willingness to learn (tracked through hours spent on future-critical skills), and collaboration network analysis to identify the "go-to people" who influence others.

What you can do: Don't wait for perfect performance data. Use proxies like learning hours and network centrality to identify high-potential talent.

Identify high-impact leavers before they become “regrettable exits”

Poonam Sirigidi, Senior Director of People Insights at Pfizer, shared how her organization took this concept even further to identify high-value employees at risk of leaving before they depart. "We’re all familiar with the term ‘regrettable exits’,” she noted, but these people are identified after they’ve already given notice or left the company. “What we’ve started doing is giving HR and managers access to a list of colleagues who will be a regrettable loss if they leave."

This approach combines multiple data sources related to potential resignations, consistent performance, potential lost impacts, leadership behaviors, and engagement data. "This data is refreshed on a daily basis because things are constantly changing and evolving," Sirigidi noted. The result? "We’re seeing so many use cases where people are leveraging this input, whether it's for talent decisions, investment decisions, or other things."

What you can do: Build a "regrettable exits" list combining performance, engagement, and flight risk signals. Refresh it daily and give managers access to intervene before talent walks out the door.

Turn restructuring into redeployment opportunities

Pfizer implemented a "reverse fetch" program during restructuring, where skills alignment tools are used by employees to help them find new internal roles rather than causing them to leave the organization

What you can do: During workforce changes, flip the script. Instead of just matching jobs to candidates, let affected employees use skills tools to discover where they fit internally.

Scaling qualitative analysis: How generative AI changes the game for “dead data”

"Dead data." That's what Mads Frank, Director of Behavioral Sciences at GSK, calls the employee objectives, feedback, and comments that sit unused in free-text fields. But today, GenAI is bringing it back to life.

Frank described how this transforms HR's ability to understand employee experience: "We've seen really good results in improving how we do qualitative research and looking at qualitative data—for example, objectives that people put in for the year ahead,” he said. These inputs, he said, may have been “written down in free text, traditionally something that's just left there as dead data.” Now, though, he says, “suddenly we have the ability to look at an aggregate and identify trends."

Frank calls this "passive listening." Instead of launching new surveys, GSK analyzes the qualitative data employees naturally generate. It's data that's been sitting unused.

The impact on decision-making has been profound. "We did some free text work where we gathered people's feedback on cultural engagement," Frank shared. "It changes the way you interact with lived culture in a much deeper, more meaningful way. You get examples, you get experiences, and it drives conversations that I've never seen before in HR."

What you can do: Start with your existing free-text data (employee objectives, performance notes, intranet comments). Use GenAI to surface insights from qualitative data that's currently sitting unused, before launching new surveys.

Make AI your co-pilot, not your captain

Christopher Cerasoli, Senior Associate Director at Boehringer Ingelheim, emphasized that while Gen AI currently performs better with qualitative than quantitative data, the key is knowing where to apply it. The ability to combine qualitative inputs using generative AI with quantitative data can yield significant results, he says.  

Tools like Visier's Vee, for instance, offer conversational interaction with workforce data that can make critical insights more accessible to HR business partners who may not have deep analytical expertise but need to make data-informed decisions quickly.

What you can do: Pair GenAI-analyzed qualitative insights with your quantitative data. Use AI where it excels (understanding free text, surfacing themes) and combine those findings with traditional metrics for a complete picture.

Read more early-adoption stories about Visier's AI Assistant Vee

Implementation guidance: Stop delaying your start for the ideal state

For organizations wondering where to begin, the panelists offered grounded advice. Cerasoli, for instance, invoked Arthur Ashe, saying: "Start where you are, use what you have, do what you can."

That might mean using AI to quickly mine existing job descriptions and development objectives rather than conducting weeks of manual analysis. Or creating simple talent flow visualizations—Cerasoli noted that charts showing internal mobility patterns captivate business leaders. It could also mean embracing "progress over perfection," prioritizing speed and direction over waiting for perfect data.

But the first step for organizations, Derler emphasized, is getting the data sorted, cleaned, organized, and integrated. “The AI layer would then, ideally, come later to ease access to all of the goodness that's underneath."

Removing bias from processes is on top of everyone's mind, Sirigidi said. The solution? Democratizing data. "You cannot be the gatekeepers of this data. You need to put it in the hands of people who make talent decisions."

What you can do: Clean and organize your people data before layering on AI tools. Then make sure you’re democratizing access (Vee can help!). Put those insights in the hands of managers and business leaders who make day-to-day decisions, not just HR. 

The path forward

The panelists agreed on what's ahead: AI will augment rather than replace human judgment in talent decisions, and managers will become more critical than ever as the bridge between business strategy and people development.

The opportunity is clear. Organizations already have hidden talent, underutilized qualitative data, and the tools to surface both. The question isn't whether to start. It's whether you'll act before your competitors do.

For more insights on workforce planning and how to mine for upskilling in your organization, check out our latest report "Strategic Workforce Planning in the AI Era."