Why ‘data is just data’ without capable managers – and what HR should do next

Strong management and accurate forecasting are often treated as separate disciplines. However, new evidence suggests they rise and fall together – and that Canadian employers who want better forecasts may need to start by fixing how they manage. 

The research, based on the management and forecasting outcomes of about 8,000 U.K. firms, finds that companies with more structured management practices – including clearer targets, better KPI tracking and data collecting – make smaller errors when forecasting both GDP and sales.  

“We are able to show that firms with higher management scores are significantly more accurate in their forecasts about macro-economic growth (GDP) and their own growth (turnover),” the researchers write, noting that this result is true even after accounting for factors such as location, ownership and age of the company.  

“To put it simply, better managed firms make better forecasts, and as a consequence, better business decisions.” 

Why forecasting is an HR issue 

For Qian Zhang, assistant professor of strategic human resources at the University of Ottawa, forecasting is squarely an HR concern, not just a finance exercise. As she puts it, “Personnel planning is fundamentally a forecasting exercise,” including when to lay off staff and when to upskill internal talent – more important than ever in today’s uncertainty. 

Zhang argues that uncertainty is an “internal capability problem,” not just about weathering external shocks. In practice, that means headcount budgets, reskilling programs and reliance on contingent talent all rest on an explicit or implicit view of the future, and when that view is wrong, employers can find themselves scrambling to fill gaps or carrying payroll costs they can’t support. 

“They can make better forecasts, and they are more confident in achieving the results that are predicted by the forecasting tools,” she says. 

“Therefore, they can make better workforce decisions or personnel-related decisions under uncertainty.” 

The Bloom paper backs up that claim, showing that firms with more structured practices around targets, monitoring and incentives tend to be more productive, more profitable and better at forecasting both GDP and their own turnover.  

Building forecasting into HR systems 

As Tatsuro Senga, coauthor of the study and associate professor of macroeconomics and corporate behaviour at Keio University, explains, when it comes to forecasting, “misalignment can be really bad.” 

For example, when a company plans for 10-per-cent sales growth but only achieves two per cent, the results are “over-capacity – insufficiently huge amounts of capital stock sitting in your factory.” 

This matters most in turbulent periods, Senga says, observing that research shows resilience to be a crucial aspect of organizational success. Well-managed companies, he explains, collect large and consistent data and KPIs, along with frequent performance monitoring, and can “adjust their policies and decisions more frequently and more quickly, and also more accurately.” 

Zhang points out that HR forecasting cannot be bolted on to otherwise weak systems, explaining that HR systems are inherently interconnected with others in the organization – and is also increasingly data-driven. 

“Basically, we are relying on past data, and then we use those data to help the firm predict the future personnel-related needs,” Zhang says. This “AI-driven HR,” she adds, can estimate not only numbers of new hires and attrition rates but also “red flags” identifying potential quitters and external market conditions that will impact the firm. 

However, she is also clear that technology alone is not the answer. Without managers who know how to gather and analyze metrics, she says, “the data is just data.”  

 “HR leaders, they should invest in managerial capability. They should not just be focusing on purchasing those expensive analytical tools,” Zhang says. 

Rather, employers should focus on distinguishing “signal from noise” and avoiding over-customization that can lead unreliable results. 

Making forecasting a shared routine 

Both Zhang and Senga say forecasting should be treated as an ongoing managerial responsibility. In well-managed firms, Zhang notes, “It isn’t just an isolated financial exercise, it’s not just about cost — it’s actually embedded in the regular managerial routines.”  

She suggests that HR build forecasting practices and check-ins into regular leadership discussions around manager tasks and reporting, including performance reviews and HR cycle planning. 

Zhang also wants HR to treat forecasting as “an open system,” meaning HR and managers should regularly and consistently revisit the tracking and forecasting systems. 

“Constantly refine the system, to enhance, to make sure the predictions are up-to-date and using up-to-date information, using up-to-date data,” she says. 

“What matters isn’t just about making forecasts, it’s about systematically learning from the process, and refining the process.” 

HR forecasting to better weather uncertainty 

The forecasting research connects that mindset to concrete episodes such as COVID-19, tracking the managerial and forecasting success of firms as they navigated that disruptive event. 

Senga points out that firms that had better management and stronger, more accurate forecasting capabilities were able to anticipate and profit from the eventual economic rebound and GDP-surge post-COVID, while those without forecasting accuracy missed out on the opportunity.  

“Collecting KPIs frequently, monitoring, and then data collection overall, [they] managed to expect this rebound better than their peers,” Senga says. These companies were ready to meet the sudden market demand with the needed inventory and staff, while more “pessimistic” firms were caught off guard and under-staffed. 

Data, AI and the limits of automation in HR forecasting 

Both experts see a role for AI, but with caveats. For Senga, he cautions against over-reliance on generic tools for data-gathering and analysis.  

“I’m not so optimistic about AI at the moment,” he says, explaining that accurate “micro” or firm-specific forecasting requires accurate, firm-specific data. 

Zhang reaches a similar conclusion from the HR side, adding that manager knowledge and proficiency in data usage is key, with applying past forecasting results to present practices an essential part of the equation. 

“Those tools only work when the manager processes are strong enough to use them,” Zhang says.  

“They should be able to use, to interpret and implement those results from the HR forecasting. That’s very, very critical.” 

Zhang adds that for HR in Canada’s tight labour market, the most important lesson from the forecasting research is around how organizations handle and communicate about uncertainty internally.  

“Top management teams, organizations, they should use uncertainty as a diagnostic, not a failure,” she says.