The HR Technology Revolution Needs a Reality Check Before It Needs a Reboot

The Future of People Management: A Balanced Perspective

McKinsey's recent vision for the future of people management presents a compelling narrative: automate two-thirds of HR tasks, leverage AI for hyper-personalized employee experiences, and free managers to focus on the human elements of leadership. The authors Asmus Komm, Fernanda Mayol, Neel Gandhi, and Sandra Durth paint a picture of frictionless organizations where skills flow seamlessly to high-value work, employees receive Netflix-style personalized development recommendations, and strategic triumvirates of HR professionals operate as value-creation engines rather than cost centers.

This vision deserves serious examination, not dismissal. But it also requires a more critical lens than the report provides. After two decades covering organizational transformation, I've learned that the distance between consulting slide decks and operational reality often measures in light-years, not months.

The Seductive Promise of Technological Solutionism

The article's central thesis rests on a familiar pattern in management thinking: technology will solve our human capital challenges. We've heard variations of this promise before with enterprise resource planning systems in the 1990s, big data analytics in the 2010s, and now AI. Each wave brought genuine improvements, but rarely at the scale or speed predicted, and often with unintended consequences that took years to address.

McKinsey's research suggesting that two-thirds of current HR tasks can be automated deserves particular scrutiny. This figure appears to derive from technical automation potential rather than practical implementation reality. A 2023 study from MIT's Initiative on the Digital Economy found that while many tasks are technically automatable, actual automation rates lag significantly behind potential due to factors including cost, change management challenges, regulatory constraints, and the complexity of integrating systems across organizational silos.

The telecommunications coaching engine example provided in the article illustrates both the promise and the problem. Training an AI system on call center transcripts and KPIs to provide personalized coaching suggestions sounds efficient. But this approach raises questions the article doesn't address: How do we ensure the AI doesn't simply optimize for metrics that may not capture quality? What happens when the AI's recommendations conflict with an employee's lived experience or a manager's judgment? Who bears responsibility when AI-driven coaching leads to poor outcomes?

The Data Foundation That Doesn't Exist

The vision of hyper-personalized employee experiences powered by integrated data lakes assumes a level of data infrastructure that simply doesn't exist in most organizations. Deloitte's 2024 Global Human Capital Trends report found that only 12 percent of organizations rate their people analytics capabilities as excellent, and just 8 percent have successfully integrated HR data with broader business data systems.

Even technology-forward companies struggle with this integration. Microsoft's experience with its Workplace Analytics product offers instructive lessons. Initially launched with ambitious goals around productivity insights, the company had to significantly revise its approach after employee pushback around privacy concerns and the realization that raw productivity metrics often missed crucial context. The company learned that data integration was the easy part; developing analytics that employees trusted and that actually improved decision-making proved far more difficult.

The article's call for a "single data lake for all business information and a stable, integrated IT system in the cloud" glosses over the reality that most large organizations operate with dozens or hundreds of legacy systems, each containing crucial data but built on incompatible platforms. The cost and complexity of integration often exceeds the projected benefits, which is why so many digital transformation initiatives stall or fail outright.

The Automation Paradox in People Management

There's an inherent tension in the article's argument that deserves more attention: the claim that automating HR tasks will free managers to be more human and provide more individualized attention. Research on automation in professional work suggests a more complex reality.

A comprehensive study published in Administrative Science Quarterly examined what happened when organizations automated routine professional tasks. Rather than automatically redirecting their time to higher-value activities, many professionals struggled to redefine their roles, felt deskilled, and experienced decreased job satisfaction. The automation of routine tasks didn't automatically generate the capability or inclination to perform more complex, human-centered work.

The assumption that managers will naturally redirect time saved through automation toward coaching and empathy-building activities ignores how organizational incentive systems actually work. If organizations continue to reward managers primarily for hitting quarterly targets rather than for developing their people, technology-freed time will flow toward whatever drives bonuses and promotions, not necessarily toward humanness.

Moreover, the article's vision of AI-enabled personal agents handling routine interactions may actually reduce rather than enhance the human element of work. Research from the University of Pennsylvania's Wharton School on AI-mediated communication found that while efficiency increased, the quality of relationships and trust-building decreased when interactions were filtered through automated systems. The small moments of human connection that build workplace relationships often occur during those supposedly routine interactions that the article assumes should be automated away.

The Ethics of Algorithmic People Management

The article mentions ethics only in passing, suggesting that HR leaders "may need to provide an ethical sounding board for senior leadership on technology usage." This dramatically understates the ethical challenges of AI-powered people management systems.

Consider the "opportunity marketplaces" the article celebrates, where AI matches employees to internal opportunities based on skills and potential. Research from scholars including Safiya Noble and Ruha Benjamin has documented how algorithmic systems consistently reproduce and often amplify existing biases around race, gender, age, and other protected characteristics. Amazon's abandoned recruiting AI, which taught itself to penalize resumes containing the word "women's," offers a cautionary tale about the challenges of bias in seemingly objective systems.

The vision of continuous monitoring and measurement that enables "real-time" workforce planning raises profound questions about employee autonomy and privacy. When does helpful coaching become intrusive surveillance? At what point does data-driven personalization become manipulation? The article's telecommunications example of AI tracking employee performance through call transcripts should give us pause rather than inspiration. Do employees have meaningful consent in such systems? Can they opt out without career consequences?

European regulatory frameworks like GDPR already impose significant constraints on employee data collection and algorithmic decision-making that aren't mentioned in the article. The EU's proposed AI Act would classify many HR applications as "high-risk" AI systems requiring extensive documentation, human oversight, and accuracy standards. Organizations pursuing the McKinsey vision must reckon with a global regulatory landscape that is moving toward restricting, not enabling, many of these applications.

The Implementation Gap That Swallows Strategies

The article's segmentation of organizations into strategists (70 percent), scalers (25 percent), and visionaries (5 percent) actually reveals a damning reality: after decades of HR technology investment, 95 percent of organizations haven't successfully implemented even the previous generation of people management technologies.

This isn't because they lack vision or investment. It's because organizational change is fundamentally difficult, especially when it involves shifting power dynamics, redefining professional identities, and requiring new capabilities from people at all levels.

The proposed "strategic triumvirate" of people strategists, people scientists, and people technologists sounds compelling but requires organizations to attract and retain talent with highly specialized capabilities in an intensely competitive market. Where will companies find data scientists who also understand organizational psychology? How will they retain AI specialists who could earn multiples of typical HR salaries in product or engineering roles?

The article acknowledges this challenge by noting increased demand for "data and HR technology specialists" but doesn't grapple with the supply constraints or the cultural challenges of integrating these very different professional orientations into a cohesive function. Having observed numerous attempts to build hybrid teams of specialists, I can attest that putting technologists and traditional HR professionals in the same organization chart doesn't automatically create productive collaboration.

What Organizations Actually Need

None of this is to argue against leveraging technology to improve people management. The status quo in most HR organizations does need disruption. But we need a more grounded, ethically informed, and realistic approach than the article provides.

First, organizations should start with the problems they're actually trying to solve rather than the technologies they want to deploy. The article's vision is solution-driven, starting from technology capabilities and working backward to applications. Better practice starts with clearly defined business challenges, explores multiple potential solutions, and then considers whether and how technology might help.

Second, successful technology implementation in people management requires equal investment in change management, capability building, and ethical frameworks. Research from BCG on HR transformation found that projects with strong change management support were 3.5 times more likely to succeed than those focused primarily on technology deployment. The article mentions this briefly in its closing principles but doesn't integrate it into the core vision.

Third, organizations need to think more critically about what should be automated versus what should remain human, even if it's less efficient. Not every interaction needs to be optimized. Some inefficiency in people management creates space for relationship-building, learning, and the kind of informal knowledge transfer that formal systems can't capture. The goal should be appropriate automation, not maximum automation.

Fourth, any vision for AI-powered people management must start with robust ethical frameworks, not add them as an afterthought. This means involving employees in the design of these systems, building in transparency about how algorithms make decisions, creating meaningful human override capabilities, and regularly auditing systems for bias and unintended consequences.

The Question of What Makes Work Human

The article's title promises an approach that is "more personal, more tech, more human." But it's worth interrogating whether these goals are actually compatible in the ways suggested.

The vision of hyper-personalization borrows heavily from consumer technology, where personalization means tailoring content and recommendations to individual preferences. But the employment relationship is fundamentally different from a consumer relationship. Employees don't just consume organizational offerings; they co-create value, negotiate terms, and exist in ongoing relationships of mutual dependence and, ideally, trust.

Research from organizational scholars including Amy Edmondson and Edgar Schein suggests that what makes organizations human isn't just personalized experiences but psychological safety, authentic relationships, and the ability to bring one's full self to work. It's not clear that AI-mediated interactions and constant monitoring enhance these conditions. Indeed, they may undermine them.

The article quotes economist Andrew Scott saying that "as machines get better at being machines, humans have to get better at being more human." This is exactly right, but I'm not convinced the article's vision supports this goal. Being more human in organizations means creating space for messiness, learning from failure, building trust through repeated interactions, and sometimes choosing inefficiency in service of relationship-building.

These very human capabilities may actually require protecting certain domains from automation and measurement rather than subjecting everything to AI-powered optimization.

A More Balanced Path Forward

Organizations do need to evolve their people management practices. The traditional HR operating model developed in the industrial era doesn't fit the realities of knowledge work, distributed teams, and the increasing importance of organizational culture and employee experience to business performance.

Technology will certainly play a role in this evolution. AI can help with genuinely routine tasks like answering basic benefits questions, scheduling interviews, or flagging compliance issues. Analytics can provide useful signals about organizational health and help target interventions. Platforms can make it easier for employees to access learning resources and connect with opportunities.

But the transformation organizations need isn't primarily technological. It's cultural, structural, and human. It requires rethinking how work is organized, how value is measured, how power is distributed, and what obligations organizations have to employees and vice versa.

The most successful organizations will be those that use technology selectively and thoughtfully in service of clear strategic goals, while maintaining focus on the fundamentally human work of building cultures where people can thrive. They'll invest as much in developing managers' coaching and leadership capabilities as they do in implementing AI systems. They'll create ethical frameworks before deploying algorithmic decision-making tools, not after.

Most importantly, they'll recognize that there are no shortcuts to building high-performing organizations. The vision of automating two-thirds of HR and freeing up capacity for strategic work is seductive precisely because it promises transformation without the messy, difficult work of actually changing how people think, behave, and relate to one another.

But organizational capability isn't a technology deployment problem. It's a human development problem. And that will remain true regardless of how sophisticated our AI becomes.

Recommendations for Business Leaders

For executives and HR leaders considering how to evolve their people management approaches, I offer these recommendations:

• Be deeply skeptical of claims about automation potential that don't account for implementation complexity, change management requirements, and organizational readiness. Technology vendors and consultants have every incentive to paint optimistic pictures. Reality tends to be messier and slower.
• Before investing in sophisticated HR technology, ensure you have the foundational data infrastructure, governance processes, and analytical capabilities to use it effectively. Too many organizations buy enterprise platforms that become expensive shelfware because they lack the fundamentals.
• Involve employees as partners in designing new people management approaches, not just subjects of them. The best solutions emerge from deep understanding of how work actually happens, not from theoretical models imposed from above.
• Develop ethical frameworks and governance structures for AI in people management before deployment, not in response to problems. Include diverse perspectives, build in transparency and contestability, and plan for regular auditing and oversight.
• Invest at least as much in developing human capabilities as in deploying technology. The constraint on organizational performance is rarely lack of technology but lack of capability to use available tools effectively and to work together productively.
• Maintain healthy skepticism about any vision of transformation that promises revolutionary change through primarily technological means. The history of management fads suggests that the transformations that last are those that align technology with deeper changes in strategy, structure, culture, and capability.

The future of people management will certainly involve more sophisticated technology. But whether it becomes more personal and more human depends less on our AI capabilities than on the choices we make about what to automate, what to measure, and what to protect as essentially human domains that resist and should resist optimization.

Those choices require wisdom, not just technological sophistication. And wisdom, unlike automation potential, can't be reduced to a percentage or deployed through a platform upgrade.