One of my favorite cartoons shows an executive looking across his desk at a downtrodden employee saying, “You do have good ideas, but we don’t do things in this company just because they happen to be a good idea”. This cartoon is funny because many of us can relate to being that employee. Innovations in technology constantly create new ways for companies to improve the nature of work. But the fact something is possible and valuable to do does not mean companies will do it.

People tend to view work practices through a lens of “familiar is easy, and change is hard”. Technological advances rarely lead to large scale innovation in work practices without some triggering event tied to external competitive threats or internal business disruptions. React and respond is one way to change, but it is rarely the most efficient and effective. A better approach is to proactively use technology to improve work by enlisting cooperation from senior leaders, middle managers, and frontline employees. Each of these groups has a different perspective on change and each plays a different role in transforming the experience of work. But what they all have in common is the human tendency to view change more in terms of avoiding risks than realizing opportunities^[i].

Senior leaders fear change will waste resources. The desire to stick with familiar technology and existing work practices is often the strongest among senior leaders^[ii]. The reluctance of senior leaders to change is related to their seniority. Leading organizations requires complex skills that are acquired through job experience^[iii]. Because it takes time to gain this experience, most senior leaders have worked for at least 20 years before achieving senior executive status^[iv]. This experience makes them more capable leaders, but it also makes them more doubting of the value of technological innovation compared to their less experienced colleagues. Anyone with 20 years of corporate experience has inevitably witnessed multiple failed technology implementations that did not deliver on the changes they promised. Senior leaders tend to be skeptical towards technology, but it is a healthy skepticism1.

The people with the greatest capacity to change work methods are senior leaders but they can also be the most resistant to change. The behavior of senior leaders largely defines the culture of a company. Because leaders set the culture, they tend to be comfortable with “the way we do things.” They have little incentive to try new work methods. And there may be no one else in the company positioned to drive them to change. As a result, companies persist in doing things the way they’ve always been done simply because leaders have always done it that way. Even if technological innovations have eliminated the constraints the led to people working that way in the first place (e.g. requiring people to commute to offices every day despite it no longer being necessary to do their jobs).

Convincing senior leaders to invest in work technology requires connecting solutions to things they care about. Part of this is showing how changing work methods can positively impact the growth, profitability, and agility of companies. Another approach is linking changes to issues that are personally meaningful to leaders. For example, one factor that lead companies to invest resources to create more equitable employee experiences for women are concerns senior male executives had for the career experiences of their daughters^[vii]. A third approach is making leaders aware of how clinging to past work practices is hindering their personal leadership ability. Leaders cannot fully inspire employees to execute the company’s strategy if employees feel unfairly constrained or needlessly inconvenienced by outdated work methods.

Middle managers fear being overwhelmed by change. One of the most critical, challenging, and underrated roles in a company is that of the middle manager. Managers are the link between senior leaders who set strategies and employees who execute them. Leaders define the business goals the company needs to achieve. Employees deliver results. Managers ensure that the results employees deliver align with the goals leaders need to achieve. The job of a manager is fundamentally about two activities. First, making staffing decisions to build a team capable of executing the company’s strategic goals. Second, aligning company business objectives with what employees want to do and what they can do. And this is only part of their job. Very few people are employed solely to be managers. Most are also expected to perform a range of individual contributor tasks and functions that have nothing to do with managing others. Given the scope of their job responsibilities, managers tend to work longer hours than other employees^[viii].

Managers are highly sensitive to any changes that might make it more difficult to perform their work or take time away from core job duties^[ix]. Enlisting manager support to improve work practices requires showing how the change will help them be more effective managers. This includes being sensitive to the time pressure managers are under. Avoid presenting technological change as a new initiative that will take managers away from their ongoing work responsibilities. Instead, present the change as a way to improve things managers are already doing. If you really want to get support from managers, show that the change will require less time than what they are already doing. Also make sure they have access to training if they need it and recognize and reward them for doing it. One of the more effective deployments of technology I have seen started with the CEO standing in front of the company’s managers and saying, “This new solution is not asking you to do anything you should not already be doing. This just makes it easier to do it and reminds us if we forget to do it.” He closed by showing how he used the solution commenting that, “If your CEO can find time to do this then so can you”.

Frontline employees fear change will make their jobs worse (or eliminate their job entirely). Most full-time employees devote at least 35% of their waking hours to work, potentially much more if they have long commutes. People literally give over a third of their conscious life to the organization. What employees want from work is to know that they are spending their lives doing something meaningful that allows them to achieve their personal and professional goals. The vast majority of employees want to do a good job. They show up with a desire to do effective work, help others, and achieve their career goals. Yet companies often make it incredibly difficult for employees to do these things. Work should be about getting important stuff done. But for many employees, work is about showing up and waiting to be told what to do, trying to track down the people, tools, and resources needed to do it, and then learning that what they were doing didn’t matter that much anyway.

Employees tend to see changes as either requiring them to spend time doing things they do not want to do or taking away resources and freedoms that they value (including their job itself). To get employees excited about using technology to change work, focus on things that demonstrate that the company values and appreciates their time, skills and potential. What employees want are things that reduce time spent on repetitive or administrative tasks, help them do their jobs more effectively, build and strengthen their relationships with others, create a stronger sense of employment security, and enable them to achieve their work and non-work goals. These things don’t just help employees to be successful, they help companies to be successful.

We excel at not doing things that are good for us

Much of the $350 billion global weight loss industries is driven the by the fact that people struggle to do things that are good for them^[i]. It can be very difficult for people to change even when they know it is in their best interest. Because organizations are groups of people, resistance to change tends to get stronger as companies grow in size. Advances in technology are making it easier to change and improve the experience of work. Technology enables change but it does not create change. It is the employees in organizations, whether senior leaders, middle managers, or frontline employees who ultimately determine whether change happens.

Note: portions of this article were originally included in my book Talent Tectonics

^[i] Ltd, RAM (2021, July). Global weight loss products and services market 2021–2026. Research and Markets Ltd 2022. https://www.researchandmarkets.com/reports/5393446/global-weight-loss-products-and-services-market

^[i] Slovic, P., Peters, E., Finucane, M. L., & MacGregor, D. G. (2005). Affect, risk, and decision making. Health Psychology, 24(4, Suppl), S35–S40. https://doi.org/10.1037/0278-6133.24.4.S35

^[ii] Oreg, S. & Berson, Y (2017). Leaders’ characteristics and behaviors and employees’ resistance to organizational change. Academy of Management Proceedings. https://journals.aom.org/doi/epdf/10.5465/ambpp.2009.44259813

^[iii] Yukl, G. (2012). Effective leadership behavior: What we know and what questions need more attention. Academy of Management Perspectives, 26(4), 66–85. https://doi.org/10.5465/amp.2012.0088

^[iv] Crist|Kolder Volatility Report. (2021). https://www.cristkolder.com/volatility-report/

^[v] Morris, M. G., & Venkatesh, V. (2006). Age differences in technology adoption decisions: Implications for a changing work force. Personnel Psychology, 53(2), 375–403. https://doi.org/10.1111/j.1744-6570.2000.tb00206.x

^[vi] York, J. (2017, March 8). The millennial expectation of technology in the workplace. Forbes. https://www.forbes.com/sites/paycom/2017/03/08/the-millennial-expectation-of-technology-in-the-workplace/?sh=9e0df614a507

^[vii] Dahl, M. S., Dezső, C. L., & Ross, D. G. (2012). Fatherhood and Managerial Style: How a Male CEO’s Children Affect the Wages of His Employees*. Administrative Science Quarterly, 57(4), 669–693. https://doi.org/10.1177/0001839212466521

^[viii] Brett, J. M., & Stroh, L. K. (2003). Working 61 plus hours a week: Why do managers do it? Journal of Applied Psychology, 88(1), 67–78. https://doi.org/10.1037/0021-9010.88.1.67

^[ix] Cavanaugh, M. A., Boswell, W. R., Roehling, M. V., & Boudreau, J. W. (2000). An empirical examination of self-reported work stress among U.S. managers. Journal of Applied Psychology, 85(1), 65-74. doi:10.1037/0021-9010.85.1.65

^[x] Sikora, D. M., Ferris, G. R., & van Iddekinge, C. H. (2015). Line manager implementation perceptions as a mediator of relations between high-performance work practices and employee outcomes. Journal of Applied Psychology, 100(6), 1908–1918. https://doi.org/10.1037/apl0000024

^[xi] Meinert, D. (2014). Five types of bad bosses: what HR can do to improve the performance of the most common types of problem managers. HR Magazine. Society of Human Resource Management.

Candidates are the most important part of hiring. Yet they are treated more like commodities than people in the applicant funnel model, as reflected in phrases such as “filling candidate pipelines with higher applicant volumes”. At each step in the funnel candidates are told what they have to do to get past the next hurdle often with little explanation of why they have to do it. They are frequently left in the dark about their performance on screening exercises, the overall status of their application, when they are likely to hear from the company, and if they have a realistic likelihood of receiving a job offer. They rarely receive constructive information explaining why they did not get hired. Instead of feeling like a partnership between companies who need people and people who need jobs, the hiring process feels like an examination where candidates must prove themselves worthy of employment.

The applicant funnel is breaking

The traditional applicant funnel model of hiring has never worked that well from a candidate perspective. We are now reaching a point where companies can no longer ignore its inherent shortcomings. The applicant funnel methodology is beginning to fail for companies and candidates alike because of the following factors:

• Applicant volumes clogging the funnel. The advent of AI tools enables candidates to apply to jobs with little to no effort. As a result, company hiring funnels are being overwhelmed by applications. It is not uncommon for companies to receive over 1,000 applications for a single role. This makes it difficult for companies to find the best candidates. It also makes it difficult for the best candidates to get noticed and find a job.

• Susceptibility to candidate fraud. Applicants are using AI solutions to pass job screening and selection tests by exaggerating their skills, experience, and interests. Even more troubling, criminal organizations are using AI to create false candidates so they can steal company secrets or financial resources. Increasing use of automated candidate screening in response to applicant volumes clogging applicant funnels is likely to further increase candidate fraud. Job seekers may feel the only way to get through company screening methods is to falsify their application information, and standardizing selection processes allows criminals to systematically develop ways to get around hiring tests and screens.

• Rewarding quality of candidate vs quality of hire. The majority of hiring technology solutions use data to predict people’s potential to be hired into the job. They do not use data to predict whether people are successful after they are hired. For example, many companies use structured internet search tools to identify candidates who are likely to do well in interviews. But there is big difference between being a good interviewer (aka, quality of candidate) and being a good employee (aka, quality of hire).

• Losing high potential candidates. Applicant funnels filled with large numbers of unqualified or fraudulent applicants increases the risk of overlooking and losing high potential candidates. The quality of who gets hired into a job is entirely determined by the quality of who applies and their willingness to persist through the hiring process. The cost of losing high potential candidates is increasing due to growing shortages of skilled employees. It is one thing to lose a qualified candidate if there are other equally qualified candidates in the applicant funnel. It is another if that candidate is the only qualified person who has applied.

• Inefficient utilization of data. At its core, hiring is about optimizing a series of decisions. On one side, are decisions made by companies to consider and hire candidates. But equally important and often overlooked, are decisions made by candidates to apply for a job, complete the hiring process, and accept the offer. Traditional applicant funnels collect a lot of data about candidates that could be used to improve subsequent company and candidate decisions, but much of it is collected in a sub-optimal way and used in a limited fashion. For example, while many companies talk about using candidate data collected during the hiring process to guide employee onboarding activities, relatively few companies actually do it.

Replacing the Applicant Funnel

Increasing applicant volumes, growing candidate fraud, failing to measure quality of hire, losing high potential candidates and poor use of data are collectively breaking the applicant funnel model. The question is what can companies do instead? The answer lies in reframing how we think about the process of hiring to put more emphasis on candidate experience and data utilization. This includes:

Viewing hiring as a tournament not a test. Applicant funnels were not created to help candidates build their careers. They were built to help companies find employees. While careers of some candidates benefit from applicant funnels, this is more a byproduct than a goal of the overall process. Fixing hiring requires reframing hiring as a partnership between candidates and companies where the needs of both parties are considered at every step of the process.

The applicant funnel model treats hiring as a series of tests where candidates must pass hurdles to prove themselves worthy of getting the job. This is not the right way to think about hiring, particularly for jobs where there are more applicants than open positions. Hiring is a tournament between candidates to demonstrate why they are the best fit for the role compared to other candidates. What makes tournaments work in sports such as tennis is participants know the rules, understand how they will be judged, and trust that everyone is being evaluated on the same criteria. They also know who won and why. Understanding the outcome provides a sense of fairness and informs participants of what skills to build to increase their likelihood of winning future tournaments. Tournament organizers also know that to attract the best players they must ensure participants have a positive experience during the tournament.

High touch recruiting processes for senior roles share many of the characteristics found in well-run tournaments. Recruiters personally contact qualified candidates and often inform them if they are one of several candidates under consideration. The steps and criteria used to evaluate applicants are explained so candidates know how to prepare. Efforts are made to ensure candidate feel respected and supported throughout the process. Candidates who do not get hired are given information about why they were not selected. Contrast this to the experience of people applying for high volume roles. Applicants for these jobs may get little information about the hiring process beyond an online set of job requirements and the steps needed to apply. Even if companies share information about how they make selection decisions, candidates are rarely given an opportunity to ask questions about the process. And little to no information is given to candidates that would help them understand why someone else got the job instead of them.

Helping candidates succeed. Hiring can be thought of as two phases. Phase one involves marketing jobs to candidates in order to get people to apply (aka, “filling the funnel”).[1] Phase two focuses on running candidates through a series of exercises designed to identify who should receive job offers. These two phases feel very different from a candidate perspective. Phase 1 treats the candidate as a customer. During this phase the company tries to make itself look like an appealing place to work. Phase 2 treats candidates like questionable merchandise, where they must convince the company they are worthy of being offered a job. This frequently results in application processes that give candidates little sense of agency, support or recognition.

Companies will never hire the people who do not apply. When a person decides to apply for a job their potential value to the organization changes categorically from none at all to some value greater than zero. Hiring processes should be designed with this in mind. This means doing things to ensure candidates know the company values their time and appreciates their interest. Hiring processes should help candidates present the best possible version of themselves provided it is rooted in reality. For example, imagine a job required a specific skill that a candidate possessed but that was not listed on their resume. An effective hiring process would alert the candidate to this missing qualification and give them an opportunity to correct it. Otherwise, the company will lose this person not because they lack the skill, but because they failed to list the skill in their job application.

Evaluating hiring based on successful employees not successful candidates. The term “quality of hire” refers to the value employees provide to a company after they are hired. It is a function of selecting candidates using data that predicts post-hire outcomes such as job performance and retention. What most hiring technology systems actually measure is “quality of candidate”, which is the likelihood that an applicant will pass a company’s selection process. The majority of hiring technology solutions use data to predict people’s potential to be hired into the job. They do not use data to predict whether these people are successful after they are hired.

There is a big difference between being a good candidate and being a good employee. A good candidate is someone who has the skills, social connections and attributes to get hired. A good employee is someone who has the skills, social acumen, and attributes to be successful in the job. These things often overlap, but they are far from the same thing. A socially awkward but technically skilled applicant might come across as a poor candidate in a job interview but be an exceptional employee when placed in the job. Hiring processes should be designed so they are able to identify individuals who may be poor candidates but exceptional employees.

What about candidate fraud?

Creating hiring processes using a hiring tournament model that gives candidates more personal attention, support and agency to ask questions and share information about themselves can also help deter candidate fraud. It is easier fake standardized screening tests than it is to fake conversations, particularly when those conversations can be compared to verifiable, unique job history and personal reference information. The challenge companies face is how to engage candidates in a meaningful, authentic way given limited recruiting resources. This challenge is one of the things that led to the creation of impersonal applicant funnel methods in the first place. What has changed is the ability to augment human conversation with AI enabled communication. The critical point is using AI to amplify human connection between candidates and companies without totally replacing it.

Replacing applicant funnels with hiring tournaments

The use of applicant funnel processes lies at the heart of much of the criticism toward hiring in general. The solution is to stop building processes that focus on company needs over candidate interests, and that emphasize quality of candidate over quality of hire. It is time to eliminate applicant funnels that treat candidates as faceless commodities and replace them with hiring tournaments that support and respect them as unique individuals. This change is within the capabilities of modern technology[2]. It will take effort but will have tremendous benefits for candidates and companies alike. And sticking with the status quo will only become more untenable as time goes on.

[1] Marketing jobs to potential candidates is a critical issue for successful hiring, but it is not the focus of this article.

[2] This was not true before cloud technology and AI data management solutions became widely available.

“Work is becoming physically easier but psychologically more difficult” – from the book Talent Tectonics

The field of occupational health arose in the early 20^th century in response to changes in work caused by the technology fueling the industrial revolution. Companies that increased employee productivity by automating physical tasks using steam powered machines often created working conditions described as “dark Satanic mills”. Accidental employee death and dismemberment was accepted as a cost of doing business. Regulations focused on ensuring worker safety arose slowly over decades in response to societal pressure placed on companies to protect the health and wellbeing of employees. Modern employees benefit from over 100 years of occupational health research focused on understanding how to perform manual tasks in ways that minimize risks to worker health. But what happens when we start automating mental tasks at a similar scale?

Recent innovations in AI have led to rapid technological automation of mental tasks related to reading and interpreting information, solving complex problems, creating written documents and imagery, monitoring and providing feedback on performance, and developing software applications. Occupational health research historically focused on understanding how technology impacts performance of physical tasks and effects this has on employee wellbeing. We have far less knowledge of the impact technology has on employee wellbeing when it is used to automate and augment performance of mental tasks. Where we are with AI now reminds me of where society was in the early 2000s during the initial days of social media technology. Many of its immediate benefits are clear, but much less is understood about its potential long-term negative impacts.

This places us in a tricky situation. Companies must embrace use of AI to redesign work if they are to remain competitive. But because AI is so new, we do not fully understand the risks it may pose to employee wellbeing. Fortunately, some early research is starting to provide insights into what these risks might be and how companies can manage them^[1]. The following are thoughts on seven types of risks identified through these studies: co-dependence, social isolation, mental exhaustion, over-trusting, algorithmic anxiety, expert paradox, and algorithmic management. Because this research is in its very early stages, I have added commentary based on my professional experience that occasionally goes beyond what is found in the studies themselves.

Seven AI Wellbeing Risks and how to manage them

1) AI Co-dependence

Having access to AI appears to increase people’s willingness to take on more challenging tasks but it can come at the cost of decreasing their self-confidence toward solving similar tasks on their own without use of AI (Lee et al, 2026; Ren et al, 2026). A key point in this research is the moderating influence associated with how people use AI to solve problems. People who use AI as a form of expert coach appear to be less susceptible to negative effects on their learning and self-confidence. This means taking time to actively engage with AI solutions to understand what they are doing and not relying on AI for every step in the process. People who use AI simply to tell them the answers and do everything for them may get things done faster, but struggle when they are asked to perform the same tasks without AI.

Wellbeing Recommendation: Train and support employees so they use AI in a manner that enhances learning and self-confidence. This means giving them time to engage more deeply with AI solutions as opposed to just using solutions to solve problems as fast as possible. And designing AI solutions to proactively share information and suggest follow on actions to encourage learning. When it is important for employees to truly understand the underlying methods or knowledge used to perform certain tasks, periodically require them to perform those tasks without access to AI tools in a learning environment. For example, when I was a math student we were required to solve certain problems without using calculators because it ensured we understood key concepts underlying the mathematical techniques being taught in the course.

2) AI induced Social Isolation

Asking questions and providing help is an important social mechanism for building relationship among coworkers. AI can weaken social connections among employees by decreasing the need for human-to-human interactions related to collaboration and knowledge sharing (Klimova & Pikhart, 2025). By weakening personal bonds between employees, AI poses risks to employee commitment, support and belonging which play key roles in reducing employee turnover and increasing workforce resilience.

Wellbeing Recommendation. Create cultural norms and events that foster rich human-to-human interaction. This includes physically bringing people together into shared spaces where they actively engage with one another to discuss common work challenges and interest (note, this is not the same as requiring people to commute to an office very day). Also encourage and support employees in creating mentoring and coaching relationships that involve ongoing human-to-human conversations about work challenges and opportunities.

3) Mental Exhaustion

Early research suggests AI can cause excessive rumination and an inability to switch off of work tasks (Bedard et al, 2026). Described as “AI brain fry”, it reflects a sense of cognitive fatigue resulting from a feeling that one can never do enough to keep with up with the pace of change enabled by AI. AI brain fry is similar to mental fatigue issues that have long been associated with software engineering jobs (Sarkar and Parnin, 2017). Mental exhaustion can result in increased mistakes, poorer sleep, and decreased energy levels. As AI enables more people to engage in software development using things such as vibe coding, mental health risks associated with software development appear to be spreading into new types of work.

Wellbeing Recommendation. Expand the use of wellbeing methods associated with software programming to other jobs where employees make extensive use of interactive AI tools to address work tasks. This includes creating policies to protect employees personal time so they do not feel pressured to constantly work, training employees on how to effectively manage work stress, and creating supportive managerial relationships so managers become a resource for support are not just a source of pressure to get more work done faster (Wong et al, 2025)

4) Over-trusting AI

There is evidence that people may trust AI solutions more than their fellow employees when it comes to judging the accuracy and quality of work. This seems more likely to occur people start to anthropomorphize AI solutions, treating them as though they were actual work colleagues (Wiles et al, 2026). This creates tactical risks related to accepting false or misleading guidance from AI solutions. It also poses social risks if people start being treated as less competent and potentially less valuable than the machines they work with. While over-trusting may not be seen as a wellbeing issue, the consequence of over-trusting AI and under-trusting people could impact employee wellbeing in the form of stress caused by social isolation and job insecurity.

Wellbeing Recommendation. Issues of over-trusting AI appear to be more common when companies design and use AI solutions in ways that encourage thinking of them as living creatures. Companies should avoid giving AI solutions human sounding names and human looking avatars, or placing them on org charts so they appear to be an employee. Employees and managers should be reminded that AI solutions can and do make mistakes. And leaders responsible for making talent decisions should be coached so they do not inappropriately compare humans to machines when evaluating job performance.

5) Algorithmic Anxiety

The term “algorithmic anxiety” refers to emotions employees experience when they feel core parts of their work are likely to be automated by technology (Shekhar & Saurombe, 2026). It is associated with a loss of self-identity as being a valuable part of the organization combined with heightened concerns over job security. The result can be decreased organizational commitment, elevated stress levels, and in some cases active resistance toward use of AI solutions.

Wellbeing recommendations. Algorithmic anxiety is rooted in fears that employees’ existing skills will no longer be valuable in a post-AI world. Algorithmic anxiety can be reduced by defining the role employees will play in the organization after AI has been implemented, clarifying how use of AI will change the skills employees need for their jobs, and providing employees with a path to make the transition from pre-AI to post-AI work. This includes communicating how AI is likely to impact the number of jobs in the organization and the expected pay levels associated with different positions.

6) AI Expert Paradox

Somewhat related to algorithmic anxiety, the AI Expert Paradox refers to situations where human experts are responsible for using and monitoring the results of AI solutions that mimic their expertise (Cotton & Scholls-Cotton, 2026). For example, asking an expert contract lawyer to use an AI solution to write legal contracts and then holding them accountable for ensuring the AI generated contracts are well-written. The Expert Paradox creates two issues. One is related to talent management: how can companies build future experts in a world where the tasks used to build the skills of early career experts have been outsourced to machines? The second issue is similar to wellbeing concerns associated with algorithmic anxiety: managing the stress experts can feel when they are asked to use an AI solution that could challenge their credibility and value as experts.

Wellbeing recommendation. Addressing the expert paradox is about showing how expert employees long-term careers will benefit from using AI solutions. Using AI in a way that challenges and builds expert capabilities and not asking experts to use AI systems merely to do things the experts already know how to do. Another issue is figuring out how to equitably resolve issues that may occur when expert employees disagree with AI recommendations and actions.

7) Algorithmic Management

Algorithmic management refers to jobs where employee goals, tasks, and performance are set and evaluated partially or completely by technology solutions (Jarrahi et al., 2023). A common example are drivers for ride-sharing apps who frequently receive trip assignments and performance reviews through an automated system with little to no interaction with human managers or supervisors. AI is expanding use of algorithmic management solutions across a range of jobs including call centers, manufacturing workers, and public safety officers. Algorithmic management poses at least two specific wellbeing risks. First, it can lead to employees being constantly measured which results in heightened sense of anxiety and an inability to take time off to “mentally recharge”. Second, because AI solutions are incapable of actually caring about another human, use of algorithmic management can increase feelings of social isolation.

Wellbeing recommendation. Companies using algorithmic management techniques can protect employee wellbeing by creating breaks when employee behavior is not being actively recorded. These breaks play a key role in letting employees temporarily switch off from working and mentally recharge. It is also important to ensure employees have a way to contact a supportive human when their work issues go beyond what can be effectively addressed by an uncaring algorithm. No person should ever feel like they are working for a machine.

It is not about if we use AI, it is about how we use AI

We are in the early stages of understanding how AI will impact employee wellbeing. The examples here are a starting point that will evolve as we gain more experience insight into AI’s impact on work. It is also important to remember AI can significantly improve employee wellbeing when used in ways that give employees a greater sense of control over their work and careers. Examples include giving employees more control over setting work schedules, building job skills, and monitoring health patterns.

Perhaps the most critical thing we can do at this point of the AI journey is increase leadership awareness of potential wellbeing issues we need to manage. This is not about slowing down the use of AI. It is about increasing use of AI in ways that recognize and positively shape the inevitable impact it will have on employee mental wellbeing.

References

American Psychological Association. (2023). 2023 Work in America™ survey: Artificial intelligence, monitoring technology, and psychological well-being. https://www.apa.org/pubs/reports/work-in-america/2023-work-america-ai-monitoring

Bedard, J., Kropp, M., Hsu, M., Karaman, O. T., Hawes, J., & Rosen Kellerman, G. (2026). When using AI leads to “brain fry”: Certain patterns of AI use drive cognitive fatigue, while others reduce burnout. Harvard Business Review. (Verify final publication details when available.)

Cotton, S., & Scholls-Cotton, L. (2016). The AI expertise paradox. Issues in Science and Technology, 32(4), 55–61. (Verify page numbers.)

Jarrahi, M. H., Möhlmann, M., & Lee, M. K. (2023). Algorithmic management: The role of AI in managing workforces. MIT Sloan Management Review. https://pages.ischool.utexas.edu/hai-files/files/publications/72/2023-algorithmic_management_the_role_of_ai_in_managing_workforces.pdf

Klimova, B., & Pikhart, M. (2025). Exploring the effects of artificial intelligence on student and academic well-being in higher education: A mini-review. Frontiers in Psychology, 16, 1498132.

Lee, E. H., Yin, Y., Jia, N., & Wakslak, C. J. (2026). Relying on AI at work reduces self-efficacy, ownership, and meaning while active collaboration mitigates the effects. Scientific Reports, 16, Article 13583.

Ren, L., Stephens, J. M., & Lee, K. (2026). The impact of AI on learners’ self-efficacy: A meta-analysis. Behavioral Sciences, 16(1), 158. https://doi.org/10.3390/bs16010158

Sarkar, S., & Parnin, C. (2017). Characterizing and predicting mental fatigue during programming tasks. In Proceedings of the 2nd International Workshop on Emotion Awareness in Software Engineering (SEmotion ‘17) (pp. 32–37). Association for Computing Machinery. https://doi.org/10.1109/SEmotion.2017.5

Shekhar, A., & Saurombe, M. D. (2026). Algorithmic anxiety: AI, work, and the evolving psychological contract in digital discourse. Frontiers in Psychology. (Verify volume, article number, and DOI.)

Wiles, E., Hsu, M., Bedard, J., & Kropp, M. (2026). Putting AI on the org chart: Evidence on delegation and oversight. Boston University & Boston Consulting Group Working Paper. https://www.emmawiles.com/storage/ai_employee.pdf

Wong, N., Jackson, V. F., van der Hoek, A., Ahmed, I., Schueller, S. M., & Reddy, M. C. (2023). Mental wellbeing at work: Perspectives of software engineers. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems (Article 713). Association for Computing Machinery. https://doi.org/10.1145/3544548.3581180

_{Cartoon inspired by Gary Larson}

The term AI literacy has become increasingly popular in the talent management world. Companies want to build AI literate workforces and hire candidates who possess AI literacy. But what is AI literacy and how do we develop it?

AI literacy reflects one’s skill using AI powered solutions to achieve meaningful goals. It can be likened to language fluency in the sense it combines factual knowledge with creative skill. Like learning a language, it is also something you can only master through immersion. It would be hard to learn to speak French fluently without living in a French speaking part of the world. Similarly, it is impossible to become AI literate without spending considerable time using AI solutions to solve meaningful problems.

The following are a few tips from my own journey developing AI literacy. Offered in the hope they may provide a source of comfort or inspiration for those of you, who like me, might find this stuff a bit overwhelming and occasionally frustrating.

A little preparation is valuable but a lot of preparation generates diminishing returns. Developing AI literacy is like learning a language in the sense that it is useful to understand underlying structures that define how it works. Knowing something about the grammatical structure of a language makes it easier to grasp how that language is similar or different from other languages you might already know. The same is true for AI in terms of understanding things that govern how AI works related to machine learning, ontologies, and data structures. But the value of learning these things quickly diminishes compared to the value one gains from immersing yourself in the use of AI.

Get used to being confused. I have worked with AI solutions since my days as an undergraduate student learning recursive machine learning algorithms. Despite being around AI methods for decades, I readily admit to not fully understanding how it works at a detailed level. Just as you do not need to know every word in the dictionary to speak a language effectively, you do not need to know everything about AI to use it at an advanced level. This is one of the things that makes AI so useful. Unlike deterministic computer programming languages that required fairly detailed knowledge to use effectively, AI programming allows people to create practical software applications without understanding how they actually work.

AI solutions are designed to help you but you need to ask them for help. The companies building AI solutions generate revenue based on how often their solutions are used. Consequently, they put a lot of thought into designing solutions that support users in getting value from their tools. If you want to do something with an AI solution but aren’t sure how to do it, then ask the solution to coach you through the process. You might even ask the AI solution if it is able to build what you want to create. But keep in mind the solutions are designed to encourage you to keep using them. Their answers tend to be overly optimistic in terms of their capabilities as an AI solution and your skills as an AI user.

Use more than one AI solution. There are substantial differences in the underlying design and data used by the major AI solution providers. If you ask two solutions the same question you will get different answers, and sometimes they can be significantly different. It is valuable to use one solution as a source of second opinions to evaluate what you are building with another solution. You may also find that one solution is more adept at doing what you want to do than another, depending on what it is you want it to do. Each solution I use has its own particular strengths. How many solutions you use will depend in part on how much money you want to spend on tokens and licenses. But I recommend having at least two solutions you use on a regular basis.

Maintain a healthy skepticism. It is bad to anthropomorphize AI solutions for many reasons, yet it is almost impossible not to think of them as being akin to a work colleague, employee or consultant. However, remember AI solutions are just complex pattern recognition algorithms combined with masses of digitalized data. They have no sense of moral obligation about whether what they are doing is helpful or harmful. Even if they act like they care, they cannot actually care. What you do with AI is entirely your responsibility. The more important the task assigned to an AI solution, the more important it is to critically review the outcomes it produces. My own experiences with AI have led me to think of it as an enthusiastic and wildly overconfident consultant. It is rare for an AI solution to say, “you probably do not want to use me for this task because you might not like the outcome”. Unfortunately, you may have to burn a lot of tokens before reaching that conclusion yourself.

You learn the most when it doesn’t work the first time. It is amazing how easy it is to get AI to do things you want. But increasing your AI literacy means learning how to use AI in more powerful and unique ways. Just like anything else in life, if you want to get better at using AI you need to use to in ways that force you out of your comfort zone. Using AI for tasks that require considerable mental effort, time, and probably a lot of trial and error. AI literacy is not about learning what AI can do, it is about learning what you can do with AI. This means learning how to overcome problems that arise when AI does not do what you want the first time you ask it to.

My journey developing AI literacy reminds me of my experiences learning a foreign language. It is about figuring out how to communicate my goals and intentions to something that does not think the same way I do. The more time I spend doing it, the more fluent I become provided I challenge myself to engage in more complex conversations focused on increasingly sophisticated topics. At the same time, I have to keep in mind it may not interpret the meaning of my words the same way I do. Which is why I need to remain cautious when asking it do something that could create serious problems if it is done the wrong way.

I have attended more than 200 HR technology conferences. All were organized by companies that sell, market or provide services tied to HR technology, with one exception. The Acadian Ventures Summit is organized by a company that funds HR technology start-up companies. The conference is not focused on selling HR technology solutions or services, at least not directly. It is focused on “bringing together investors, founders, innovators and executives who are building the new architecture of work”. There were very few HR leaders at the conference who oversee use of HR technology solutions. In this sense, the Acadian Venture conference might be likened to a meeting of film directors and producers discussing what films to create based on their knowledge of movie goer preferences, but without any actual movie goers in attendance. In this analogy, the movie goers are the HR leaders who purchase HR technology.

The Acadian conference is designed to create connections and encourage information exchange between people who build HR technology companies and people who financially invest in HR technology companies. The agenda provides a lot of time for informal networking. It also includes organized activities that promote conversation such as hiking and pickleball. Because the Acadian Ventures conference is structured differently, the insights gained at the conference tend to be different from what one might learn from attending other HR technology conferences. Here are a few from this year’s conference.

Greed is good at grounding conversations in reality.

Technology investors are primarily interested in the commercial potential of technology solutions. They care about HR technology features, but care far more about how much companies will pay to use these features. This creates a pragmatic tone at the Acadian conference emphasizing the tangible, practical value of HR technology. This is different from other conferences where one often hears HR tech vendors talking with HR tech analysts about the wonders of HR tech solutions, with little focus on whether the solutions address problems that matter that much compared to all the other challenges companies face.

HR technology professionals sometimes talk about “falling in love” with the problems their solutions are addressing. However, many of the problems companies need to solve are not the kind of problems one tends to fall in love with. HR organizations spend a lot of time and money addressing administrative tasks and labor issues they rarely discuss outside of their own offices. It is the rare person who wants to have a dinner table conversation about the challenges of outdated benefits policy manuals or the slow pace of manual training needs analysis methods^[i]. But these are the sort of operational problems that prevent HR leaders from addressing more interesting topics related to creating agile, high performing workforces. HR technology companies that solve these problems tangibly increase the time and resources HR leaders have to address more interesting problems. Which is why HR leaders often buy these pragmatic solutions instead of the more “glamorous” solutions demonstrated on the main stages at other HR technology conferences.

The Acadian Conference featured a lot of companies solving problems that are specific, tangible, complex and in the view of most people, boring. For example, one of the most successful HR technology start-ups in the last 10 years primarily focused on one issue: reducing time spent scheduling candidate interviews. Scheduling candidate interviews is not the kind of problem most people want to fall in love with. But solving it enables HR leaders to shift their focus from reducing time and cost of hire to increasing post-hire performance and quality-of-hire, which is the kind of problem one can happily learn to love (it is certainly one of my passions).

Data is the fuel that powers AI innovation (and some HR technology vendors want to control that power).

In the past, HR data was primarily viewed through a lens of reporting and process automation. HR vendors were open to sharing the customer data in their systems because it elevated the value of their solutions. For example, an HR payroll vendor might share job code and employee attribute information with an HR learning vendor because it made their solution more valuable to the customer. If data in the payroll solution are also used to ensure employees receive training required for their jobs, then the payroll solution becomes more important to the company overall. It was not always easy to share data between vendors due to technical and security challenges, but philosophically most vendors approached data sharing with a supportive mindset.

AI is changing this open-minded attitude toward data sharing. AI solutions use data processing algorithms to automate and augment a growing range of work tasks. Data that was once shared to enable process automation and reporting is now being used to build agentic AI applications. These applications are changing the nature of work while also becoming a major source of revenue for HR technology companies who get paid each time the agents are used. As a result, some HR vendors are increasing restrictions on accessing data in their solutions to have more control over what gets built with the data and who gets paid for using it. To illustrate, let’s revisit the previous example of a payroll vendor sharing job code and employee profile information with a learning vendor. But now instead of using the data to coordinate training delivery, the learning vendor is using the data to build an agentic coach that helps employees develop skills to advance their careers. The payroll vendor might think “that learning company is generating revenue by creating AI solutions that depend on data in our system. They should pay us for this access.”

This topic was a source of a lot of conversation at the Acadian Conference because it has profound consequences for the growth of the HR technology market. Taxing and restricting access to customer data sharing is fundamentally about controlling revenue streams and market share. The HR vendors seeking to exert control over customer data in their solutions are unlikely to admit this although everyone knows it to be true. Instead, they offer all kinds of reasons, some valid and some dubious, about why they need to restrict data access more now that they did in the past. The issue I am most concerned about related to this “data power grab” is it could dampen HR technology innovation, particularly for smaller entrepreneurial companies who may not be able to afford the data access fees set by established vendors.

The people who will ultimately decide how this plays out are HR technology customers. It is still their data even if it is stored in a vendor’s solution. To use an analogy, this is a bit like if you put your money in a bank and the bank told you they were going to start charging you a fee when you took out the money based on how you planned to spend it. If you are using your money to buy “bank approved purchases” then the fee will be lower than if you spend it on “competitor offerings”. I don’t know how HR leaders will react to this shift, but I know how I would feel if it was my data. It will be interesting to see how this evolves over time.

Bigger HR Technology companies innovate largely through acquisitions.

All HR technology companies are committed to innovation, but as companies become larger innovation becomes more difficult. Reasons for this include the operating costs to maintain legacy technology solutions, the inherent bureaucracy of large organizations, and concerns about protecting existing revenue streams. As a result, acquisitions are a critical source of innovation for large HR technology companies.

Larger HR technology vendors tend to acquire smaller companies that their customers are already using. Understanding this dynamic is important for HR leaders seeking to advance their organizations by adopting next generation HR technology solutions. Do not expect your legacy vendors to lead the way when it comes to creating the next generation of HR technology. Look to smaller more agile, entrepreneurial companies for the next “killer app”. The best way to get legacy vendors to provide undelivered features long promised on their product roadmaps could be to work with smaller companies who can build the road for them.

This does not mean big legacy HR vendors are standing still. They are building some highly innovative things, but their innovations tend to be more internally focused and are rarely leading edge compared to smaller companies. There are cases where “almost good enough” HR technology is good enough particularly if it comes free as a feature in a larger system you already own. But to use an analogy, if you want a cutting-edge, high fashion outfit that will make you look fantastic at your next major social event, you are unlikely to find it in a big box retail clothing store.

The people make the place.

The fundamental element that defines any conference is the people who participate in it. Acadian Ventures is led by Jason Corsello and Thomas Otter, two people whose careers have been spent guiding development of next generation HR technology solutions. Many of the people at the conference bet their careers and personal finances on the success of HR technology companies. There is a difference between people who became leaders by climbing the corporate ladders of existing organizations and people who became leaders by building entirely new organizations. This difference shows up in how they talk about innovation and the future of work in general. They have an uncanny focus on understanding customer needs and finding creative, cost-effective ways to support them because that is the only way to be successful as an entrepreneur. They also tend to have a great sense of humor because it is a valuable trait for surviving the stress of growing a company from nothing. If you have a passion for using technology to improve the world of work, it is hard to imagine a much better way to spend a few days then hanging out with the kind of people one finds at an Acadian Ventures conference.

^[i] Acadian Ventures has invested in successful startups that built highly effective solutions addressing these problems and many others like them.

Partial workforce data is often worse than no data because it gives leaders confidence to make bad decisions.

Workforce analytics involves collecting and interpreting data to guide workforce decisions. At some point, people started calling it “workforce intelligence” instead of workforce analytics. One could argue that workforce intelligence is different from workforce analytics based on the kind of data being analyzed or how that data is used, but I suspect the primary reason people use the word intelligence is to make analysis of workforce data sound more strategic and less tactical. The problem is using the word intelligence to describe workforce analytics is misleading. It is also potentially harmful because it downplays the importance of considering business context when making decisions based on workforce data.

You can’t unsee what you just saw

There are many definitions of intelligence, but the most practical is “using the resources one has to effectively solve the problems one faces”.^[i] Having access to data does not make a person intelligent. It makes them informed. The risk of giving leaders workforce data is it can make them think they have enough knowledge to act intelligently without fully grasping the situation behind the data. For example, when business leaders see data that suggests the ROI of employees does not justify the cost of employing them, bad things may happen even if the data is misleading. Consider the following true story. A company had grown through acquisition but was not achieving its revenue goals. Consequently the ratio of workforce costs to profit was below financial targets. Business leaders felt pressure to streamline the workforce, so they asked the workforce “intelligence” team for data showing employees’ job titles, functions, and salaries. Based on this data, the leaders decided to reduce costs by laying off people in roles that appeared to be non-critical to revenue generation. After the layoffs were communicated, the sales team informed these leaders they had fired people with specialized knowledge critical to winning deals. In an effort to increase profitability the company got rid of talent that was central to generating profit. The decision to get rid of these people made sense based on data the leaders had about job titles and salary costs. But what the leaders did not have was contextual knowledge about how these people impacted revenue growth.

The leaders in this story made stupid decisions but they were not stupid people. They were smart people who made confident decisions based on interpreting accurate workforce data in the wrong way. What they lacked was knowledge of information that lay beyond the numbers.

Intelligent and unintelligent applications of workforce data

Intelligent use of workforce data requires a lot more than data analytics. It comes from considering data in the right context. To illustrate this concept, consider the following stories going from unintelligent to intelligent use of workforce data.

Using workforce data without any business context. A management consulting firm convinced the CEO of a global company they could reduce bureaucracy by requiring every manager to have a minimum of five direct reports. Managers with smaller teams had to add more direct reports or have their team merge with other teams and lose their managerial responsibilities. Leaders of different divisions in the organization were given workforce data reports highlighting teams with fewer than five members with the directive to merge, expand or eliminate them to comply with the corporation’s anti-bureaucracy initiative. This created problems for division leaders who had intentionally given high-potential employees smaller teams so they could develop managerial skills before assuming roles with larger spans of control. The “one size fits all” requirement that every manager have at least five direct reports ignored the context behind the creation of these smaller teams, lowered the morale of the employees, hindered development of future leaders, and increased turnover of high-potential talent.

Using workforce data without enough business context. Hiring freezes are a method used to control finances because they swiftly halt growth in headcount and salary costs. What leaders implementing hiring freezes do not see is the financial losses they create. This starts with wasting time and resources spent recruiting candidates only to tell them they cannot be hired. When a company implements a hiring freeze it often loses top talent to other companies including their competitors. Hiring freezes also ignore the fact that the reason companies pay employees is to do things that will increase profits and decrease operating costs. The problem with hiring freezes is not that they never make sense. The problem is they are often implemented based solely on workforce data showing costs saved by eliminating new hires, without considering the financial gains new hires will generate. If the logic of using hiring freezes to improve financial performance was totally sound, then companies could become infinitely profitable by getting rid of their entire workforce.

Using workforce data with the right amount of business context. A major challenge in low-margin industries is determining the right number of people to employ to maximize profit and growth without generating excessive workforce costs. A retail organization combined workforce data with contextual store performance data to determine the optimal number of store managers to hire in a region. Historically the company strove to keep costs low by only hiring a new store manager when an existing store manager left. The analysis showed that when a store manager left the engagement of employees in the store suffered, turnover increased, and store sales declined. Pressure was also placed on other store managers to cover the store until a replacement manager was hired, and consequently the performance of adjacent stores suffered as well. The analysis revealed the cost of waiting to hire store managers until after an existing manager left was greater than the cost of employing an extra “floating” store manager in the region who could immediately step in and run a store when a manager quit. This story illustrates the benefits of going beyond workforce data like staffing and salary to incorporate contextual business operations data that illustrates how people impact business performance.

Using workforce data with local contextual business knowledge. Sophisticated workforce analytic techniques can identify patterns in data that people could never uncover on their own. On the other hand, people can make inferences based on contextual knowledge that computer algorithms will never consider. This point was made in an analysis of turnover data for a chain of hotels. The turnover in one hotel suddenly increased, leading the workforce analytics team to apply a range of sophisticated methods to figure out if something had changed in terms of how employees were being hired and managed. The data failed to provide any clear insights, so one of the team members asked to interview the hotel staff. The hotel manager commented that the local bus company had changed its routes and a bus that used to stop in front of the hotel now stopped a quarter mile down the street in front of another hotel. No amount of advanced data analysis would have led to this simple insight that was obvious to people with contextual knowledge gained from working in the hotel.

Intelligence includes knowing what you don’t know.

Giving workforce data to leaders who may not fully understand the business context underlying that data can give them confidence to make stupid decisions. This is why it is bad to present workforce data as workforce intelligence. A far more accurate title would be “a limited description of workforce characteristics lacking potentially important contextual business knowledge”. I realize adopting such a long and ponderous title is never going to happen, but it is important to keep in mind when analyzing workforce data. Workforce intelligence involves a lot more than workforce data. It requires knowing the context behind the data, and much of that information is not found on spreadsheets and data lakes no matter how big or deep they may be.

^[i] Hunt, EB (2010). Human Intelligence. Cambridge Press.

This article discusses why linking healthcare coverage to employment status damages the economic potential and adaptability of the US workforce. It is written in the hope that maybe, just maybe the US government will begin to seriously address an issue that is sapping productivity from the US economy while creating massive financial and mental stress for US citizens. Discussing who should provide access to healthcare for US citizens is not the focus of this article. The focus is purely on the economic problems that arise when a country’s laws and structures link a person’s employment status to affordable healthcare.

A short history of employer sponsored healthcare

“Saint Peter don’t you call me because I can’t go, I owe my soul to the company store”
from the song Sixteen Tons by Merle Travis

From around 1880 through 1940 many US employees in the mining, manufacturing, textile and lumber industries lived in towns owned and maintained by their employer. These “company towns” were ostensibly created as a benefit to provide employees and their families with safe, affordable living conditions. Some company towns did appear to be run with the welfare of employees in mind. Yet many were used to control and exploit workers. Employees had to accept rents and services offered by the company regardless of cost. If employees quit or were fired they lost both their income and their home including access to services like medical care. Company towns were largely eliminated through the worker’s rights movements of the early 20^th century. But one practice remains that is reminiscent of company towns: employer provided healthcare and health insurance plans.

The United States is one of the only developed countries in the world whose regulations and social expectations assume most people will get access to healthcare through employer sponsored plans. The existing US healthcare system was not designed in any wholistic sense. It is a complex result of competing political, social and economic interests going back to the 1800s. The employer sponsored healthcare plans we have today can largely be traced to an unintended consequence of salary restrictions put in place during World War II. Since employers could not raise salaries, companies and unions developed employee healthcare benefit plans to attract and reward talent. The current practice of linking employment status to healthcare benefits was an unforeseen result of government regulation. It was not a result of free market capitalism.

An anxious workforce is an unhealthy workforce

Access to affordable, effective healthcare has a major impact on people’s overall happiness and quality of life. People in other countries might find it odd to base career decisions based on healthcare plans, but this is a major factor in the job choices of US employees. It is also common for employees to cling to jobs that are underutilizing their talents for fear of losing access to affordable healthcare. Conversely, employers often avoid increasing the size of their workforce because of the high costs associated with providing mandatory healthcare benefits.

It is in the best interest of companies for employees to have access to quality healthcare. A healthy workforce is a productive workforce. But requiring companies to provide healthcare coverage to employees is like requiring companies to give employees family housing. It forces companies to expend resources managing a complex and sensitive topic that is unrelated to the core focus of their businesses^[i]. It decreases labor mobility by hindering transfer of talent within the economy. And it increases the negative impact of unemployment because families lose both their income and access to affordable healthcare at the same time. Healthcare costs are one of the greatest concerns of employees faced with mounting job insecurity and are one of the top causes of bankruptcy in US households. People can control much of their spending, but healthcare is a cost that people have little ability to control.

A healthy workforce is a fluid workforce

Fluid movement in labor markets is critical to building adaptable societies, particularly during times of widespread job disruption caused by technological or economic change. Yet just when the US is entering a time of massive workforce disruption caused by AI, the US labor force is freezing up. Rather than seeking new opportunities, employees are clinging to jobs out of concern for their financial wellbeing. And companies are holding off on hiring in part because of concerns linked to the rising costs of employer provided healthcare.

Loss of healthcare is one of the biggest threats to the physical and financial wellbeing of US employees facing job loss. Providing people with a sense of economic security during job transitions leads to finding better quality work. It also decreases the negative effects that fear of unemployment has on employee performance. This is not just good for employees it is also good for companies. When employees feel supported in making job transitions, they are more likely to consider career changes. This enables companies to hire employees who might otherwise remain in a less optimal jobs for fear of what they could lose if they left their existing employer. It also increases confidence for employees to take risks to build their careers knowing that losing one’s jobs while not desirable, is unlikely to devastate the lives of their family.

If the United States wants a flexible, adaptable labor market capable of thriving in a world of accelerating technological and economic change, then we need to break links tying affordable healthcare to employment status. Employees need access to affordable healthcare, but that should not come through their employer. It should come through a mixture of private and public healthcare services delivered directly to US citizens that is unrelated to where people work^[ii]. When it comes to healthcare, the only thing companies should be responsible for is creating a safe work environment that does not harm employee health and wellbeing.

^[i] With the possible exception of some organizations in the healthcare industry.

^[ii] What this should look like is another topic entirely. One challenge is many insurance and healthcare companies benefit financially from employer sponsored healthcare. In fact, the complexity of employer sponsored healthcare plans is one reason US healthcare costs far more in comparison to other countries even thought it does not generate better health outcomes. Companies that make money from the existing US healthcare system are likely to fight tooth and nail to preserve the status quo.

What is rewarding about being an advisor.

The companies I agree to advise are building solutions solving significant workforce challenges through innovative use of technology. They are also addressing HR topics where I have considerable knowledge since to be an effective advisor one needs adequate expertise to provide useful advice. The most rewarding advisory experiences are marked by the following characteristics:

• Strategic Input. Company leaders actively solicit my opinion on product design and solution marketing. We typically meet by phone at least once a month. They also fund travel so I can meet in person with their leadership teams at least once a year.

• Customer Engagement. They facilitate opportunities to meet directly with their customers to better understand the value of their technology and explore ways to improve its effectiveness. When appropriate, I am brought into sales conversations where my expertise helps build customer confidence in the credibility of their solutions.

• Collaborative Marketing. We engage in joint marketing activities such as social media posts, webinars, conference presentations, and in one case, co-authoring a book. What makes this work is the in-depth knowledge I have gained from spending time with their team to fully understand their solutions, future vision, and customer success stories.

What is frustrating about being an advisor.

I wish I could say all of my advisory experiences have been fulfilling. Sadly, in a few cases I entered into advisory agreements with companies at the encouragement of their investors or partners only to regret the time I put into the relationship. These experiences shared the following characteristics:

• Transactional Mindset. The company leaders did not want advisors. They wanted commission-based salespeople to sell their solutions to friends and colleagues. Several companies offered me a percentage of any sales revenue made by making introductions to people I know. For a variety of reasons, I declined such offers. There is nothing wrong with hiring salespeople, but it is disingenuous to call them “advisors” or “board members” in an effort to disguise their financial motives.

• Leadership Arrogance. This seems to be most common in companies that had significant early success, particularly when it is the founders’ first start-up. I was surprised how little interest leaders in some companies showed toward my knowledge of the HR environment and HR technology market. A good advisor does not simply shower leaders with praise for what they have done. They challenge leaders’ thinking to help them realize the future they could have. That includes telling them how they might screw it up.

• Fickle Commitment. If you read the fine print in many advisor contracts, the company retains the right to terminate the contract with no advance notice or reason. I signed a few contracts like this in good faith that the leaders were committed to a lasting relationship, only to have the contract abruptly terminated after I promoted them in the market but before I received any stock vesting or other compensation for sharing my time and reputation. It is possible that this behavior was driven by the private equity companies investing in the organization and not the founders themselves. Regardless, it makes one feel like a disposable, unappreciated rung on someone else’s ladder to success.

Advice on becoming an advisor.

On balance, I have gained far more than I have lost from the experience of advising HR tech companies. The word “experience” is intentional because the chance of making significant money through advising is relatively low. Most companies seeking advisors are start-ups and do not have a lot of cash they can offer up front. As an alternative, they offer stock options but the likelihood of these becoming worth much is low. An experienced venture capitalist told me less than 5% of HR technology start-ups achieve significant success from a financial standpoint. The majority fail completely or get stuck with relatively flat growth indefinitely (the advisor called these the “walking dead”).

I believe the primary reasons for being an advisor should be because you like the company, believe in the value of its solutions, and enjoy the work they want you to do. With that in mind, I have become much more prescriptive in terms of the nature of the advisor relationship I want with companies. If a start-up has no real interest in my advice and primarily wants access to my contact list and market reputation to sell their solutions they can look elsewhere. But if they are building an innovative technology solution with the potential to positively transform work at scale and they want to work with someone who has done this before, then please send me a note.

“Computer, open a hailing frequency, translate message, and route to Star Fleet Command”. Agentic Artificial Intelligence (AI) refers to technology solutions that use AI functionality to autonomously perform complex tasks that were previously performed by people. My first exposure to Agentic AI happened as a child watching re-reruns of Star Trek and seeing Lieutenant Uhuru give verbal orders to the communications computer. What was once science fiction, has now become commonplace. Agentic AI is widely used for tasks ranging from creating web pages to booking flight reservations and has the potential to create massive transformations to work itself. Its predicted impact ranges from utopian visions where people no longer toil under the drudgery of “soul crushing” work to apocalyptic warnings of mass unemployment as machines automate jobs at scale. Some credible experts even worry Agentic AI systems could achieve a form of self-consciousness and end humanity. My own views fall somewhere in the middle of these predictions. These are my hopes, doubts, and fears toward Agentic AI.

Why Agentic AI is different from previous automation

The things that make Agentic AI possible create inherent risks in its use compared to previous technologies. These include use of:

1. Non-deterministic algorithms that allow it to guess at answers when faced with partial or missing data. This makes it resilient when dealing with the inherently messy nature of the world around us. The challenge is these guesses are not always right and can lead to incorrect actions based on false information (aka, AI hallucinations).

2. Iterative pattern recognition algorithms that enable use of machine learning to improve its performance over time based on feedback on whether its actions or answers are correct. The challenge is how AI measures the concept of correct may not align with how people define correct. For example, many people use Agentic AI solutions to book flights. When booking complex travel or arranging trips for passengers with unique needs, it is common for these solutions to provide less than optimal options which often leads us to request to speak with a human agent. This is because Agentic AI travel agents struggle to understand our needs in the way a human can in terms showing empathy from one person to another.

3. Natural language, image and noise processing algorithms that interpret text, language, imagery, and sound in ways that are often as good or better than human sensory perception and interpretation. This enables AI to appear like it is a person and makes it highly flexible in terms of the information it can interpret. The challenge is there are significant risks to anthropomorphizing machines. The most foremost being the fact AI solutions can not actually care about a living thing. Caring at its core, is a gift of life in the form of time and attention from one living thing to support another living thing. Agentic AI can act like it cares, but it can never truly care in the human sense of the word.

AI agents can be linked up with other AI agents enabling automation of increasingly complex tasks. Self-driving cars, for example, are a complex form of Agentic AI. The potential Agentic AI has for transforming work is extremely exciting. But a career using technology to improve work has also taught me three things about any form of automation: it creates tremendous efficiency and time savings, it inevitably generates false hopes and unmet expectations about its capabilities, and it enables doing bad things quickly at a massive scale. And it often does all three at the same time.

My hopes for Agentic AI.

“We ask people what part of their job sucks and then see if we can automate it so they don’t have to do it anymore.” This statement was made by someone leading a project focused on applying Agentic AI within their organization. When you ask people what they like least about their day-to-day work, they often describe boring repetitive tasks (filling out expense reports), time-consuming mindless tasks that require paying attention to a lot of small details (e.g., reviewing payroll data for accuracy, shift scheduling), physically demanding tasks that are repetitive and/or potentially dangerous (e.g., stocking inventory), or tasks that involve answering simple transactional questions from people over and over again (e.g., explaining HR benefits enrollment processes). These tasks are ideal use cases for Agentic AI. Agentic AI solutions already exist to support all of these examples and many more like them. This is my hope for Agentic AI. That it will be used to free up employees’ time so they can be paid to focus on more fulfilling work that involves developing new capabilities, creating new things, performing intrinsically enjoyable activities, and building productive relationships with customers and colleagues.

My doubts about Agentic AI.

At their core, Agentic AI solutions are digital data processing systems. They collect data, analyze and process it, and then spit out new data in the form of information for people or as digital inputs to guide actions of other technology systems. The capabilities of Agentic AI are limited to the digital data they can access. This data falls into three categories.

• Public data such as that found on the internet in unprotected or easily accessible sources. This is the main source of data used to train the commercial large language models (LLMs) that fuel development of most Agentic AI solutions. This data can be likened to a farmers market in the sense that its often disorganized but everything is discoverable so people can access it.

• Business data kept behind company firewalls to protect commercially valuable or legally sensitive information. This data can be likened to the Pentagon where everything is held in locked file cabinets behind locked doors in a highly secure location. This data has a lot of potential value for Agentic AI applications but it is hard to access.

• Expert data found in people’s heads. This data is the most valuable for companies because it is the most commercially differentiating. It contains new ideas not yet in in the market, expertise gained through experience that other companies do not possess, and detailed knowledge about customers and partners gained through personal relationships. This data can only be used by agentic AI if it is converted from organic neurons in a person’s brain to digital bytes on a computer. Capturing expertise is a notoriously difficult task. And experts may not be eager to download their brains to a computer if it means they may be replaced by an Agentic AI solution^[i].

We are in a period of rapid development of Agentic AI solutions that perform tasks using available public data or by accessing data companies are willing to share with the public for a fee. This means we will be spending a lot more time interacting with customer service AI agents, buying products from Agentic AI interfaces, and relying on Agentic AI computer programmers to build software applications. Development of Agentic AI solutions is moving much more slowly for tasks that require access to business data that companies do not want to share widely. And development often slows to a crawl when Agentic AI is used to automate tasks that require accessing expert data that has not yet been digitalized.

Although Agentic AI has been used to justify workforce reductions affecting thousands of professional jobs^[ii], I suspect we will hit a wall when it comes to using Agentic AI to automate entire categories of professional work. Over the last few years companies have built impressive Agentic AI tools to perform complex professional tasks such as reviewing legal documents, writing computer programs, answering complex HR questions, and providing training. But these agents do not replace everything employees do in their jobs. For example, computer programmers typically spend less than half their time at work writing computer programs. The rest of the time is spent working with other programmers and product designers to ensure the right programs are being written in the right way, including deciding whether a program should be written at all. Automating “the rest of the time” is going to be incredibly difficult.

Companies have been using technology to automate work for decades. As result, a big part of most professional jobs involves dealing with things that are not repetitive and predictable. If you ask someone to describe their typical day at work, they often say “there is no such thing as a typical day”. Every day brings new challenges and unforeseen events. The most comprehensive job manuals contain a small fraction of the things professional employees actually do. Some interesting AI enabled solutions are being created to analyze work tasks and capture expert knowledge. But we are a long way from having the data needed to create Agentic AI solutions that could totally replace the need for humans in most professional jobs. When I expressed these doubts to another HR technology expert, he said “but look at all the examples we already have showing how Agentic AI can automate work for specific, narrowly defined use cases”. To which I replied, “how much of your day-to-day work is specific and narrowly defined”?

My fears about Agentic AI.

“In the future getting service from an actual person will be a luxury reserved for the very wealthy”. This prediction by my brother should strike fear in the hearts of those of us who are not among the very wealthy. Sincerely caring about people lies at the heart of good customer service and employee experiences. As mentioned earlier, AI can act like it cares but is incapable of truly caring in a human sense. Knowing another human cares about your needs and wellbeing has a profound impact on both our happiness and our mental health. The same is true when it comes to interacting with people vs technology. Humans are social creatures and we suffer when technology prevents us from having meaningful social interactions with other humans.

The cost savings generated by replacing human service agents with AI service agents are too tempting for companies to ignore. Most of us have experienced the frustration of being stuck in customer service phone trees where we are required to “push 3 for this” and “press 7 for that” when what we want is to talk to a person who cares about helping us solve our problem. Agentic AI is going to take this to an entirely new level. Companies are already using agentic AI technology to replace frontline retail, hospitality and dining service staff with automated solutions. Thanks to development of self-driving cars, automated airline check-in, automated restaurant agents, and self-service hotel apps one could travel from a hotel in Los Angeles to a hotel in Phoenix and without ever talking to an actual human. If this trend continues, much of the world will eventually be like a giant automated vending machine where traveling, shopping and dining become transactional experiences largely devoid of human contact. This is not a world I am excited about living in.

AI technology will improve our lives in many ways. But there are many historical examples that demonstrate the harm done when technology is used to centralize profit and power among a small class of individuals with little regard or regulation to protect the wellbeing of society in general. Many technological inventions of the 19^th century industrial revolution led to horrific living and working conditions for thousands of people while generating phenomenal wealth for a few business owners. This lasted until the 20^th century when workers’ rights movements created health, safety and wage protection regulations. Agentic AI has the potential to impact millions of people working in service jobs in a manner similar to what happened to workers in the textile and manufacturing industries in the late 20^th century when improvements in supply chain technology enabled offshoring production jobs at a massive scale. Innovations in technology that enable companies to increase profitability by fundamentally changing how they produce and deliver goods and services can create unemployment levels that devastate communities for generations. Agentic AI poses this risk for wide classes of service jobs. Managing this risk will require reframing the value placed on the uniquely human skills that enable service workers to create caring relationships with others, whether those relationships last for years or simply brighten a person’s day for a moment through a kind smile offered while purchasing a cup of coffee. Our society has historically devalued the skills of service workers going so far as to label many service jobs as “unskilled” even though they are highly demanding occupations. Anyone who has worked in a customer facing job knows how difficult it is to consistently make people feel good about the services you are providing. And all of us have experienced the joys and frustrations that come from being on the receiving end of good or bad customer service interactions.

Two things will help us elevate the value of employing humans to deliver service to other humans and resist the economic temptation of entirely outsourcing service work to AI agents. First, we need to get better at measuring the value of customer service skills and end the tyranny of paying service workers fixed, hourly rates as though they are interchangeable widgets with no unique talents or abilities. Second, we need to stop acting as though Agentic AI solutions are technological replications of human beings. This includes discouraging the tendency to give AI agents human names and endow them with fabricated personalities or mimic human features in an effort make them seem like an actual person. Believing that AI “does what people do” lures us into using AI in unhealthy ways that devalue the importance of authentic human interactions both in service settings and in life in general. The current trend to treat AI as though it were a replacement for people reminds me of ads from my childhood that encouraged consumers to think of margarine as a replacement for butter. Not only does margarine taste worse than butter, we have learned it is often bad for our health. Margarine is not butter, and AI agents are not people. Instead of designing agentic AI solutions to act like people, we should design them to act like what they are: uncaring iterative complex pattern recognition data algorithms. Or more simply put, computers.

^[i] The 1996 science fiction novel The Sparrow described a category of specialists who shadowed subject matter experts to analyze their knowledge in extreme depth so their work could be automated using artificial intelligence. These people were called “vultures” by the subject matter experts who resented having to share information with someone who would ultimately use it to eliminate their job.

^[ii] There is debate about whether these reductions were truly caused by use of Agentic AI, or Agentic AI was simply a convenient excuse to justify reducing workforce costs for reasons that had nothing to do with automation.

TED talks have become one of the most influential sources of behavioral science knowledge in our society.^[i] For example, three TED Talks by Dr. Brenee Brown, Dr. Amy Cuddy, and Simon Sinek referencing psychological research have been collectively watched more than 2 billion times. By comparison, the most popular American Psychological Association peer-review article ever published has just over 30,000 citations and I suspect has been read by fewer than 100,000 people^[ii]. One TED Talk can dramatically increase the influence of a single study, creating awareness across much of the general public around its core findings. TED Talks have also helped turn researchers into minor celebrities leading to high profile speaking sessions at HR and company leadership conferences. As an applied psychologist, I appreciate what TED Talks have done to promote the value of behavioral science to a wider audience. At the same time, the ability of TED Talks to make certain research findings highly popular and catapult a few researchers into celebrity status does not necessarily mean they are always making the world better.

A brief explanation of TED Talks.

The birth of TED Talks traces to a conference in 1984 designed to share information about Technology, Entertainment, and Design (TED). The TED organization is now a global company coordinating thousands of events, presentation, blogs, articles and other media focused on the belief that “powerful ideas, powerfully presented, move us to feel something, to think differently, to take action”. TED Talks follow an extensive set of guidelines including being limited to less than 20 minutes. TED also provides coaching to speakers on how to present. As a result, TED Talks tend to have a highly similar structure to the point it has become the target of parody. The TED organization accepts nominations for speakers and offers suggestions on how to be accepted, but is not transparent on how speakers are selected other than stating “we will evaluate your nomination and will share this information with staff members and third parties selected to review speaker nominations.”

There are currently over 250,000 Ted Talk recordings on scores of topics ranging from poetry to driverless cars. Many of the most popular TED talks address behavioral science topics related to work. When preparing this article, I gave Claude and ChatGPT the prompt, “You are a graduate student in industrial organizational psychology and have been asked to identify the 10 most influential TED talks that provide leadership or career advice based on behavioral science research.” The systems provided overlapping lists containing a total of 12 talks. The talks ranged from a low of 3 million views to a high of over 700 million views. Many of the observations in this article are based on the content of these presentations^[iii].

The value of powerful ideas, powerfully presented

The value of the TED format is it makes behavioral science knowledge easy to consume and widely available. Most behavioral science research studies languish in obscurity on the pages of peer-review journals. TED Talks give a few lucky researchers the chance to broadcast their findings to the world. Being a researcher myself, I hold no ill will toward my fellow scientists who have promoted their research through TED Talks.^[iv] Who wouldn’t want the opportunity to share their work on such a powerful stage? In this sense, TED Talks remind me of television shows like “America’s Got Talent” or “Eurovision” but instead of giving artists the chance to demonstrate their musical skills to the world, TED gives academicians a chance to showcase the value of their research to the public.

The TED Talk format has also proven effective in helping researchers simplify their study findings into easily understandable insights and recommended actions. Presenters frequently explain their research by engaging the audience through the power of storytelling and emotion. The sessions are designed for the general public and can be understood by someone with no formal psychological training and no practical experience working in the field of human resources. TED Talk presenters often spend almost as much time sharing amusing and emotional stories and anecdotes as they do explaining the research. TED Talks are entertaining and easy to listen to, and consequently people listen to them. This makes them markedly different from many presentations that discuss interactions between psychological science, workforce productivity, and employee performance.

The problem with powerful ideas, powerfully presented

The risks posed by TED Talks are related to the same things that make them valuable. To use a phrase from executive coaching, their strengths are their weaknesses. These include:

• Making it the most popular does not mean it is the most impactful. TED Talks are designed to give the audience an “aha” moment by providing some unique or unexpected insight. The most popular ones related to work tend to focus on research findings that may seem counter to popular beliefs about job performance. They also focus on somewhat narrow psychological findings that are widely overlooked in part because they are so specific. Examples from two popular TED Talks include research that identified situations where sharing ideas hurt employee performance or instances where showing vulnerability helped build resilience.
  
  There is value in learning about the psychological findings showcased in TED Talks, but it can also call attention away from other research into the psychology of job performance that is more impactful but perhaps less interesting. For example, if there is one thing companies and people could do to improve workforce performance and career success, it is to put more energy into setting effective goals and clarifying ongoing role expectations through constructive dialogue^[v]. You are unlikely to surprise an audience by telling them employees have greater success when they know what they are supposed to do in their job and receive guidance on how to do it effectively. Nor will it surprise people to learn that companies are bad at communicating job expectations. Research about effective goal setting, while highly valuable for talent management, is unlikely to be popular on the TED stage because the topic sounds boring and familiar. Even though using goals to ensure role clarity has a far greater impact on employee success than knowing when to share ideas or display vulnerability to coworkers.
  
  Having a company build talent management programs focused on much of the research highlighted in TED Talks is a bit like the person who asks their doctor if they should take more vitamin supplements when the real problem is they smoke two packs a day. You aren’t going to make things worse, but you are not changing the things that really matter.

• Statistically significant does not mean practically useful. TED Talk guidelines state that “scientific claims by speakers should be based on data that has survived scrutiny by experts in the field. Speakers should be transparent about the basis for any factual claims, and the scientific evidence for them.” Most behavioral science related TED Talks reference specific studies and occasionally show data charts from the research. Although they rarely share detailed information about the data samples or study methods.
  
  It would not make sense to delve into statistical methods in the context of a TED Talk. There is not enough time and most of the audience would find it boring, confusing or both. Instead of talking about effect sizes and moderators, TED presenters often support the validity of their research with anecdotal stories. The problem with this approach is it can make the impact of certain research findings sound far more powerful than they actually are. One widely discussed example is a highly popular TED Talk that implied making simple changes to body posture has a major impact on people’s physiology and subsequent performance. This claim vastly overstates the effect size observed in the studies. Unless a person has horrific posture, suggesting that adopting a new pose is going to transform someone’s performance is a bit like telling a jogger that one of the best things they can do to run faster is change the brand of their running shorts. It might have some impact, but for most people the difference will be trivial compared to other actions they might take.
  

• What happened there may not replicate here. Many TED Talks take research findings collected in unique situations and apply them to the world in general. For example, a TED Talk watched by over 3 million people argued against the motivational value of money by claiming “here is what science knows...if-then rewards often destroy creativity…The secret to high performance isn’t rewards…but that unseen intrinsic drive to do things for their own sake.” The research studies referenced in this talk involved things such as paying college students to play games instead of having them play them for fun. That is a lot different from paying employees to perform their jobs. It is a stretch to think these findings will generalize to corporate settings. Using financial rewards to incent performance is a complex topic. To quote Dr. Barry Gerhart, “Pay for performance when it works seems capable of producing spectacularly good results. And when it does not work, it can likewise produce spectacularly bad results”. But suggesting that if-then financial reward structures do not play a role in maximizing employee performance levels is not just scientifically wrong. It is practically ridiculous.
  
  It is difficult to assess the risk of over-generalization in TED Talks because they rarely provide enough information about the studies referenced in the presentation to make an evaluation. In many cases, they do not give the names of the studies and instead just say some version of “research shows” and perhaps present a few graphs and charts with little explanation of the methods used to collect the data. The TED format does not provide time to provide a proper explanation of research methods and data, nor should it. But it would be nice if speakers were required to include a list of research citations enabling the audience to look at the source data.

The main risk TED Talks pose to the field of HR are not about the content in the talks themselves. The main risks are the potential of TED Talks leading companies to focus on scientific findings about job performance that are interesting but not particularly important in the grand scheme of managing workforces. Or even worse, encouraging companies to adopt talent practices based on research findings collected in settings that do not generalize to their organization. In my book Commonsense Talent Management, I compare creating healthy workforces to creating healthy bodies. There are more things we ideally should be doing than we have time or resources to do, and the greatest benefits tend to come from doing a few basic things consistently vs constantly chasing new “scientific discoveries”. TED Talks are an entertaining way to learn about specialized research findings but they are also a potential source of distraction for leaders focused on building more effective workforces.

Talk Show Expertise

A few years ago, I was at a conference sitting next to the CHRO of a large global company watching a TED Talk presenter give a keynote presentation. The CHRO commented that this presenters clearly never worked in an HR function. She was not dismissing the value of their knowledge but noted they seem ignorant to the realities and constraints faced by leaders in companies. She likened many HR conference presenters to experts one hears on talk shows such as Dr. PhilandDr. Oz who provide short, quotable advice to people with no real understanding of their health history, family dynamics, or socioeconomic realities.

The problem with “talk show” expertise, which is how I would describe many TED Talks is usually not what is being said. The problem is the talks over-simplify complex problems and highlight research that is interesting but not necessarily highly important in the grand scheme of managing workforces and building careers. TED Talks are designed as much for entertainment as they are for education. It is important to remember this when watching them. Using TED Talks as a source of talent management and career advice is akin to getting medical advice from a doctor speaking on a talk show. You may learn some interesting things but be cautious in how you apply them to your own life.

^[i] The term behavioral science is used broadly and includes psychology, sociology, and behavioral economics.

^[ii] The most popular article based on citations is Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 1173–1182. I confess to not having read it myself.

^[iii] I do not provide the names of these talks as my goal is not to criticize specific TED speakers. The focus is on critiquing the nature of TED Talks in general.

^[iv] I will admit to also feeling a bit of envy and longing to be on the TED stage myself.

^[v] Locke, E. A., & Latham, G. P. (Eds.). (2013). New developments in goal setting and task performance. Routledge/Taylor & Francis Group. https://doi.org/10.4324/9780203082744

Why CFOs are like lousy sailboat captains

An analogy can be made between leading a large workforce and captaining one of those old clipper ships with multiple masts and dozens of different sails. Both require managing a myriad of actions to achieve maximum productivity. Changing the direction of a clipper ship without losing speed requires coordinating actions across a lot of people to alter the ship’s course while still keeping wind in the sails. Similarly, quickly changing the financial performance of an organization without damaging workforce productivity requires changing people’s actions without undermining their performance. When a clipper ship captain gives the order to turn hard to starboard, they do not want everyone on the crew to do the exact same thing. They are telling each crewmember to alter the particular set of sails they manage in a way that leads to the ship changing direction. Officers monitor the crew to ensure members are taking appropriate action to change the ship’s course.

Finance leaders in large companies often behave like incompetent clipper ship captains. Issuing blanket directives to decrease costs that might be likened to the sailing equivalent of telling everyone to drop all the sails, throw the anchor overboard and yank the anchor line to move the ship in a new direction. The ship may change direction quickly, but it is now sitting dead in the water robbed of momentum. Four common examples include blanket hiring freezes, insensitive cost saving initiatives, counterproductive compensation strategies, and misinformed downsizing decisions.

Hiring freezes – encouraging managers to hire fast and fire slow. Hiring freezes control costs by prohibiting managers from bringing new employees into the organization for a set period of time. A hiring manager may be on the verge of making an offer to a candidate only to be told “the headcount is no longer approved”. Hiring freezes undermine one of the oldest bits of management wisdom to “hire slow and fire fast” by taking time to hire high quality talent but swiftly exiting employees who are undermining company performance. To illustrate this point with a story, I knew a hiring manager who spent several months searching and evaluating candidates for a critical position. He finally found the ideal candidate when the finance department issued a hiring freeze to control costs. As a result, the candidate accepted a job at competing organization. His manager told him in the future you need to hire people as fast you can or risk having finance take away your headcount budget. The finance department’s history of hiring freezes created a culture that punished managers who emphasized quality of hire over speed of hire. Furthermore, because headcount budgets where regularly frozen by finance to save money, managers were reluctant to fire under performing employees lest they be unable to backfill the position. In sum, the actions of the finance department encouraged managers to hire fast and fire slow, which is the opposite of good management.

Cost savings initiatives – sacrificing performance for efficiency. Cost saving initiatives designed to decrease operating costs are often designed by people who do not directly work with customers to generate revenue for the company. As a result, they over-emphasize outcomes important to internal support functions and under-emphasize outcomes associated with service and sales. Consider the following story about a health care organization I worked with. Someone in corporate headquarters felt the clinics were spending too much money on contractors to remove trash from patient care rooms. They eliminated this program to save operational costs, shifting the task of garbage disposal to the physician assistants. Treating highly trained healthcare professionals as though they were custodians did not go over well with the frontline staff. The initiative to reduce operational costs damaged employee morale and decreased the time health care providers had with their patients which led to far more costly actions caused by increased staff turnover and decreased patient satisfaction. The financial spreadsheets used to guide cost efficiency decisions rarely have information showing how these decisions will affect the performance of frontline staff who are supporting customers and growing sales.

Sales compensation strategies that discourage selling. The fundamental purpose of sales compensation is to increase sales volumes. Yet many sales compensation strategies discourage employees from selling. For example, I saw a compensation plan that rewarded employees with a one-time bonus of $5000 if they closed a product deal worth at least $30k in annual revenue. If a salesperson sold a deal for $29k they got nothing. As a result, salespeople did not spend time on accounts that had any chance of coming in under $30k. Ignoring lower value opportunities regardless of how easy the sale might have been. This compensation strategy only made sense if the company truly did not want any deals under $30k, which was not the case. Another example is capping sales commission at a certain level. For example, saying salespeople can only receive commissions on up to $1 million annually and after that they get no commission for additional sales. Caps like this are often put in place because financial leaders think it is inappropriate for salespeople to make more than a certain amount of pay. But all this does is discourage salespeople from maximizing revenue for the company. Not to mention demotivating them.

Downsizing – ignoring the value of employees. Companies that downsize tend to perform more poorly compared to organizations in the same industry that avoided workforce reductions^[i]. These lower performance levels persist for years following layoffs and can create a sense of resentment among employees who remain. ^[ii] Nevertheless, there are times when companies have to quickly reduce workforce costs. In such situations, restructuring decisions often focus on reducing headcount costs without adequately considering employee value. This can lead to catastrophically bad decisions. For example, after a market downturn a manufacturing company eliminated a department working on a new line of products that they could no longer afford to develop. What they did not consider was the department was staffed with some of the most experienced engineers in the company who had been moved into this department because of their capabilities. It was only after firing them that the company realized they possessed crucial skills the company needed. This led to hiring the employees back as contractors at much higher rates with much lower levels of commitment. Restructuring decisions should not be based solely on employee job titles, cost centers, salaries, and demographics. They should also consider data about employee skills, experience, relationships, and capabilities. Keeping in mind there is cost to not employing someone even if it is harder to estimate than the cost of keeping them. Otherwise, companies could just fire everyone and be infinitely profitable.

Finance plays a critical role in ensuring the financial viability of an organization. Money is the lifeblood of companies, and hard financial decisions are necessary to ensure a company’s success and survival. But there are countless examples that suggest many finance leaders do not fully understand or appreciate the damage that misguided financial decisions have on workforce productivity. These financial decisions may provide a good quarterly story to tell Wall Street but they tell a much different story to employees within the company.

• The company’s leaders don’t trust you. We don’t believe you are competent enough to know whether the long-term revenue generated by a decision is worth the short-term cost.

• The company’s leaders cannot control the workforce. We have to issue blanket corporate financial policies because leaders lack discipline to ensure smart financial decisions within their departments.

• The company is performing so poorly we need to sacrifice long-term gain for short-term survival. We have to take these extreme measures because our strategy and current practices are failing.

• The needs of managers and employees are less important than the needs of shareholders. To keep our stock from dropping this quarter we need to hit our short-term financial targets, which is more important than supporting the long-term career goals of our employees and managers.

• Company leaders do not care about employee motivation. We base our reward structures around financial results that will help executives achieve their goals regardless of whether they make sense from an employee perspective.

One way to avoid these problems is to minimize use of corporate mandated company-wide financial policies. Instead communicate cost saving and compensation requirements to specific departments and let people closer to the frontlines determine how to achieve the financial targets. In those cases when it is necessary or most effective to issue corporate wide financial policies, spend time understanding how this will impact workforce productivity and seek ways to moderate potential negative outcomes.

For many years HR departments have been justly criticized for not doing enough to understand the financial side of the business. The same can be said for many Finance organizations failing to understand what drives workforce productivity. Decisions that make sense based on financial spreadsheets can seriously undermine workforce performance when they are rolled out to employees whose lives are impacted by them. The partnership between HR and Finance should extend both ways with both groups seeking to learn more about each other. Positive employee experiences do not come from working for financially struggling organizations. The main reason companies employ people is to hit financial targets. But companies will not hit their financial targets if they do not effectively manage the people they employ.

^[i] Cascio, W.F., Young, C.E., & Morris, J.R. (1999). Financial consequences of employment-change decisions in major US corporations. Academy of Management Journal, 40, 1175-1189.

^[ii] Huang, G.-h., Wellman, N., Ashford, S. J., Lee, C., & Wang, L. (2017). Deviance and exit: The organizational costs of job insecurity and moral disengagement. Journal of Applied Psychology, 102(1), 26–42. https://doi.org/10.1037/apl0000158

Worst practices. There are no best practices for returning to offices. What might work in one company could fail in another. However, there are two things we can be certain will not work:

• Never going to the office at all. Research on virtual teams found that having in-person meetings positively impacts collaboration, commitment, and morale. Employees do not need to be together all the time to be effective. But there is value in periodically bringing people together to meet face-to-face. Furthermore, people miss getting together. Remote workers tend to have higher commitment and wellbeing than in-office employees, but the social isolation of remote work can also take a toll on the engagement and mental wellbeing of employees. In office work is also valued by early career employees seeking to build work relationships through in-person interactions with colleagues. A100% remote work model is neither optimal or desirable, nor is it likely to be sustainable over time.

• Requiring employees to commute every day. Most employees do not want to return to the “commute to the office just to be in the office” cultures that existed prior to the pandemic. Organizations that require employees to swipe an access key card every day to prove they are in the office are likely to see high performing employees quit and high potential candidates turn down offers. Employees want the flexibility to decide when to go into the office based on the nature of their job, work location, living arrangements, and personal preferences. Not to mention the burden commuting places on employees in terms of transportation costs and lost time. By one estimate, employees who have to commute daily spend 223 hours a year just getting to and from work. That translates to about six unpaid 40-hour workweeks. Daily commuting is particularly challenging for people with disabilities and employees who have primary childcare responsibilities (who tend to be women). Leaders that advocate returning to “how it was when everyone went to the office every day” should be challenged on whether where people sit is more important than what people do. What does it say about a company when it restricts employee freedom just so leaders can cling to work habits of the past, particularly when many of these leaders do not go into the office themselves?

The four uses of office space. The are four reasons why companies want employees to use offices.

1. As a resource that provides an effective workspace with access to materials and tools needed to perform job tasks.

2. As a reward that allows employees to work in an aesthetically pleasing environment or giving them access to recreation and entertainment resources (e.g., cafeterias, gyms, ping pong tables).

3. For collaboration by providing a place for people to interact and build working relationships.

4. As a means of control to manage employee performance. Some leaders assume if employees are physically present in an office then they are more productive. This assumption is wrong.

A mindful return to offices requires examining the role the office plays relative to the first three of these reasons. Does an employee want or need a dedicated workspace outside of their home? Does their job require access to specialized equipment? Do they put value on working in a “nice office” or appreciate having access to office amenities such as food or exercise equipment? And most complex of all, how important is it for them to be physically next to the other people working in the office?

The fourth reason of using the office as a means for control is a different problem. Companies that monitor how long employees are in the office and evaluate them based on where they are sitting should rethink their leadership models, not their return-to-office policies. Managing people based on how much time they spend in the office is an ineffective, highly controlling and potentially exploitive form of leadership. The amount of time employees spend in an office says little about their commitment or performance. Arriving early or staying late at an office does not mean someone is more productive. It may mean they take more time to get work done compared to their peers. Similarly, leaving early does not mean someone lacks commitment, but it may mean they have to pick up their children from childcare. There are very few jobs where it makes sense to evaluate employee performance based on where someone sits and how long they sit there.

The goal is to create better interactions, not to spend more time in the office. Some leaders calling for a return to offices claim people are not as creative, collaborative or productive when they work remotely. This is not true based on data collected since the mass move to remote work that occurred in 2020 as well as research predating the pandemic.Remote workers are often happier and more productive than office workers in general. On the other hand, meeting face-to-face does affect the quality and nature of communication. The question is not whether employees should meet in person. They should. The question is whether this requires commuting to an office every week.

Many people believe working in an office creates more frequent and effective live interactions between employees. Live interactions are defined as two or more people interacting at the same time. They include things such as real time texting, phone conversations, video chats, and in-person conversations. These can be contrasted to asynchronous interactions such as e-mail. ^[1]. Live interactions are critical to supporting employee collaboration and building employee commitment. And in-person live interactions tend to create the most immersive social experiences. As the world becomes more remote, in-person live interactions can become less frequent which further increases their value. Strategic use of office space is a powerful way to create live interactions. But simply forcing people into the same office building is no guarantee that they will have productive live interactions.

A mindful return to office strategy does not focus on returning to the office. It focuses on maximizing the occurrence of effective live interactions. This includes making employees aware of the importance of meeting in-person periodically. Employees who refuse to come into the office for important meetings are not just decreasing workforce performance, they are also hurting their own career development. This should be communicated to them in advance. Employees should be given the choice of whether to come into the office but should be aware of the consequences of their choice both good and bad.

Sitting together is different from working together. Requiring people to go into the same office every day does not mean they will talk to each other more frequently than people who work remotely. People are quite capable of ignoring the people sitting around them. Remote workers are actually more aware of the importance of scheduling conversation with their coworkers than people who work in the office. In addition, while in-person meetings have the potential to be highly effective, they can also be a waste of time. Everyone has spent time in a meeting wishing they could leave and do actual work. Leaders who overemphasize being in the office can also create dysfunctional organizational norms that confuse being physically present with being productive. Putting people in offices can negatively impact the power structure of a company. People whose desks are near the executive offices are more likely to randomly interact with senior leaders. This gives people influence based on where they sit instead of being based on what they know or what their formal role is in the company. In sum, there is a dark side of in-office cultures that can be attenuated by effective use of remote work models.

Interactions are more important than offices. It is important to bring people together for in-person meetings to collaborate on shared projects, work though potentially difficult conversations, and build the kind of human connection that only occurs when we directly look someone in the eye. But one need not commute to an office to have these sorts of live interactions. A mindful use of office space focuses on balancing the strengths of remote work with the benefits of in-person live interactions. Communicating and holding meetings in way that makes people feel included whether they are in the office or offsite. Empowering employees with flexibility to decide when they need to commute to the office and when they can be more effective working remotely. And minimizing the role that a person’s geographic location plays in determining their qualification and ability to perform a job. When it comes to doing the kind of skilled work that allows for remote work, an employee’s ability to use the knowledge in their head to solve the company’s problems should be more important than whether they can physically transport their head into the company’s office building every day.

Being together in an office can enable highly valuable live interactions. But one need not commute daily to have the sorts of live interactions we associate with offices. Nor do live interactions need to be in-person to be effective. Face-to-face meetings are not always better than virtual meetings, and it is often far easier to communicate remotely than it is to communicate in person. This is why employees frequently text co-workers even when they are sitting next to each other.The best approaches to returning to the office balance the benefits of remote work with the value created by in-person interactions. Ensuring employees return to the office in a manner that is mindful instead of mindless

^[1] There is another dimension reflecting the number of participants. For example, participating in a one-on-one interaction is more immersive than participating in a 30-person meeting.