The performance review is dead! Long live the performance review

Wharton professor Peter Capelli once raised the question: Is employment in an organization a contract or a relationship?

In a knowledge worker context, it is a bit of both. The employment contract specifies job responsibilities, competencies, and skills required to perform the job. However, it is also a relationship because feedback and performance are often dependent upon peer and manager interactions and perceptions about performance. In the absence of hard data and well-defined metrics, measuring actual output for an employee makes the performance review process fraught with pitfalls.

A Harvard Business Review article in 2016 summed up the performance review process and outcomes as one that both managers and employees hate. A McKinsey survey pointed out that employees felt their managers did not grade their performance correctly, while managers felt their employees were performing to their job description and rarely exceeding expectations.

The process continues to be cumbersome and highly subjective but cannot be dispensed with completely, as it is necessary to calibrate alignment to company objectives and reward employee performance.

The problem with the current state of performance reviews

Getting performance reviews right is more an art than a science. Stakes are high as review outputs are inputs into other processes such as succession management and compensation planning, for example.

There is a lot of variation in how people perform and while they may be highly productive in some aspects of their job as measured by their hard skills, there is not an objective way of measuring their soft skills.

This leads to difficult year-end conversations, especially in environments where managers are forced to stack rank and calibrate employees against supposedly similar peer groups. In such companies, arbitrary rules on the percentage of employees that must fall into various categories often result in a more disengaged workforce.

Additionally, there are implicit and explicit costs for all parties. A Gartner study in 2019 showed that employees in large organizations spent ~40 hours on performance review-related activities while Managers spent about 10% of their time on similar activities. Gallup noted in 2019 that it cost organizations of 10,000 or more employees anywhere between $2.4million to $35million a year in lost working hours to take part in performance evaluations.

A 2022 survey by Willis Towers Watson of 837 organizations worldwide, of which 150 were in North America, found that only 16% of North American respondents felt they were effective in managing and paying for performance though driving organizational performance was a key objective for performance management for 93% of the respondents. Only 44% said their program was meeting the objective. Similarly, nearly 72% said supporting the career development of their employees was a primary objective, but 69% said their performance management program was not meeting that objective.

These data points underscore the dissatisfaction with the current state of the performance review process and a call to come up with creative new ways to measure and manage performance. However, for large companies, scrapping the entire performance review process is also problematic, as managers need some form of an evaluation process to make sure employees are rewarded fairly and appropriately for their contributions.

Correlation between performance management and organizational trust, Betterworks

What's a company to do?

Not surprisingly, over the past few years, talent-forward companies have been laying the groundwork for a more equitable and fair process and building trust. The updated process is structured around more frequent touchpoints between managers and employees throughout the evaluation period, adding more technology to support the process, providing more transparency into the process to build trust, as well as better training and guidance to both parties on giving and receiving constructive feedback. The chart above from from a Betterworks survey shows the correlation between performance management outcomes and levels of trust across the organization.

Ultimately a more comprehensive approach should have the following elements:

1. A rolling process (rather than an annual process) that consists of frequent check-in conversations that are aimed at providing more feedback and generally making the annual conversation more of an ongoing discussion. The idea is that this higher degree of engagement allows for clarity in expectations, improved goal setting and performance, and ultimately translates to more productivity that benefits the company as a whole as measured not only by improved employee engagement scores but also by the company’s bottom line.

2. Mechanisms to increase peer-to-peer feedback, especially in the context of alignment to company values and continuous two-way feedback between the manager and the managed. Caveat emptor, though, as an unintended consequence of these elements could be to unleash a barrage of feedback, as the Wall Street Journal pointed out in a recent article, that could very easily overwhelm the recipient and could increase the administrative burden as well.

3. Greater transparency into the evaluation framework and process, which should be tied to company objectives and outcomes. This is especially important in the context of new performance evaluation and management tools and technology coming onto the market that may leverage generative AI.

4. Ongoing training for managers and employees to set appropriate expectations at the beginning of the year and revisit those in frequent conversations, and guidance to all parties on the best ways to provide constructive and actionable feedback.

5. No productivity measuring tools, like measuring the number of lines of code a software engineer produces, which could lead to unintended consequences such as code bloat.

Future thoughts

Looking forward to what is next for the performance review process, as companies begin to evaluate organizational workforce models from more of a skills centric perspective, it seems quite certain that future performance discussions will likely include a large component on skills and skills based career progression.

Another big trend is of course the use of Generative AI tools and models in various programs and processes. It is tempting for leaders to consider using Generative AI tools for performance management from a process efficiency and cost or time saving standpoint. However, while Generative AI can be used in some parts of the process such as setting individual development goals or summarizing feedback, it would be wise to be very thoughtful and careful in determining which tools are used, where in the process they are deployed, and making sure there is no erosion of trust in the organization as a result of using Generative AI.

Daya Nadamuni is an experienced workforce strategy leader with a proven track record of success at Adobe and Walmart.