How can using data help improve your performance reviews?

As a young employee, I never understood the purpose of annual performance reviews. To me, they were pro forma exercises that had little to do with day-to-day career development. But at least they never made me cry.

That’s not the case for the 34 percent of millennials in one recent survey who said a performance review had driven them to tears.

Annual reviews can serve important functions in workforce management, but too many companies rely on ad hoc processes riddled with inconsistencies, influenced by a host of irrelevant factors, and plagued by hidden biases. But the good news is that new tools, and access to new types of data, can transform performance reviews into occasions for real insight. As the leader of a company of more than 500 people, and the founder of an HR platform with insights into 25 million employees, I’ve seen the transformational power of using data in the review process.

As companies resume regular performance reviews post-pandemic, organizations have a unique opportunity to develop a more fair and productive methodology. Considering that the review process costs as much as $35 million a year in lost working hours for an organization of 10,000 employees, it’s worth taking the time and effort to get it right.

Here’s what leaders should consider when preparing for performance reviews in 2023.

The first step to creating a fair and useful performance review process is to gain clarity of purpose. First and foremost, they’re tools for assessment, not for individual growth. This is often misunderstood. The harsh reality is that annual performance reviews are terrible tools for employee development, which must be an ongoing and individualized practice.

By contrast, performance reviews are most effective when they are upward-facing—or used to give leadership a snapshot of workforce strengths and resource utilization. Used right, annual reviews can help companies fairly distribute merit pay increases, and identify strong candidates for promotion and other opportunities.

But they’re not a replacement for the kind of continuous dialogue between manager and team that’s critical for real development.

A well-designed review process starts with clear definitions of what excellence looks like in different roles and departments. Company and human resource leaders must agree on what the company wants to achieve (hard targets) and how they want employees to behave (soft targets). Then they need to make sure that these standards are applied consistently in every department so that the data collected can be used throughout the company.

As an engineer by training, I’m partial to using OKRs to set short, inspirational quarterly goals measured by quantitative key results. But other frameworks—from Key Performance Indicators to benchmarking against industry standards—are equally valid. What’s most important is that performance reviews shouldn’t be a guessing game.

The advantage of defining what “good” looks like is that you can minimize the role that other, often superficial variables play in reviews. Hidden bias can unfairly penalize employees for factors ranging from gender and race to cultural differences and personality traits that don’t impact job performance. For example, if an introverted programmer works independently, yet receives a pro forma review that scores her low on “communication skills,” she might rightly wonder if that review is unfairly holding her back.

Once clear goalposts have been set, it’s critical to incorporate people data—not merely instinct—into the assessment process. At an organizational level, reliance on gut-level assessments makes it difficult to accurately measure performance across departments, complicating promotion and merit increases. Meanwhile, ad hoc attempts to incorporate hard numbers are often good-intentioned but unhelpful. Traditional scorecards and surveys can be scattershot, at best.

What’s often overlooked is that today’s companies gather vast amounts of digital data that extends well beyond HR basics like turnover, absenteeism, and salary. This rich people data, with the help of AI and smart algorithms, can offer an unprecedented window into performance.

Consider the challenge of peer benchmarking. Too often, managers have ranked employees based on ad hoc comparisons with fellow department members (the classic “ABC” approach). This is highly subjective and treats departments as separate silos.

Smart platforms can now look beyond titles and job descriptions and map discreet skills across an organization to ensure fair evaluations. This apples-for-apples approach ensures that a skilled writer, for instance, is recognized regardless of job title or business unit.

Grounding annual performance reviews in data also protects against a host of other biases that have been shown to impact managers’ assessments. Halo bias, for example, leads managers to rate employees favorably in all areas because they excel in one.

Recency bias leads managers to lean more heavily on achievements in the immediate past than at the beginning of the evaluation period. Hard numbers prevent these false impressions from compromising the review.

Using data, you can ensure employees are being assessed on relevant skills, that those evaluations are accurate, and that they’re being compared to the right peer group. Data can help identify the true “water carriers” in your organization—the people who consistently do great work without fuss or fanfare—and reward them to your mutual benefit.

We all know that poorly conceived performance review processes can be irrelevant, erratic, bureaucratic box-checking exercises. At their worst, reviews can even be counterproductive, inspiring cutthroat competition among colleagues who are supposed to be collaborators and encouraging avoidance instead of resolution of troublesome issues.

But it doesn’t have to be that way.

Using planning and people data, companies can craft fair and transparent review processes that generate real insights for employees, managers, and HR strategists. Now that performance reviews are back after a pandemic lull, the answer is not to scrap annual assessments but to raise our expectations.

Ryan Wong is an engineer turned CEO of Visier, a people analytics company.