This is What Peak Performance (Management Season) Looks Like

By David Wohlreich and Tristan Orford

Well, it WAS the beginning of the year when we started writing this piece, and work (including as part of the performance cycle) bogged us down for a bit. We’ll still indulge in a bit of overdue post-holiday reflection, which for us involves the books we read and games we played. Hades and Shadowlands are terrific, and one of your authors is still (yes, still) eagerly awaiting more supply of the newest generation of graphics chips for his PC so he can explore some of the newest games in their full glory. For many companies, it’s also the heart of performance management and compensation season (or, at time of publication, just after those cycles have completed)! Hooray!

Ok, there probably aren’t many people who greet performance management season with thrown confetti and trumpets. For some, it’s a lot of work on top of already heavy product development calendars. For others, it seems like a giant waste of time. And for a very, let us say “vocal,” cohort, performance management is little but a tool in managerial oppression, bureaucratic futility, or both. Even among organizations who have a strong commitment to performance management, there’s vibrant debate over what it looks like, or should look like, what data to collect or create, and even if participants in the performance management cycle should see the mechanisms of it at all.

At its best, a performance management cycle is a time for reflection, development, calibration, and meaningful engagement across and within teams.  And it should be one of the more meaningful interactions between managers and the people on their teams.

This time is also particularly meaningful for us folks in the world known as total rewards (compensation, benefits, perks, etc.) due to the fact that performance is usually – or should be – the basis of making decisions around compensation which come shortly thereafter.

We won’t cover this entire branch of HR here (nor do we have the expertise to do so), but there are a few topics we often see either implicitly or explicitly tied to total rewards that we’d like to explore to hopefully help others navigating this always-interesting and very dynamic (and contested) space.

1. Performance management – yea or nay? Should companies stop this stuff?
2. Should we connect pay and performance discussions or should those be totally independent?
3. Everything old is new again; didn’t we just get rid of performance ratings?
4. How much data, what kind of data, to what end?

Our posts can sometimes be long, we get that. If you’re in a hurry, if the technical parts sound like mumbo-jumbo to you (you might be right), or if you plain don’t like longer-form writing, you can focus on the bolded segments and get most of the main points we’re looking to make. 

The usual disclaimer

While we may allude to legal considerations, we are not lawyers, and any professional is advised to consult an experienced employment or labor attorney prior to engaging any area with a lot of complexity – like pay or benefits. Similarly, while both authors are employed as total rewards professionals, these words are our own and are informed by as broad a review of the space as we’re capable of. They do not necessarily represent the views or actions of our current or any past employers. While both authors have experience in global rewards, we are both Americans and recognize our frame of reference is anchored in U.S./California employment practices (we try to take a global mindset, but won’t always be successful). We also work in largely pay-for-performance environments and so are further embedded in that perspective.

1. Is the exercise of evaluating performance valuable in the first place? Aren’t there other things we should be doing with our time?

There’s an enormous body of work concerning performance management that’s largely or entirely separate from total rewards: succession planning, training and development, organizational design and development, talent acquisition and strategy, employment litigation mitigation, and much more. Almost anything you want to do at a company, business unit, function, team, or product level involving people requires some understanding of how those people are doing – what skills they have, how they perform, who’s doing well and ready for a new challenge and who’s failing. It’s also critical to reconcile the needs of the business and the person through this exercise. Hopefully there is overlap between the two and an opportunity to build on that, but either way, the understanding of performance is a crucial part of the process. There are incredible HR leaders with deep expertise in this space who are frankly much more equipped to discuss these items than your authors. We’ve been fortunate to work with several.

Before we go further, it’s worth sparing a moment to address the implications for the philosophy that sits at the center of pay programs in most organizations: pay for performance. The centrality of performance is evident in the name, and there is a clear need to get a good handle on that key performance input for any “pay for performance” program. Even organizations with pay models centered on other variables (like tenure) still leverage performance in how many key talent management decisions are made, such as who gets put in which roles. And that, of course, translates into pay. One of the benefits of a pay for performance system is that it provides us with an ability to evaluate and continually improve our talent outcomes. With good intent and sound methodology, this means we can make our organizations ever more effective and equitable over time. And, the imperfections of today aside (we don’t have all of the answers yet on what a perfect system looks like, and it likely varies based on context), it’s the most viable option we have to allow us to make that progress. Therefore, we’d encourage folks to think about this as an investment in a better future world of work. The work may not be as grand as the vision, but it’s a necessary part of achieving it. In summary, investing in better understanding performance, with data, is one of the best ways to ensure you’re following a pay for performance philosophy.

Now, some will criticize pay for performance itself as a system for managing pay in an organization. Various critiques are leveled against it, not least among them Dan Pink’s famous TED talk and book about motivation and rewards. The TLDR is that by offering extrinsic rewards for performance in complex work, we risk crowding out their intrinsic willingness to do work and be creative. We’ll likely return to this topic in a future piece, but for now we’ll offer two main comments. 

First, Dan Pink makes a point! But it’s worth not taking it too far. In other words, yes, we should be wary of crowding out intrinsic motivation. But that doesn’t mean we should never use extrinsic rewards, just that we shouldn’t over-rely on them or use them in places where they don’t make sense (i.e., in very open-ended problem spaces). The most troubling application of this that your authors have seen is when business leaders (not at our employers, to be clear) seem to use Pink’s work as a justification to pay employees less! “A bonus won’t motivate great work from folks, so let’s not waste money on bonuses” will make any rewards professional’s blood run cold.

Second, we should also distinguish between compensation as motivation vs. as recognition. The difference here is subtle, but important. It’s true that the track record of motivating specific outcomes is checkered, so worth approaching with caution. Yet all companies recognize that not all roles are the same in terms of the mix of skills required, difficulty to execute, and impact on the organization. Different people and roles bring different value, and if we don’t at least attempt to reflect this, then we risk losing people to other organizations who will. The function of compensation as a “hygiene” factor, namely that people expect it to at least meet a baseline expectation, is well-supported, and paying for performance is a part of meeting that baseline. 

Said more simply, you can’t pay for passion – but if you’re not willing to pay your top performers well, they’ll take that passion to other employers who will.

Yet even when viewed exclusively through the lens of total rewards, performance management is critical to a key task almost every HR or compensation professional is asked to do on a regular basis: differentiate rewards while paying people fairly (or equitably, or defensibly, or whatever term resonates).  The degree to which you can justify different pay decisions, positions, and outcomes is often directly tied to your performance management maturity. This applies equally to individual outcomes and programmatic or statistical outcomes. 

By way of example, consider a situation where at a hypothetical company, women received higher bonus payouts as % of target than their male colleagues last year.

Given just the facts above, we don’t know if we have a problem or not because we don’t know what’s leading to these outcomes. It could be:

1. Similar performance isn’t codified the same – i.e., same performance, different evaluation (a gap in performance management calibration)
2. Similar performance evaluations aren’t translated into rewards the same – i.e., same performance, different pay outcomes (performance management is calculated, but pay and bonus decisions aren’t aligned with performance)
3. Different performance is accurately recorded and properly translates into differentiated rewards – i.e., different performance, different pay outcomes (Employee A is a stronger performer than Employee B, specific women had a better year than specific men)
4. Other, possibly confounding variables – i.e., perhaps men are more likely to get promoted given the same profile as their female colleagues, resulting in more high-achieving but under-leveled women and more developing but higher-leveled men
5. Some combination of the above

Each of these can also contain more levels of detail, but the problem is that without performance data, you don’t even know whether you should be going down path A, B, C, D, or E. And this isn’t just academic; the practical measures you might take to correct any issues you discover will be totally different depending on the source of the problem. Without performance information, your ability to spot and address issues in the short term will be limited, and in the long term you won’t have effective ways to prevent issues from recurring – or getting worse. In other words, you can’t identify and address the root cause if you don’t have visibility to the roots.

At its core, a data-rich performance management process is a way to make visible (particularly at scale and across groups) what is impossible to view or manage intuitively beyond an immediate team. None of this is to say that collecting performance data is a panacea, but without it, progress is much, much harder.

2. Should performance management and compensation processes be linked or not?

How closely performance and pay should be linked also gets debated a lot. Basically, it goes something like this:

• Person A: “We’re a pay-for-performance organization; of course the two should be connected and we should evaluate those connections as managers and with our teams.”
• Person B: “If we connect performance and pay, then people won’t feel safe exposing their vulnerabilities, which is bad generally and means they won’t develop.”

Both of these perspectives contain some relevant insights. And the good news is that there is a solution that can help both achieve their desired end result.

Firstly, we should acknowledge that performance and compensation are usually linked, and there should be a really clear connection between them.  Again, we’re assuming a pay for performance organization here, and it’s also worth noting that this usually shows up in more pronounced ways in incentive pay like bonuses than it does in salaries. Some organizations lean farther into this – the top engineer can make 10x what a “normal” engineer earns. Some are less comfortable with that kind of profound deviation; maybe the top engineer earns at most 20% more than their peers. But almost every organization identifies a need to pay more for better work – whether that “better work” is delivered at higher scope and complexity (leading to promotions and different role definitions) or as variations on a core job function (being stronger, more valuable, more externally-desired worker). So then it becomes a question of how (not if) to link performance and pay. 

While this sounds obvious, it’s worth exploring, as often organizations have unexamined assumptions about the connection between pay and performance – and that can lead to quite painful misalignment. Added to this complexity is simply the difficulty in creating a definition of performance that will resonate with everyone. While often assumed for so-called “knowledge” work, this ambiguity surrounding performance applies to more manual work as well, as elements like speed and throughput vs. quality and overall uptime could cause two workers in similar roles to have very different perspectives of “high performance.” Given this context, it’s very unlikely that a single interaction will capture and deliver the totality of performance for a given worker.

It’s something of a cliché that performance discussions should happen regularly and interwoven throughout the year, and that’s true. But there is value in a formal and structured performance calendar as well. And here we recommend two structured touch points, one to align on performance, and a second to discuss compensation.

Typically, the first conversation will be longer, as there are many related topics that will likely make sense to cover. Alluding to comp decisions can be OK here, i.e., setting directional expectations around comp outcomes, but it’s important to caveat that nothing is final at this point. This is more to avoid people carrying misconceptions (positive or negative) that can cause them angst. And the degree of expectation setting should be relative to how much context and experience the manager has. In other words, if you aren’t sure, better to stay silent. Often this expectation setting is colloquially spoken about in terms of “no surprises” (or at least, no bad surprises) when it comes to the subsequent compensation discussion, and we’d agree that’s a good way to think about it. 

Many companies will subdivide this first conversation even further, for example using some version of a quarterly model where there are touch points to discuss performance at multiple points during the year. That’s great, and in fact the more regular this is, the more space there is to discuss development without the overhang of the impact on pay. The key focus is to ensure there is alignment on performance (not necessarily agreement – employees don’t have to agree with the evaluation of their performance to understand and align on how it has been evaluated) before going into the compensation discussion. We’d recommend ending those conversations with something like: “We can, of course, revisit this topic if you ever have questions, but since this conversation is about understanding my assessment of your performance, we want to make sure you’ve gotten as much information out of it as possible. To that end, are there any questions you have about my assessment of your performance this year that I can help clarify?” 

The second conversation should go deep on comp. This is the time to present the numbers and explain the reasons behind them. The content discussed during the performance discussion should at least be alluded to so that the linkage between the two is clear to the individual. For both managers and HR professionals, a few items to be really grounded in before having this discussion. If any are unclear, it’s a sign for more alignment as a management team before having the conversation:

1. The major inputs to the decision, namely the performance context discussed previously.
2. A baseline knowledge of how the various comp programs (salary, bonus, etc.) work and why someone participated in them or not and why they got what they did is crucial.
3. Overall ownership of how the manager’s inputs were factored into the ultimate decision. If there’s a possibility that, when pressed, “someone else made me do it” is the explanation, that’s a warning flag.

We’d recommend ending this conversation with something like: “We can, of course, revisit this topic if you ever have questions, but since this conversation is about understanding your compensation, we want to make sure you’ve gotten as much information out of it as possible. To that end, are there any questions you have about your pay this year that I can help clarify?”

3. I thought we got rid of performance ratings, aren’t we going back to the bad old ways?

If you’re at an organization that regularly gathers performance data and where this isn’t the topic of debate, this section will be less relevant for you. For the rest of us, let’s get into it.

This topic has currency given the performance management trend of the 2010s was moving away from performance ratings, “taking people out of boxes,” and moving to a more freeform manner of talent assessment. If you happened to miss this, basically a number of organizations, including but not limited to the technology industry, moved away from an annual performance review model of performance management. As part of that, many also did away with summary performance ratings, or in many cases didn’t implement any means of collecting that data if they hadn’t previously done so. There were some good intentions behind this shift, and some real underlying problems to solve, but we’ll argue that there are some reasons to further refine the model.

First, the legitimate grievances. Below is a representative but not exhaustive list of things that performance ratings are often associated with, and which are objectively worth avoiding.

• Using a single data point (e.g., an indicator of performance in the most recent period) to drive multiple talent management or rewards programs (e.g., salary increases, bonuses, stock grants, role changes, development opportunities). Most of these programs are different enough that they need different data points to inform them.  At the very least, one single data point is way too little to inform all of them effectively.
• Labeling people, especially with a number. Being labeled is, at best, a mixed experience, and being labeled by a number is not something most humans relish.
• The risk that the performance label (meets expectations, needs improvement, etc.), is all the employee will hear – and they’ll miss the nuance of the conversation. Related to the above, if managers use the category/label as the way to have  a performance conversation with someone on their team, it can often short-circuit the shared understanding which is really what you’re trying to achieve.
• Using a number or rating as a substitute for thorough communication between a manager and someone on their team about that person’s performance. An objective assessment process and a human conversation are good complements to each other, but neither does a great job substituting for the other.
• Only talking to people about their performance once per year, after the assessments have been finalized and often as part of delivering compensation outcomes. This is a lot of weight to put on one interaction, and there is a high likelihood of misunderstandings building up in the interim.
• Using performance data for something it wasn’t initially designed for. As an example, it’s not uncommon for organizations to use these ratings as a means of assessing who should be let go in a workforce reduction. While perhaps part of the picture, this data certainly shouldn’t be used in isolation for such a purpose.

You’ll notice that none of the above are inextricably tied to the process of collecting performance data. Rather, they are understandable but avoidable outcomes of how programs using that data are sometimes managed.

To this end, it’s really important to clarify what is meant when people talk about ratings, and distinguish ratings (a very specific and emotionally charged topic) from the much broader and objective concept of performance data. While this may come across as an overly pedantic distinction, in this case it’s critical. If this isn’t done, then every discussion on the topic of performance data risks an emotional hijack, circling back to concern about ‘going back to ratings.’ This will in turn make a coherent and productive dialogue on the topic, and progress on it, very challenging. We’d recommend something like the following to help get really clear on what is important to do, and not do, at your organization:

1. Articulate the features you want to avoid (e.g., the list above).
2. Document the features of a performance and reward system you’d like to have.  Examples might include: differentiated rewards outcomes, holding people accountable for expectations, and pay equity.
   1. Corollary, these require data on performance to operate effectively, especially at scale.
3. Build a system for 2 and show the measures to avoid 1 which will be designed into your programs. For example, ensuring that your managers can draw a distinction between different assessments and how they should be used. You could even use data to help hold them accountable for doing so!

While all that sounds simple, we suspect that making progress on this point will not be a linear path for you. Getting alignment on a complex and emotionally charged topic is rarely easy. Hopefully this helps shed some light on a viable route, and good luck!

4. What amount of data is the right amount to collect?

This topic comes up often in the area of performance management as any system will almost by definition create data – and those of us called to design performance management programs have a nearly infinite number of variables to consider and deploy to assist the company in having robust performance management discussions and making real talent movements based on them. Of course, we facilitate this process, but we don’t own it. No one from HR or Compensation is going to have as robust and context-rich an understanding of an individual employee’s performance as their direct management and team. Making that knowledge visible, calibrated across the organization, and trackable over time is one of the key benefits of the performance management cycle.

To get a bit more specific and granular:

One is never enough

Performance cannot be a single measure. There are both time dimensions (what’s the measurement period, how often will you evaluate, and how is historical information treated) and components of results, approach to work/impact on others, and professional craft. Performance over time is critical, as it takes time to evaluate results of intervention and sustained strong performance is a key input to promotion decisions. At the same time, performance three years ago is largely irrelevant other than as part of a trend – no one can rest on their laurels, but so should everyone get a chance to improve. You likely have many talent outcomes you want to drive across your organization, many pay programs, promotions, development, and more. You will likely need many data points to inform these programs effectively. Starting with the most important handful of these is OK, you can always build from there.

Invest in clear definitions

Terms like “performance” are really broad and hence can be really confusing to try to nail down if left open-ended. A common example is the differentiation between individual results/performance in a given time frame and craft (how skilled someone is at a given role). These are not necessarily the same thing, although they are often connected. If not clearly distinguished, the conflation between the two can cause many people to lose patience with the process of gathering performance data.

While a totally comprehensive definition is not possible or desirable, getting as clear as possible is worth doing. What specifically is being measured, over what time period, how should it be assessed, what will it be used to inform?

Take five

Whole treatises have been written about the ideal performance management scale, numeric or descriptive (1 – 5 or Below Expectations/Meets Expectations/Exceeds Expectations), and the truth is we don’t have the answer. There may not even be one correct answer beyond “it depends,” (which every business leader is probably tired of hearing from their compensation folks). That said, balancing specificity with managing diminishing returns and false precision often leads companies to a scale with more than three elements and fewer than ten. There is a reason a 5 point scale is the most common performance rubric out there, even if it’s much-maligned. More important than the scale, of course, is how you use it – and even more importantly, how you train for it, calibrate it, and ensure it’s working. If you have a 5 point scale but everyone is a 3 or a 4, you don’t actually have a 5 point scale. 

Can’t we just turn all this over to computers?

There’s a lot of interest in formulaic approaches to compensation and performance, given increased scrutiny related to pay equity and often lean rewards teams. Purely formulaic approaches, though, where there’s no manual intervention, are incredibly hard to build and calibrate, and ironically require a great deal more direct intervention to ensure inputs are fully validated. There is also an inherent limitation to assessing performance of most roles, namely that it’s effectively impossible to do so in a totally quantitative way. The goal is to be as consistent as possible, but this still requires human judgment. That said, consistency and fair treatment are the goals for any performance management process. And collecting the data consistently does allow data analysis to help supplement human memory and judgement. Time invested in ensuring that outcomes are in fact equitable, validating and testing data outputs, and ultimately working to make sure the performance management ecosystem aligns with the values and expectations of the business is never wasted time.

Thanks for reading, and even if this doesn’t lead to a shout of joy the next time your performance cycle kicks off, hopefully it at least reinforces that the time invested in it is worthwhile.