Ahead of the Curve: The Future of Performance Management
How much do you love the annual performance review? If you are like me, and most other people for that matter, I bet you greatly abhor this yearly ritual.
For a lot of employees and managers, the annual evaluation, rating and ranking of employee performance is a time wasting, demotivating, and excessively subjective exercise that does more harm than good in most cases.
Instead of helping employees to improve their performance, it actually demoralizes them as they are constantly worrying about their rating and ranking and what it means for their compensation.
The issues to do with the ineffectiveness of the performance review are not new.
Instead, they have become more visible owing to the evolution that the corporate environment has undergone over the last two decades.
Unlike the past where employees were treated like droids who were only supposed to follow orders from their superiors, the modern employee is expected to be more knowledgeable, and to use this knowledge to solve problems and exercise independent judgement.
Owing to the changes in the business environment, modern employees are providing value to companies in intangible ways that are difficult to identify and quantify using performance-management systems from the industrial era.
While employees and managers alike know that current performance management systems are outdated and ineffective, about 90% of companies around the world still continue to rate and rank their employees from performance evaluations, and what’s worse, they then use these performance scores as the basis for making decisions on how to compensate employees.
This does not mean that managers are content with this ineffective performance management system.
Many would love to do away with it, but there’s one problem keeping them from doing it – they do not know how they will manage employee performance once they do away with the annual performance review.
Will employees become more complacent? Will performance slack?
Most importantly, what tool with they (managers) use to make decisions on how to compensate employees?
Dissatisfied with the current systems, some companies have started looking for alternative ways to manage employee performance in the post-performance review era, and some new, interesting ideas are coming up.
For instance, Google has jettisoned the practice of basing employee compensation on performance reviews.
Netflix also abandoned the idea of measuring performance against annual objectives, instead opting for a more fluid system that gives ongoing feedback about their employees.
Other companies such as Atlassian have done away with the traditional approach in favor of a new system that automates evaluation activities.
While different companies are approaching this issue in a different manner, some patterns have started to emerge showing what the future of performance management might look like. Some of these patterns include:
- Some companies are redefining performance. Instead of trying to differentiate between the majority of employees who fall within the average, they are putting greater focus on the outliers.
- Companies are making their performance management more objective by opting for systems that continually collect data about employees, rather than relying on the once-per-year evaluations that are more subjective in nature.
- Companies are delinking employee compensation from performance reviews, opting instead to base compensation decisions on the employee’s impact.
- Companies are shifting from backward-facing performance reviews that happen once a year to frequent events that happen as-needed. This approach is more focused on helping employees grow compared to the traditional approach.
To understand the future of performance management and how companies can stay ahead of the curve, let’s take a more detailed look into these emerging patterns.
The current performance management systems are based on models whose aim was to specialize and continually optimize disconnected work tasks.
These models were introduced over 100 years ago by the likes of Frederick W. Taylor, and this is part of the reason they don’t work today, because they were built for a different kind of work environment.
While these Taylorism-based performance management systems evolved in tandem with the nature of work over the course of the next century, their basic fundamentals did not change much.
For instance, a measure such as the number of plastic bottles a worker produced within a single working day became more complex and evolved into a balanced performance score based on key performance indicators (KPIs).
Just like the initial measure of the number of bottles produced, these KPIs were connected to the company’s overarching goals.
As organizations become larger and work became more complicated, the performance management systems also evolved and acquired layers of complexity, while still remaining based on the same fundamental principles.
Over time, the number of KPIs measured and weighted to come up with an employee’s performance score increased, while the impact of these KPIs grew smaller and smaller. This, unfortunately, created two challenges.
First, the increased number of KPIs to be measured and weighted made the process too cumbersome and decreased the accuracy of the information gathered.
In many cases, employees are actually asked to self-report this data. In addition, this increase in KPIs reduced the impact of each KPI.
This means that in some cases, the performance of employees gets measured based on KPIs that contribute to less than 5% of the employee’s performance.
This dilutes the employees’ focus because they have to focus on insignificant KPIs because these KPIs still contribute to the employee’s performance score.
Based on these performance scores, the employees are then rated against each other, and where necessary, the scores are adjusted based on distribution guidelines.
In most cases, this is done using the Gaussian distribution curves, also referred to as bell curves.
The distribution guidelines of bell curves assume that the bulk of employees fall around the average (meeting expectations), while small numbers of employees fall below the average (underperforming) and above the average (over-performing).
This model seems appealing, and is most cases used to determine compensation, with majority of employees receiving average compensation, those who over-perform receiving slightly more compensation, and those who do not meet expectations receiving slightly lower compensation.
Unfortunately, while Gaussian distribution curves are intuitively appealing, they are not an accurate reflection of reality.
According to research conducted in multiple fields – including the arts, sports, business, and academia – shows that more often than not, talent-performance profiles resemble power-law distributions rather than bell curves.
Power-law curves are sometimes referred to as Pareto curves, because they are based on the Pareto principle, which states that 80% of results come from 20% of the effort.
A study conducted by Herman Aguinis and Ernest O’Boyle in 2012 found that in most companies (except in industries where the work is highly manual and low tech), the top 5% of employees exceed the performance of average employees by 400%.
This research is what inspired Google to adopt its unfair pay policy.
In a bid to ensure retention for its top performers, Google pays employees based on their impact, rather than their performance score.
This means that it is possible to find two Google employees doing the same work earning dramatically different amounts of pay, sometimes with the pay varying by up to 300% – 500%.
Sometimes, you will even find overachieving junior employees earning more than average performers in senior levels.
Google believes that since majority of the impact comes from only a small percentage of employees, the key to maintaining high performance is to identify this percentage of employees and compensate them in proportion to their impact.
If you are thinking of applying Google’s unfair pay policy in your company, there is one thing you should keep in mind about power-law distributions.
Since majority of your impact comes from only a small percentage of your workers, there is no point in trying to differentiate between the performance of majority of your employees who meet expectations but do not over-perform.
Finding the differences in the performance of this majority and compensating them differently in proportion to their differences in performance will do little to improve overall performance, because these people do not contribute much to the overall performance in the first place.
In this case, it makes sense to get rid of performance reviews and ratings, since they only irritate and demotivate employees without having any significant impact on improving performance.
These companies are ditching annual goals in favor of more fluid and flexible objectives, replacing annual or semiannual feedback sessions with more frequent feedback discussions that happen as-needed, abandoning-backward facing rating and ranking systems for a forward-facing coaching and development approach, and placing greater focus on team performance over individual performance.
In essence, these companies are redefining what performance means to them.
They have realized that there is nothing to be gained from identifying and quantifying the small differences in performance among the majority of employees who meet expectations but do not over-perform.
Instead, they know that the best way to improve overall performance is to identify and focus on the over-performers and underperformers.
These companies have also realized that conducting annual ratings of employees based on the Gaussian distribution curve does little to improve the overall performance.
Therefore, instead of demotivating their employees with this much-hated ritual, they are opting to abandon it altogether.
COLLECTING THE RIGHT DATA
The shift to new performance management systems will require employers to start gathering the right kind of employee data.
One of the reasons that the current performance management systems do not work – and part of the reasons why employees hate them – is because they are very subjective.
Evaluating employees once a year often results in inaccurate analysis of employees.
A better way to evaluate employees would be to use crowdsourcing systems to collect their performance information continually throughout the year and from different sources.
This would give better insights and result in a more accurate evaluation of employees.
One company that is doing this is European e-retailer Zalando.
Zalando has implemented a tool that crowdsources employee performance feedback in real-time from multiple sources, such as campaigns, meetings, project launches, completed projects, problem-solving sessions, and so on.
This tool allows employees to request for feedback in real-time from their colleagues, supervisors, and even customers.
Through an online app, this tool allows these people to give both positive and critical feedback about the employee in a playful, informal and engaging way.
The responses provided on the app are then weighted based on the amount of exposure whoever gave the feedback has to the employee.
Since this data is collected in real-time, throughout the year and from multiple sources, it results in a more accurate evaluation compared to the current approach to annual reviewers where managers have to try to remember details from months ago when evaluating their teams.
General Electric has also started using a similar tool known as PD@GE which allows employees and their supervisors to keep track of their performance throughout the year.
The tool allows employees to request for feedback from multiple sources, and then maintains a record of the given feedback.
The feedback language on the tool is more focused on coaching and development instead of criticism.
General Electric hopes that by providing such continuous feedback, this tool will allow employees to continually improve their performance throughout the year, rather than waiting till the end of the year to be told where they need to improve.
The point here is that using such tools, companies can automate the collection of employee performance data, make this data more accurate and objective, and improve the effectiveness of using this data to help employees improve.
Since feedback is given in real-time, it is more accurate and credible to employees.
In addition, the fact that feedback is given throughout the year makes it easier for employees to make smaller changes to improve their performance, rather than being asked to make huge improvements at the end of the year during the performance review.
Most of these tools are also more focused on providing feedback that will help employees grow, rather than criticism that does not help the employee in any way.
We can expect that with advances with technologies such as artificial intelligence and machine learning, such tools will become even more efficient at collecting employee performance data and feedback.
In addition, such tools will also make it easier for companies to accurately identify the overachievers who make the biggest contribution to the company.
For instance, based on aggregated feedback data, the tool implemented by Zalando will automatically show the top 10% (adjustable) performers.
The tool also identifies the outliers whose performance is far removed from the average performance.
This provides managers with a more objective way of identifying the star performers and those who are truly lagging behind.
The best part is that such tools are relative easy and cheap to build and implement, which means that we might see more of them in future.
However, they are not without their challenges.
For instance, there is the risk that employees could try to game the system with the aim of being ranked as a star achiever or preventing a competing colleague from making it to this ranking.
However, advancements in artificial intelligence and semantic analysis might make it harder to game these systems.
In some cases, there is also the risk that employees might not be so enthusiastic about these systems, feeling like their every move is being watched and evaluated.
Such challenges will need to be resolved as these tools become more mainstream.
DELINK COMPENSATION FROM RATINGS
Delinking compensation from ratings seems to be counterintuitive.
For the last century or so, compensation has been based on ratings and performance evaluations, and on the face of it, this seems quite logical.
The higher your performance, the better you need to be paid, right?
With this approach, mean performance levels are matched with average salary rates in the market.
Those who beat average performance levels would also beat market rates, therefore earning more, and those who performed poorly would get lower than average salaries, thereby acting as a disincentive for low performance.
Unfortunately, this approach is consistent with the Gaussian curve distribution, which as we saw, is not an accurate reflection of reality.
The above approach has a number of challenges.
Using the above approach allows companies to compensate employees differently while remaining within an overall compensation budget.
Unfortunately, this sometimes leads to managers reverse engineering ratings so as to remain within budget.
For instance, let’s say that Peggy outperforms Alice by 5x.
However, the manager does not want to pay Peggy 5x what Alice is earning so that he doesn’t exceed the budget.
In this case, during the performance review, the manager might reverse engineer the ratings, lowering Peggy’s ratings and slightly increasing Alice’s ratings such that it appears from the ratings that Peggy’s performance is double (rather than 5x) that of Alice, which is also commensurate with Peggy’s compensation compared to Alice’s.
However, the problem is that Peggy already knows she works way harder than Alice. In this case, she might assume that her efforts are not being recognized.
This could end up making Peggy cynic, demotivating her, and making her less collaborative.
In addition, it is good to understand human psychology and how it plays out in regards to these ratings.
Generally, people are more afraid of potential losses than they are motivated by potential gains. For instance, the pain of losing $100 is greater than the joy of gaining $100.
When performance ratings are linked to compensation, employees pay greater attention to how much they are potentially losing out on rather than how much more they could potentially earn by improving their performance.
Unfortunately, this serves to demotivate employees rather than encourage them, yet these performance ratings differences result in only very small variations in compensation.
Considering that only a small percentage of employees are outliers, there is not much sense in linking compensation to performance, since this will only serve to demotivate majority of your employees.
This is why many technology companies are doing away with bonuses that are pegged on performance ratings, and instead opting for bonuses that are pegged on overall company performance.
This gives employees the freedom to focus on work and experiment with new, innovative ideas without fear that a marginal rating differences will affect their salary.
However, to keep the outliers who contribute the most motivated, these companies also offer special rewards to overachievers.
This way, companies eliminate the anxiety of compensation from the majority of employees, while at the same time having a way for rewarding employees who put in outstanding work.
In addition, researchers like Dan Pink have already shown that employees are more motivated by things such as purpose, autonomy and mastery than they are by money.
Instead of relying solely on compensation to motivate your employees, you can motivate them by assigning them priority projects and customers, giving them access to assets, empowering them and giving them greater autonomy over their work, recognizing their effort, and so on.
By severing this link between performance and compensation, managers will also be able to pay greater attention to inspiring their teams and building their capabilities without worrying about tracking and rating performance.
GREATER FOCUS ON COACHING
In the industrial age, performance management was pretty straightforward.
An employee was required to produce 500 bottles or 1000 pins per day.
Poor performance meant that the employee was not hitting their quota, and improvement simply meant going back to hitting or exceeding their quota.
Today, however, with constantly changing objectives and with jobs increasingly requiring employees to exercise their judgment, performance management has become a lot more complicated.
Today, performance management systems should pay greater attention to coaching employees, and doing so at scale.
Coaching makes it easier for employees to stretch themselves, work with constantly changing goals, and take on greater responsibility and autonomy.
Different companies are coming up with different ways of implementing employee coaching at scale.
For instance, we saw that General Electric is changing the feedback language within their PD@GE tool so that it is more focused on coaching rather than criticism.
Companies like Zalando are using crowdsourced feedback to let employees know what is working and where they can do better.
Several other companies are focusing performance discussions on what employees should do in future, rather than what they did in the past.
More and more companies will have to start embracing this to approach if they are to remain ahead of the curve when it comes to performance management.
Current Taylorism-based performance management systems were developed for the industrial age and are thus becoming outdated and ineffective as the future of work becomes digital.
Already, companies in industries such as finance, technology and media, which are leading the pack when it comes to embracing the digital future of work, have started pioneering new, innovative approaches to performance management that are suited to the future as works becomes digital.
If you want your company to remain competitive when it comes to attracting and retaining top talent and optimizing performance, you will need to quickly follow suit.
Of course, these patterns that we have discussed above will play out differently depending on the company, and we can expect that other patterns will emerge as more companies realize that the current model is becoming obsolete.
What we can be sure of, however, is that change is coming, and the best thing you can do is to start thinking of how you can start transforming your performance management system before the times catch up with you.
The opera has long been regarded as the epitome of entertainment among the wealthy, as well as a …
In this article, we will look at 1) what is Soundcloud?, 2) how to list successfully on …