ERE: 50 Top Problems with Performance Appraisal – the Most Dreaded HR Process

imageIt was fascinating for us to read about the Top 50 problems with Performance Appraisals on ERE.  It was even more amusing to identify how FairSetup addresses most of these problems.

The article written by Dr. John Sullivan starts out with a promising quote:

90% of performance appraisal processes are inadequate. – survey

OK, something we already know.  He goes on to say:


In conversations with HR leaders and employees, the talent management process that suffers from the most disdain around the world is the performance appraisal. It’s one of the few processes that even the owners of the process dread.

Most ignore the shortcomings of performance appraisals and suffer through it, but that’s hard to do once you realize how incredibly expensive the process is. In 1996, Frederick Nickols estimated the cost at just under $2,000 per employee. My estimate, which includes a managers preparation time, employee time, HR processing time, opportunity costs, and advances in technology, still puts the process cost at over $2,500 per employee per year. If you choose to take on the challenge of revising your performance appraisal process, the first step is to fully understand the potential problems associated with it.

$2,000 per employee…  given that Performance Appraisals can take days and sometimes weeks to complete, that is hardly surprising.  And I am pretty sure that Mr. Nickols didn’t take into account either opportunity costs, impact on culture, and other consequences that are more long-term and thus more difficult to quantify.

Moving right along, let’s skip to some problems:

1. Don’t assess actual performance — most of the assessment that managers complete focuses on “the person,” including characterizations of their personal “traits” (i.e. commitment), knowledge (i.e. technical knowledge) or behaviors (i.e. attendance). While these factors may contribute to performance, they are not measures of actual output. If you want to assess the person, call it “person appraisal.” Performance is output quality, volume, dollar value, and responsiveness.

Performance: In FairSetup, by virtue of quick cycles that involve expectation-based assessments, the focus is not just on actual performance, but also on the impact that that performance has on the system.

2. Infrequent feedback – if the primary goal of the process is to identify and resolve performance issues, executing the process annually is silly. A quality assessment/control program anywhere else in the business would operate in real time. At the very minimum, formal feedback needs to be given quarterly, like the GE process.

Regularity: FairSetup works atop existing workflows.  So, if a team has weekly meetings, then performance-related data is collected on a weekly basis.  However, because of the frequency of data collection, there is only one data point collected per cycle: did a person perform to expectations?

3. Non-data-based assessment — most processes rely 100% on the memory of those completing the assessment because pre-populating the forms with data to inform decisions would be too difficult (cynicism). In addition, most assessment criteria are “fuzzy” and subjective.

Data-Based Assessments: it’s impossible to do purely data-based assessment except in the very few cases when there are clear KPIs (Key Performance Indicators).  With FairSetup the idea is to measure how one performs with respect to expectation.  The short cycles make for a regular “fuzzy” approximation that is as close to a data-based assessment as possible, as both the evaluator and the evaluee perceive the measure as “fair”.

4. Lack of effectiveness metrics — many accept that the goals of the process are to recognize results, provide feedback to address weaknesses, determine training needs, and to identify poor performers. Unfortunately, rarely do process owners ever measure their processes’ contribution to attaining any of these goals. Instead, the most common measure relating to performance appraisal is the percentage completed.

Metrics: as previously, expectations (especially among functionally dependent or similar peers) provide the best form of metric. 

5. Lack of accountability – managers are not measured or held accountable for providing accurate feedback. While they may be chastised for completing them late, there is no penalty for doing a half-assed job or making mistakes on them, which is incredibly common. One firm attempting to remove a troublesome employee found that the manager had rated the individual the highest within the department and awarded them employee of the year.

Accountability: the connection between impact and compensation creates direct accountability to all on the team – including the manager.

Process-related Problems

6. Disconnected from rewards — in too many organizations, getting a merit raise, bonus, or promotion is completely disconnected from an employee’s performance appraisal scores. When there is a weak link, employees and managers are not likely to take the process seriously.

Rewards: under FairSetup, individuals are compensated by way of a bonus, which is connected to their impact as calculate using the FairSetup model.

7. No integration — the process is not fully integrated with compensation, performance management, development, or staffing (internal movement). A lack of integration and coordination leads to duplication and missed opportunity.

Integration: under FairSetup, the process is integrated into the career path using levels.  This makes it easy to set expectations for what people need to do to be able to move up the corporate hierarchy.

8. Individual scores exceed team performance — without controls, quite often the average score of team members exceeds the actual performance of the team (i.e. the team reached 80% of its goals but the average performance appraisal for its members was 95%).

Individual vs. Team: because the bonus that is paid out using the FairSetup model generally connects to KPIs of the business unit, it doesn’t matter how people are evaluated: they get what their share of compensation in relation to their peers.  If everyone performed at 100%, but the team only did 80%, then everyone gets 80% of what they could have received.  This affords relative transparency and performance evaluations.

9. Each year stands alone — each performance appraisal by definition covers a finite period of time. However, if the goal is to assess potential and identify patterns, an employee’s performance must be assessed over multiple years.

No History: under FiarSetup, impact is accumulated over time.

10. No comprehensive team assessment – although individuals on the team are assessed, there is no simultaneous overall assessment of the team. Often contingent workers on the team are not addressed at all.

Team Assessment: this one is a little tougher unless you have clear KPIs to determine the team’s success.

11. A focus on the squeaky wheel — most performance appraisal systems focus on weak performers. There is significantly less focus on top performers and thus there is no system to capture their best practices and then to share them with others.

Squeaky Wheel: under FairSetup, the goal is outcomes.  So instead of focusing on the negative, the team generally acknowledges the shortcoming of underperformers and fills in the gaps while using self-throttling and evaluations to ensure that credit (and compensation) are distributed fairly.

12. Little legal support — performance appraisals may be an executive’s worst enemy in grievances and legal proceedings. Even though the process may be flawless, poor execution by managers often results in performance appraisals that do not aid in a disciplinary action. Errors may include “unfettered discretion,” improper handwritten notes, generalizations about race, gender, or age, and appraisals that do not match the performance data. At my university, a study demonstrated that while Asians got the highest performance score, they somehow managed to get the lowest average pay raise. When the HR director was confronted, he was furious that anyone would calculate and expose the obvious discrimination.

Legal Support: FairSetup affords clarity with notes being taken along the way when expectations are missed or exceeded.

13. No second review — even though the process may have impacts on salary, job security, and promotion, in many firms the assessment is done by a single manager. If there is a second review, it may be cursory, and therefore not ensure accuracy or fairness.

Transparency: in FairSetup, the process is visible to all participants, so while the manager is the one officially recording the review, it is done in a way that everyone can contribute to.

14. Not reliable or valid — most process managers do not regularly demonstrate with metrics that the process is consistently repeatable (reliable) and that it accurately assesses performance (valid).

Validity: the validity is afforded by the fact that every measure is effectively validated by the manager, individual reviewed, and the team.

15. Cross-comparisons are not required — one of the goals of the process is often to compare the performance of employees in the same job. Unfortunately, most appraisal processes (with the exception of forced ranking) do not require managers to do a side-by-side comparison, comparing each member of the team with one another.

Comparisons: the peer-based nature and use of impact (which is calculated in relation to other participants), ensures ongoing comparison not just between employees performing the same function, but it also creates an opportunity to compare employees that have different skillsets.

16. Assessments are kept secret — although a salesperson’s performance ranking may be posted on a wall, performance appraisals are often kept secret. An overemphasis on privacy concerns might allow managers to play favorites, to discriminate, and to be extremely subjective. Keeping ratings secret allows managers to avoid open conversations about equity.

Transparent: once again, FairSetup is transparent to everyone thus allowing for open outcomes-focused conversations.

17. Process manager is not powerful — often the process is managed by lower-level HR administrators without a complete understanding of performance and productivity.

Peer-power: nor should they be powerful.  Peers are the best way to drive behavior, so the process manager’s job becomes to do what they are actually good at: managing communication and expectations.

18. No process goals — the overall process operates without clear and measurable goals, and as a result there is little focus.

Focus: focus should come from both the top and the bottom (upwards management).  FairSetup ensures that everyone is aligned towards the same goals and, if such are not identified, forced their identification.

19. Not global — most processes and forms are “headquarters centric,” failing to address cultural, language, and legal differences.

Eh…  I don’t think we can comment on this one just yet, except to say that fairness seems to be a universal value and a common ground when cultural differences are present.

20. Forced ranking issues — although forced ranking has some advantages, using it may result in significant morale and PR issues.

Forced Ranking: forced ranking makes sense when it reflects performance in a fair manner.  That said, that is almost never the case – especially if a company does a good job hiring talent.  What generally happens in a homogenous group is that, rather than seek to be at the top, people fight not to be at the bottom (as was the case with Microsoft – we discussed this here) and  In FairSetup, if everyone is working hard, everyone can have equal impact.  But if there are differences, the system naturally reflects the ranking.

21. No ROI calculation — HR fails to do a periodic business case justifying the value added compared to the time and the cost of the process.

ROI: FairSetup connects performance to compensation making ROI transparent. Consider what happens if you take 100 people each of whom makes $100,000 per year and you increase their performance by 5%.  You just generated $500,000.

Instrument (Form) Problems

22. Doesn’t address diversity — all too often, the same appraisal form is applied to a large but not homogeneous group of employees (i.e. all hourly, all exempts, all managers etc.). As a result, the assessment form does not fit the job. Only management-by-objective-type approaches address individual needs.

MBO: management-by-objective (MBO) is somewhat different from the concept of “expectations” in FairSetup, but it is similar enough to reflect objective-driven diversity.

23. The process does not flex with the business – rarely does any portion of the appraisal process flex to address changing business objectives.

Adjustment to Context: FairSetup increases the variable pay component making compensation significantly more elastic and thus focusing everyone on optimal adoption to business objectives.

24. The factors are all equal — most forms treat all assessment factors as if they are of equal importance. Instead, they should be weighted based on their relative importance in a particular job (i.e. a janitor’s customer service rating should be weighted lower than for a salesperson.

Levels: FairSetup uses levels (see level of expertise) to reflect relative worth.

25. Inconsistent ratings on the same form — it is not uncommon for managers to put one level (high, average or low) of ratings in the Likert scale portion of the form, but another level of rating in the “overall assessment” box. The final narrative portion of the assessment may contain still another completely different level of assessment.

Simplicity: FairSetup is simple using a single measure that results in aggregate calculation.

26. Disconnected from job descriptions – in many cases, the factors on the form are completely different from the factors on an employee’s job description, bonus criteria, or yearly goals. This can confuse employees and cause them to lose focus.

Job Description: expectation by peers or by function connects the system directly to the job description.

Manager/Execution Problems

27. Managers are not trained — in most organizations, managers are not trained on how to assess and give honest feedback. If the process includes a career development component, it is even more likely that managers will not know how to enhance the career path of their employees.

Training Managers: training is necessary, but groups become more autonomous, making the manager’s job easier.  Moreover, with FairSetup, managers can be trained to manager rather than to scrutinize.

28. Managers are “chickens” — some managers will do almost anything to avoid tough decisions or confrontation. Some provide no differentiation and spread “peanut butter” (an even distribution) to avoid it, while others give everyone “above average” ratings. Some managers will provide feedback that is extremely vague in order not to offend anyone. Rarely if ever is anyone immediately terminated as a result of the process.

High-Stakes: under FairSetup, instead of having a high-stakes discussion once a year, the manager and reports communicate on a regular basis.  Furthermore, the collaborative process of evaluation is one that makes negative discussions easier to have and allows them to be had among peers before managers step in.

29. Gaming the system — often managers artificially rate individual employees to save money or to keep employees from becoming visible for promotion. Some selfishly give a score just below that required for a pay increase, while others give scores just above the point where they would be required to take disciplinary action.

Hard to Game: FairSetup establishes ongoing communication making it very difficult, if not impossible, to consistently game the system.

30. Recency errors — managers, especially those who don’t consult employee files and data, have a tendency to evaluate based primarily on events that occurred during the last few months (rather than over the entire year).

Time-Independence: evaluations happen on a regular basis.  Generally, if a team meets weekly, we recommend that FairSetup is used to collect information on a weekly basis.

31. Corporate culture issues — subjective appraisals can restrict cultural change in organizations. In some organizations, there are cultural norms and values that influence performance appraisals. For example, in one organization new hires were automatically given an average rating for their first year, regardless of their actual performance. One top performing hire I knew abruptly quit after receiving this cultural gift.

Immediate Feedback: under FairSetup, there is no delay, so such issues are generally avoided.

32. Inconsistency across managers — some managers are naturally “easy raters” while others are not. As a result, employees working under easy managers have a better chance of promotion due to their higher scores. In firms that rely heavily on the narrative portion of the assessment, having a manager with poor writing skills may hamper an employee’s career. Without “benchmark” numbers to set as a standard, inconsistency is guaranteed in large organizations.

Inconsistency: this one is a bit tougher and depends on a specific deployment.  However, as the manager rates openly in front of other people, reports often themselves exert influence on managers to ensure fair treatment of themselves and their peers.

33. Managers don’t know the employee — managers of large and global organizations, as well as newly hired and “transferred in” managers may be forced to do appraisals on employees they barely know. Recently promoted managers may be forced to assess their former friends and colleagues. Following a merger, managers are likely to be confused about whether to focus on the whole year or just “post-merger” work.

Easy-in, Easy-out: peer-based evaluation process shifts the burden of knowledge on peers, who are less likely to change than the manager.  At the same time, the system keeps a history of performance thus making it significantly easier for a manger to enter into a new team.

34. Secret codes — I did some work with an army unit where by custom literally everyone got a perfect numerical score. So assessments by higher-ups were made as a result of interpreting “code words” in the small written narrative portion of the assessment. Unfortunately, if your commander didn’t know the code words, your army career was limited.

Simplicity: the only thing that matters is expectation.  And, at the end of the year, there is a quantified history eliminating the danger of “interpretation”.

35. Mirror assessments — most people, and managers are no exception, have a tendency to rate people like themselves more positively. This can result in discrimination issues.

Collaborative Assessment ensures that such does not happen.

36. Managers are not rewarded — managers that go out of their way to provide honest feedback and actually improve the performance of their workers are not rewarded or recognized.

Uniform Rewards: managers are rated as well as other team members.

37. Managers don’t own it — managers often feel they don’t own the process, so they invest little in it and proceed to blame HR for everything. Managers would embrace it instead of grumbling if they were presented with a positive correlation proving that managers who did excellent performance appraisals were among the highest performers with regards to business result and bonus awards.


Employee/Subject Problems

38. High anxiety — because the process is so subjective and no benchmark performance numbers are set in advance, uncertainty can cause many employees high levels of anxiety weeks before the process. Managers may also be anxious because of the uncertainty related to the an employee’s reaction. I know one employee who sincerely thought she was going to be fired prior to her assessment but ended up being the highest rated employee on the team. Employees should have an accurate idea of their assessment long before any meeting is scheduled.

Ongoing assessment removes the anxiety – instead of having one high-stakes discussion, there are multiple small-stakes assessments.

39. One-way communication — some managers simply give the employee the form to quickly sign and they don’t even solicit feedback. Many employees are intimidated by managers and the process, and as a result, they say nothing during or after the appraisal.

Full accountability: managers are accountable to reports – their job is to manage, to allocate resources, etc.  The focus on outcomes drives reports to keep their managers accountable.

40. Self-assessment is not possible — if an ambitious employee wanted to self-assess their performance midstream (in order to improve), most processes do not provide access to the instrument. Providing each employee with a virtual assessment scoreboard and performance management process would be an ideal solution.

Ongoing assessment with regular self-assessment directly addresses this issue.

41. No alerts — most processes do not allow an employee to be notified midstream should their performance change to the point where it was suddenly dramatically below standards.

Regular Assessments: if an employee performs below expectation, they know very quickly and have an opportunity to adjust.

42. No choice of reviewers — although there are a few exceptions (Sun), in most cases, unlike with 360 reviews, employees are not allowed input into who does their assessment.

Reviewers: nor should they.  Peers reviewing or having input should be those who are impacted by your work.  Under FairSetup, peers drive expectations.

43. One-way process — in most cases, employees also have no input into the factors that they are assessed on, how often they are assessed, and what type of feedback they can receive. It is unfortunately even rare for a process manager to routinely survey their users for suggestions on how to improve it.

Regular Clarifications: when expectations are missed, the first question is “why?”.  This causes an immediate clarification on inputs used for assessment.

44. No appeal process — employees who disagree with her appraisal are seldom given the opportunity to challenge the results with a neutral party.

Appeals: there are generally two three situations when a report and a manager disagree on assessment:

  • long-term evaluation – a long time has passed before an assessment is performed.  FairSetup allows to avoid these problems by ensuring regular assessments.
  • short-term evaluation – a report is evaluated in a short cycle.  Generally, because the consequence is low for both parties, they can agree to disagree and focus on adjusting to ensure disagreements do not happen in the future.
  • consistent short-term evaluation – a report is consistently evaluated in a manner that causes disagreement.  In this case perhaps the report and the manager are not a good fit.

45. Retention issues — the ultimate cost of an “unfair” assessment may be that it actually drives your top employees away because, for example, there was no differential in recognition and rewards for their superior performance.

Exactly. FairSetup increases retention by ensuring that all parties provide feedback into the system that they perceive it as fair.

46. Many possible emotional consequences — if performance appraisal is blotched, you can expect a decrease in employee engagement, trust, employer brand strength, teamwork, and innovation contribution. Employee referrals from disgruntled employees will probably also drop.

Emotions can not be avoided, but when they happen on a small scale, corrective measures are easy to take. 

Timing Issues

47. A time-consuming process — most of the forms are incredibly long and time-consuming. As a result, some managers routinely recycle “last year’s” evaluations. If HR is required to sit in on the sessions, the amount of wasted time increases significantly.

Quick: FairSetup is extremely fast. It collects just one retroactive measure (assessment) and one proactive measure (self-throttle) on a regular basis.  Fast and easy.

48. It is historical — the process is focused on capturing feedback about last year rather than on discussing necessary changes to job and skill requirements that must necessitated by the business strategy.

Forward-looking: FairSetup focuses on expectation, which are intrinsically forward-looking.

49. Not coordinated with business cycles – some appraisal dates do not coincide with the end of major business periods or seasons when all other business results are tabulated and reported.

Regular: the regular nature of FairSetup appraisals means that the system will collect information during both the high times when revenues are coming in, as well as during the quieter periods.

50. Not simultaneous — if appraisals are done on the employee’s anniversary date, the entire team will not be assessed at the same time.

Regular: once again, the regular nature of FairSetup addresses this in a systemic manner: by allowing ongoing collection of appraisal data.

Read the List on ERE.

2 thoughts on “ERE: 50 Top Problems with Performance Appraisal – the Most Dreaded HR Process

Leave a Reply to Dhanonjoy Paul Cancel reply

Your email address will not be published. Required fields are marked *