Making Performance Appraisal Effective

If someone at work says, “You did a great job,” you’ll probably accept that comment without question (and think “I’m glad someone noticed!”).  However, if another person comes to you and says, “You messed that one up,” you may reject that comment (and think “how would you know?”).  The problem is that neither judgment – praise or censure – was supported with any valid information.

An ineffective performance appraisal can feel the same way.  Most organizations use some process to collect information to support an appraisal (employee self-assessment, feedback from clients and colleagues, managerial observations, etc).  Even when this information is valid, it is often not sufficient to deliver an effective appraisal, because other key elements are often missing.

It’s critical to recognize that virtually all appraisals are subjective.  Simply, when it comes to evaluating a person’s work – unless you can count it, weigh it on a scale, or measure it with a ruler – the appraisal process is always subjective.  This subjectivity often bothers people, as they want these judgments to be “fair.”  So, the question is “how can we make it more appraisal effective?”  Following are some steps you can follow.


1. Make sure goals are explicit, descriptive, and shared.

This step is often the point where appraisal starts to derail.  Unfortunately, most organizations do a pretty poor job setting expectations – a task that is not really all that difficult.  You can only set expectations about three things – actions, work product, or impact on organization goals.  The least useful and easiest to measure is actions; the most useful and difficult to identify is impact on results.  In my experience, it’s possible to identify the expected organization impact from virtually any job, but most managers don’t typically have the skill or will to do this – and they settle for actions or work product goals.


2. Make sure measures are explicit, specific, and shared.

Once expectations are clear, the next step is to agree on measures.  If you identify a link to organization results, then that measure is clear.  If not, then you must identify relevant performance measures.  Basically, there are four categories of performance measures (for either actions or work products) – Quality, Quantity, Time, and Cost.  The key here is to remember that these are categories, not actual measures, and each category has multiple measures.

-quantity (how much, many?)

-time (elapsed time, touch time, intervals)

-cost (design, production, acquisition, maintenance, TCO)

-quality (more than 50 including accuracy, completeness, validity, reliability, simplicity, etc.)


3. Collect relevant, valid performance information.

Once goals and measures have been established, the next step is to collect relevant, valid information.  Unfortunately, this step can be difficult, especially with jobs where is work product is intangible (information, influence, conflict resolution, etc.)  When there is a tangible work product, make sure to collect samples.  When there isn’t a product, direct observation is the best standard.  Finally, if nothing else is available, use self and other behavioral reports – but make sure to link those reports to expected goals and make sure to look for actions, and not just unsupported judgments (“he was great!?”)


4. Compare performance with goals.

Once you have the information, the next step is to analyze it.  I think of analysis as comparing actual with expected performance, for the purpose of understanding variance and its causes.  Sometimes the variance exceeds goals and sometimes it meant that the person fell short of goals – but your analysis is intended to identify what happened and to try and understand why (skill limits, roadblocks or conflicts that the person should address, and roadblocks and conflicts beyond the scope of that role).


5. Make your judgment.

Once the analysis is complete, the next step is to judge the performance – meet, miss, or exceed the goals.  Simply, your judgment should be evident from your analysis.


6. Describe both analysis and judgment so any reader can understand.

IF you’ve done the previous steps effectively, the actual written appraisal is fairly easy to produce.  When I was a corporate manager, I always used the following format – news, analysis, and commentary.  The first section was the news – a simple and descriptive report of work product and behavior, described in terms related to the agreed performance measures.  The second section (typically the longest) was the analysis – met/missed/exceeded and why, with descriptions of obstacles.  The final section (always the briefest) was the rating and judgment – this score for these reasons, with information always drawn from the analysis section.


My experience – as both manager and coach – is that if you develop your performance appraisals using these steps, then people will always understand the appraisal and see it as valid – even if they don’t agree with it.