Frank Hobson Consulting

practical human resources support

Review of Staff Appraisal Systems

Appraisals - give your scheme an MOT

Introduction

Just about every organisation has an annual appraisal system. They come in a variety of forms; some are part of extensive performance management structures, some just feed into personal development plans, some focus on personal targets, others on competencies or values, some on all these. Most end with a single overall performance rating; frequently linking into pay in some way. Even if pay is not affected ratings easily become shorthand for promotability.

Appraisals are a foundation stone of the psychological contract and play a crucial part in operational effectiveness. How an individual thinks their performance is valued feeds directly into motivation and retention. If people collectively think that the appraisal ratings are not consistent across departments there can be an insidious under-mining of morale.  

So appraisal schemes are highly important - yet they are often poorly designed, under-maintained and rarely monitored. It is important that your scheme is delivering the messages you want it to, providing the appropriate level of discrimination and doing so as accurately as possible.  A complete renewal of a scheme can be time-consuming but that is no excuse for not monitoring it. If the scheme is not delivering it is better to know and to start planning its replacement. However, a complete rebuild may not be necessary and, below, I describe some of the techniques you can use to give your scheme an MOT.

The most visible element in your appraisal scheme is the overall rating that is delivered at the end. Staff know that this is what will be recorded and, they assume, become shorthand for their standing in the organisation. Some people argue for rating-free appraisals but, as with those no-one-must-win school sports, that misses the point. Even if there is no linkage to pay, the overall rating provides a bottom line to the performance balance sheet without which staff can easily draw the wrong conclusions. Obviously, which performance aspects are recorded, and how well they line up with day-to-day management processes, is fundamental to the success of any scheme but here I am just covering the rating element and how you can spruce that up.

What do the ratings mean and how many should you have?  

Ratings are usually supported by definitions and guidance notes. Some will never have been very clear. Others will just have become tired and over-familiar. Confusion often arises as to whether ratings are relative or absolute. If, say, 10 per cent of employees get the highest rating is that because you intended to highlight the top 10 per cent or is it that, coincidentally, only 10 per cent have performed at that level. In the first case the percentage will never change. In the second everyone could, theoretically, get that rating. The confusion can be compounded where performance is assessed against personal, annual targets. Does achieving all your personal targets earn a top box rating? Or are other factors important? This must reflect the way the organisation is managed.

If performance against annual targets is a significant factor the logical presumption is that everyone starts each year at satisfactory and has to re-earn their better ratings; just as this year's premiership winners will start next season with zero points. If you want your staff and managers to think this way you need to spell it out very clearly as managers resist moving staff back down the ratings unless they really go off the rails.  

Almost all schemes use absolutist terminology up to fully satisfactory performance. Thereafter, some carry on in this vein using titles such as exceeds, greatly exceeds, etc., whereas others change to relative terms such as above average, exceptional or outstanding for the higher ratings. Both approaches fall into disrepute if too many score in the highest category, so it is better to use terminology that reflects this reality. If everyone gets a top box where is the incentive to improve? If there is a performance link to pay this bunching is especially unhelpful.

How many rating levels should you have?

There is no right or wrong number - it depends on what you are measuring and what you are going to do with the scores. Too many levels can lead to arguments and false differences. Too few and you end up with no useful discrimination between individual performances.

Review the definitions

Once you have decided what you want your ratings to show, review your definitions and guidance notes. Managers do not read long definitions carefully after the first or second time and revert to using gut feel for the overall rating. In trying to cover all eventualities and be fair to everyone rating definitions can easily become overcomplicated. One technique is to identify the basic factors that are being assessed (achieving targets, team working, initiative etc., etc.) and then, from each box definition, pull out the phrases that cover that factor at each level. Lay them out in a matrix (a bit like some competency frameworks) and see whether there really is a clear and logical progression up the ratings for each factor. Clearer and simpler definitions lead to better compliance and consistency.

Once you have your definitions, stand back, read them out loud and ask yourself what impression they give. The two most common problems arise from definitions of standard performance and outstanding (or their equivalents). Most schemes have a category meant cover the majority of staff who are working at a fully effective level but all too often it becomes a pariah score if its definition is too dismissive and the one below reads as failure. Conversely, if you over-egg the top box definition you can make it sound like a job specification for the next Pope and managers will be frightened to use it.

Do you calibrate the scoring?

If staff believe that an appraisal scheme is not being fairly and consistently applied any pay policies it supports are undermined. Of course, some staff will always believe that other departments score more generously but you should, at least, try to ensure it is not true. Most organisations (hopefully all of them) check appraisal scores for equal opportunities risk.  But do you monitor scoring patterns across locations, departments, grades and service? Modern HRIS systems should make this easy. Why not provide annual feedback on scoring patterns? After all, most managers never see any marking but their own (but do make sure your feedback is meaningful and statistically valid).

Why not audit a cross-section of appraisals? Analyse the consistency of target setting and the linkage between detail comments and the final rating to provide confidential feedback to managers. For deeper insight you can evaluate, or rank, a structured sample of staff using some alternative method and compare that to the appraisal scores. Perhaps, talk a selection of managers through some of their scoring or get them to carry out a pair-comparison exercise.

In summary, performance labels provide one of the major messages to employees both about their own standing and about how employees generally are treated.  It is worth taking steps to ensure the message is valid.

     Frank Hobson

Back to list of articles