Consistency at a price can also be achieved by forced distribution or ranking as described earlier.
Analysing and Assessing Performance 161 Even if the Flanagan concept of critical incidents has not survived as a specifi c assessment technique, it does provide the basis for evidence-based performance management – analysis and assessment processes that rely on factual evidence rather than opinion.
Visual methods of assessment
An alternative approach to rating is to use a visual method of assessment. This takes the form of an agreement between the manager and the individual on where the latter should be placed on a matrix or grid as illustrated in Figure 11.2, which was developed by Ann Cummins of Humanus Consultancy for a client in the fi nancial services sector. A ‘snapshot’ is thus pro- vided of the individual’s overall contribution, which is presented visually and as such provides a better basis for analysis and discussion than a mechanistic rating. The assessment of contri- bution refers both to outputs and to behaviours, attitudes and overall approach.
High achievement, but behaviours,
attitudes and approach need to
improve
High all-round performance
Not meeting requirements
Positive approach but poor level of
achievement Achievement
of role objectives
Behaviour, attitudes, overall approach to work
Figure 11.2 A performance matrix
The review guidelines accompanying the matrix are as follows:
You and your manager need to agree an overall assessment. This will be recorded in the summary page at the beginning of the review document. The aim is to get a balanced assessment of your contribution through the year. The assessment will take account of how you have performed against the responsibilities of your role as described in the role profi le; objectives achieved and competency development over the course of the year. The assessment will become relevant for pay increases in the future.
The grid on the annual performance review summary is meant to provide a visual snapshot of your overall contribution. This replaces a more conventional rating scale approach. It refl ects the fact that your contribution is determined not just by results, but also by your overall approach towards your work and how you behave towards colleagues and customers.
The evidence recorded in the performance review will be used to support where your manager places a mark on the grid.
Their assessment against the vertical axis will be based on an assessment of your performance against your objectives, performance standards described in your role profi le, and any other work achievements recorded in the review. Together these represent ‘outputs’.
The assessment against the horizontal axis will be based on an overall assessment of your performance against the competency-level defi nitions for the role.
Note that someone who is new in the role may be placed in one of the lower quadrants but this should be treated as an indication of development needs and not as a refl ection on the individual’s performance
A similar ‘matrix’ approach has been adopted in a fi nancial services company. It is used for management appraisals to illustrate their performance against peers. It is not an ‘appraisal rating’ – the purpose of the matrix is to help individuals focus on what they do well and also any areas for improvement.
Two dimensions – business performance and behaviour (management style) are reviewed on the matrix, as illustrated in Figure 11.3, to ensure a rounder discussion of overall contribution against the full role demands rather than a short-term focus on current results.
This is achieved by visual means – the individual is placed at the relevant position in the matrix by reference to the two dimensions. For example a strong people manager who is low on the deliverables would be placed somewhere in the top left-hand quadrant but the aim will be movement to a position in the top right-hand quadrant.
Analysing and Assessing Performance 163
Management style
Business performance
Low High High
Low
Figure 11.3 Performance matrix in fi nancial services company
A performance matrix used in Unilever is shown in Figure 11.4. This measures the ‘how’ of performance on the vertical axis and the ‘what’ on the horizontal axis. The matrix model also contains guidelines on the possible actions that can be taken for each assessment quadrant.
HOW
WHAT Inconsistent in
meeting agreed individual business targets
Possible actions:
• reward
• set milestones
• provide feedback
• training
• coach to improve delivery
Possible actions:
• recognize and reward
• challenge/stretch
• expose
• coach
Possible actions:
• set milestones
• provide feedback
• coach/monitor/track
• decision to continue or end employment
Possible actions:
• recognize and reward
• provide feedback
• mentor/coach to improve
• acknowledge contribution
Consistently meets agreed individual business targets Consistently
meeting expectations
Inconsistent in meeting expectations
Figure 11.4 Assessment and action matrix – Unilever
Overall analysis of performance
Those who do not believe in ratings argue that performance management is essentially about analysis rather than evaluation. The aim is to reach agreement about future action rather than
to produce a summarized and potentially superfi cial judgement. The aim of overall analysis is to reveal strengths and any areas for development or improvement. Managers are expected to reach an understanding with their team members as a result of the analysis that will ensure the latter will appreciate how well or not so well they are doing. The analysis should also identify the high fl yers and those who are failing to meet acceptable standards.
Businesses with performance or contribution-related pay schemes may disagree with this overall approach on the grounds that ratings are necessary to inform pay decisions. The major- ity (73 per cent) of the respondents to the e-reward 2004 contingent pay survey depended on performance ratings to indicate the size of an increase or whether there was to be an increase at all. Even those without such pay schemes like to follow the traditional path of summarizing performance by ratings ‘for the record’, although they are not always clear about what to do with the record.
An overall assessment may be recorded in a narrative consisting of a written summary of views about the level of performance achieved. This at least ensures that managers have to collect their thoughts together and put them down on paper. But different people will consider dif- ferent aspects of performance and there will be no consistency in the criteria used for assess- ment so it is therefore necessary to have a framework for the analysis. This could be provided on a ‘what’ and ‘how’ basis. The ‘what’ is the achievement of previously agreed objectives related to the headings on a role profi le. The ‘how’ is behaviour in relating to competency framework headings. The results for each ‘what’ and ‘how’ heading could be recorded follow- ing a joint analysis during a review meeting. This approach could usefully be based on a per- formance matrix assessment as described earlier. But it should always be remembered that the purpose is to generate information that will lead to planned development or performance improvement actions. It is these that must be recorded but this should include statements of why they are necessary and what they are expected to achieve.
Coens and Jenkins (2002) asserted that performance is largely driven by the system and ques- tioned the ability of anyone to distinguish adequately an individual’s performance from the situational constraints. They pointed out that: ‘In a given year some people perform better and some worse. A single individual’s performance may be better one year, worse the next and somewhere in the middle in the year following. These differences, however, may not be the result of some people trying harder or anything else signifi cant. It may just be the random hap- penstance of events and factors that impact individual performance.’ They therefore argue that: ‘Rather than think we can rank or effectively rate people we change our assumptions, we accept that while results are measurable, discrete differences in what can be directly attribut- able to individuals [are] not measurable. We may only be able to recognize, with a healthy scepticism, people who “stand out” in various settings, thereby warranting special attention.’
It is not necessary to agree with all these rather challenging statements, but there is something in the conclusion that we can recognize people at either extreme but cannot accurately distin- guish performance differences in the bulk of people lying between those extremes. Managers
Analysing and Assessing Performance 165 can in effect tell an individual that s/he has done exceptionally well and that s/he will therefore be included in the talent management programme, or they can inform another individual that s/he has not done very well and that they must discuss what needs to be done about it. The others can be told that they are doing a perfectly good job and discussions can take place on how they can build on their strengths or on any learning activity (preferably self-directed) that might help them to do even better.