Working In Uncertainty
A framework for accelerating management systems
First published 19 February 2004.
Designing management systems
Imagine you've set out to design a management system from scratch. You might start by writing down some rough ideas about how you would like it to be: effective, fast, responsive, proactive, adaptive, stable, low cost, simple. Perhaps you want to do something more than annual target setting and monthly reviews of plans against targets.
In recent years the Beyond Budgeting Round Table (BBRT) has claimed that organisations can be more ‘adaptive’ if they abandon budgetary control, and other management systems involving fixed absolute targets, and adopt different practices. The BBRT have reported many fascinating case studies of companies that have stopped using fixed targets. (The Beyond Budget ‘model’ is a collection of generalisations about the case studies expressed as principles, but these cover a wide range of practices.)
Separately I have written about ‘Dynamic Management’, which is defined as management in which the goals may change and be uncertain. Much of the material in ‘A new approach to management control: Dynamic Management’ is about techniques for accelerating management processes.
The BBRT case studies and Dynamic Management show that there are many alternatives to annually fixed targets and strongly suggest that accelerating management is an important part of the way forward.
The purpose of this paper is to propose a framework with which we can make sense of these examples and of alternative designs for management systems. The framework can be used to plan and track progress in the evolutionary development of management systems.
Five key dimensions
Imagine an organisation. It has people, structures, processes, knowledge, and so on. There are things that might happen in the future that will affect the organisation and for each it might be that some kind of management action might be advantageous for the organisation.
Forward view: The earlier management consider action in relation to a future event the greater the choice of alternative actions. (They have all the options that will be available later, plus those available only to the early bird.) It may be that the time it takes to implement certain potential actions means they are only available if an early decision to act is taken. Also, some chains of events are more easily influenced early on, before they have built up momentum.
Forward view is improved by (a) looking further ahead, (b) considering more of the set of possible futures, and (c) being better able to identify the most likely futures and focus attention on them. Most of the time we are not as good at predicting and controlling the future as we think. Our view is too narrow and should be wider. This is the main way of improving the forward view.
Scope: The greater the extent of reconsideration the greater the potential advantage. (Management have all the options that are available from limited changes, plus those available only to those who dig deeper.)
Scope is usually improved by moving along some continuum of deeper consideration. For example, one might increase scope by reconsidering goals up to a higher level in a hierarchy of goals, or by reviewing things you have previously considered more and more certain. Another possibility when looking at learning from performance indicators and experience is to progress up the following scale:
Numbers we report: What do we have to say to our shareholders, employees, the regulator etc? (And, how will they react and how should we explain it?)
Reality: What is the situation, really, and what is going on?
Understanding: Why is this happening? This is the first level at which modelling, perhaps informally, becomes important.
Prediction: What could happen in the future?
Action: What can we do that works, and how well does it work?
Input: The greater the amount of relevant information used the better the potential outcome of the deliberations.
Frequency: The more frequently management considers action the earlier actions can be considered, on average, and the higher the value of options. (Management can time their exercise of options more precisely, leaving decisions closer to the last possible moment.)
Speed: The faster the deliberations the less the delay between information being obtained and action resulting, and so the lower the risk of obsolete action.
These simple observations are the basis of this framework. The greater the Forward View, the more proactive the system is. The greater its Scope, Input, Frequency, and Speed the greater its responsiveness and adaptiveness. The other features of management systems can be understood as ways to increase these five.
In judging a management system against this framework we must recognise that there may be more than one cycle, and each cycle will have its own Frequency, Forward View, and so on. Some of the most important cycles (usually because of their good Forward View and Frequency) will be informal, making it hard to judge them.
Obviously, there's a problem. You could easily sacrifice quality of decision making for speed. Progress in Forward View, Scope, Input, Frequency, and Speed are only beneficial if they are not offset by a drop in decision quality.
Getting information, thinking about it, and communicating new plans all takes time to do effectively. The speed with which these can be done limits the Forward View, Scope, Input, Frequency, and Speed of a management system.
Beyond Budgeting cases and examples of Dynamic Management
The main characteristics of ‘Beyond Budgeting’ case studies can be understood in terms of their effect on information processing and the five dimensions:
Removal of fixed absolute targets: In a typical late 20th century management control system targets and resource allocations are set annually and rarely adjusted. In between target setting other things can change, but not the targets or resource allocations. If targets and resource allocations are allowed to change more frequently then the Scope for the monthly or quarterly deliberations is increased.
Relative and aspirational targets: Organisations that have done away with budgets often use targets expressed in relative terms compared with competitors, or aspirational targets (i.e. way above any realistic expectation for achievement). The effect of these is to split the concept of an ‘objective’ into two parts. One part shows that a particular performance characteristic is important and indicates the direction that is desirable e.g. more profit, less pollution, longer life expectancy. The other part is the specific value that is currently used for planning e.g. profit of £34m this year, pollution down to 100 tonnes this year, life expectancy up to 82 years by 5 years from now. Typically, companies have changed the first part very infrequently (not more than once a year) but the second part much more frequently (monthly or quarterly). This separation enables the specific planned objective levels to be changed more easily and often.
Increased Frequency: Along with this increased Frequency goes a reduction in the duration of deliberations. Each review takes less time because the issues are fresh in everyone's minds and changes tend to be fewer and less dramatic.
Not negotiating targets: Targets, if used, tend to be aspirational i.e. higher than any reasonable outcome. Also, reward is often on the basis of performance taking into consideration the conditions faced rather than by comparison with originally agreed targets. There is little or no negotiation of targets between levels of management and this dramatically speeds up the whole planning process.
Rolling forecasts: These improve the Forward View. The crucial difference between a rolling forecast and a budget in a budgetary control system is that the rolling forecast is a view of what may actually happen in the future, whereas the budget is a desire. The most useful rolling forecasts look at a range of outcomes and at the events that may cause them.
Faster accounting: In several examples companies have moved closer to real time accounting, which increases Speed.
Decentralisation: A few of the case studies include decentralisation. In the right circumstances it is possible to increase the total amount of management brain power available by distributing decision making. Also, if the alternative would be for decisions to be passed up the line for consideration then decentralised decision making can be faster.
Other methods of increasing management thinking throughput are discussed in Dynamic Management, and include:
ergonomics/information design, for example replacing a sea of figures with graphs that directly support the required thinking;
accelerated forms of risk and uncertainty management;
multi-level teams to integrate strategic and tactical reviews;
assembly of solutions from existing components, which is possible when planning is frequent and managers have developed a repertoire of potential action plan ideas;
legal terms that promote rethinking more frequently; and
open-ended architectures so that change can be made frequently as a regular part of business rather than by irregular upheaval.
This paper has presented a very simple way to look at management systems and think about how to design and improve them. The emphasis is on being quick and proactive. The various alternative management systems can be understood in terms of how well they perform on each of the five dimensions and how they achieve that performance.
Links and related reading
The Beyond Budgeting Round Table is at www.bbrt.org.
‘Beyond Budgeting: how managers can break free from the annual performance trap’ by Jeremy Hope and Robin Fraser is published by Harvard Business School Press.
‘A new approach to management control: Dynamic Management’ appears on www.dynamicmanagement.me.uk.
‘Risk management and Beyond Budgeting’ is on www.internalcontrolsdesign.co.uk, along with ‘Managing risk and uncertainty in Beyond Budgeting implementations’.
Made in England