Driving a car is in a way quite similar to steering an organization: when you know how fast you drive (an indicator) you in fact know nothing (yet). You want to know when you will arrive at your destination; you will want to know whether you are driving in the right direction; you want to know if there is enough gas left, and so on. Steering an organization well is therefore only possible when the organization meets the conditions for effective management:
- there needs to be a goal (or destination) – without purpose and direction steering makes no sense at all except when we are driving just for the fun of it;
- we should possess information about the organization and its environment – without this information we cannot determine where we are which makes it rather impossible to determine the remaining part of the route;
- there should be enough time and capacity to process the information – when we drive very fast and another car suddenly appears from the right, we need sufficient time to respond: brake, swerve etc.)
- we should have a model of the system that we control – without knowledge of the car (or system) we do not know what handle to pull to move forward
- there should be sufficient possibilities and authority to intervene – to brake, throttle, steering left or right, horn).
The information should meet a variety of criteria
We also need information about the qualitative state of the system and its control measures. After all, when the dashboard indicates that the parking brake is defective, we know that we cannot use it when we find ourselves on a slope. In addition, the information should meet a variety of criteria to be as effective as possible. If the oil temperature indicator is hidden in the glove compartment, this makes little sense. Finally, we need a mechanism that does not just detect sudden degradations or improvements, but also detects successive little improvements and deteriorations.
Frogs jump straight from a pan containing boiling water, but they do nothing when water is cold at first and then gradually heated. In that case, they are boiled to death. This is a rather gruesome example to illustrate that people can react almost similarly: we exhibit a strong response to major or sudden changes while we show a temperate or lukewarm response towards a series of small(er) changes. Another phenomenon also comes into play here and could worsen the situation: once we embark on a certain course, we find it hard to take a step backwards on the basis of a few small or ‘soft’ instructions. We often need strong stimuli for that.
The context, which we must give to indicators to ensure effective steering, consists of:
The four perspectives of an organization (financial, customer, processes and learning/growth) provide a handhold for intensifying and enriching the information needs. Organizations typically choose to begin with shaping the information needs associated with the financial perspective. Although this information provides a picture of the actual course of events afterwards, it provides little predictive value for future performance.
Nonetheless, controllers often still believe that it is possible to create a predictable forecast of business results based on financial indicators. This should be discouraged: this method entirely misses the point because financial information is at the end of the process chain. The beginning of the process chain (e.g. how many orders are currently in the pipe line and what is their value) usually provides a much better indication on future (financial) performances.
Best is to start at “the front” and not just randomly take bits and pieces from all four perspectives. A performance indicator showing the employee satisfaction (from the learning/growth perspective) is not very useful to the organization when certain financial figures, such as turnover and costs, are not at hand. It is important though that the indicators from all four perspectives are eventually linked together to clearly see that financial performances are the actual result of satisfied customers, smoothly running business processes and inquisitive, knowledgeable employees.
Indicators should be supplied with relevant dimensions or angles. This concerns for instance dimensions such as products, customers, time (year, quarter, month) or sales channels. By linking indicators to dimensions, Business Intelligence system users can create all sorts of reports and analysis from different perspectives quite easily (OLAP).
Indicators are divided into two types and must be interconnected:
- Key Result Indicators (KRIs);
- Key Performance Indicators (KPIs).
KRIs relate to the goals and mission of an organization, such as market share or profitability. They do, however, not enlighten us on how to achieve those goals or in other words which performance driving powers underlie the performance and which strategy we need to get there. For example, a higher profitability can be achieved through realizing an increase of cross selling to customers, or by making better use of resources. The connections between indicators are of vital importance for an organization in order to be able to actually test and eventually to carry out its strategy. These connections are often not just made at once – due to a high level of complexity – and therefore organization end up dealing with this pragmatically and apply the relationships incrementally.
Figure 17: the management of a housing corporation has visualized the performance drivers on a dashboard.
After having determined an initial set of linkages, we’ll work with them for say a few months to a year. During this period, new connections may become visible which we will include in the Strategy Map and in the Business Intelligence system. “Strategies do not always need to be thought out in advance, they can also arise more or less naturally” (Mintzberg, 2004).
This method of working does however not fit very will with how we commonly think about strategy development. The prevailing thought is that strategies should be systematic, considered and thought out in advance: first think, then act. Ask any CEO of a company which, after a dynamic period of reorientation achieved success, what strategy they followed and he or she will invariably answer that the strategy had been thought out in advance. In reality though numerous ideas and alternatives were developed and swept aside during a process of trial and error. Crafted strategy development is about alternately thinking and doing, about trial and error: a Business Intelligence system and its organization in particular, should be aligned with this non-linear process. The system therefore is never complete providing that the organization stays in motion and its people remain motivated to achieving continuous performance improvement!
To actually be able to steer, we need clear standards (or norms) that indicate whether something is a ‘problem’ or not. Goals must be SMART . For example, in 80% of the cases, the waiting time should not exceed 48 hours. Or the repair time for standard repairs may never be longer than one day, whilst for repairs that require new parts to be ordered the repair time may not exceed five days. In the latter case, the relationship with the manufacturer is essential for achieving the set goals. In this specific example we see that by setting standards a key success factor that was initially ‘hidden’, namely securing clear agreements with the manufacturer, emerges. When we implement Business Intelligence, we will at some point go beyond the boundaries of the organization.
Standards and averages
Working with averages as a standard should be strongly discouraged because it might cause for outliers to cancel each other out. Instead, it is better – if possible – to work with a margin within which the indicator values should be covered at the lowest level of detail before moving on cumulating and laying down standards. Standardizing indicators using this method fits well with the principle of management by exception, whereby only the exceptions give rise for further analysis and intervention.
Information and knowledge should, at several levels, lead to (improvement) actions, either directly or after some time, otherwise the Business Intelligence system is rather useless. If the system does not lead to action, this might be due to it not containing the right information; is too difficult to work with; lacks good standards and so on. Other reasons may be lacking a performance oriented culture or managers and employees who do not possess the skills needed for achieving successful Business Intelligence.
That is precisely why we also design possible actions to be taken once an indicator goes beyond or below a certain level and exceeds the standard. This can be an additional incentive for professionalizing not only Business Intelligence, but also the entire organization. If possible, we begin with defining action there where the action should take place (at the lowest possible level), for example the turnover of an individual customer in the past week. When the turnover lags behind the revenue on other customers, something might be wrong. Possible actions then are: performing further analysis on customer data; contacting the customer; inquire with colleagues; making an offer in a letter, or just wait and do nothing.
It may be that this one less loyal customer is a harbinger of an exodus of customers. If that is the case, we also require actions at ‘higher’ levels and of a different calibre. Before we expand those actions to a wide spread group of customers, we first perform detailed analysis to find out what is actually going on and why. Possible actions at higher levels are: reducing prices; introducing new products; focussing on specific markets; advertizing campaigns to improve reputation, and so on. If we ‘stick’ at the lowest level too long, chances are we will fight a running battle and not survive.
When designing actions, this too should be taken into consideration: when do we switch levels? Practice shows that appointing various possible actions is a major differentiator to determine whether an organization is truly intelligent and capable of actually improving performance.
6. External information
Some indicators are strongly influenced by external factors. These indicators need to be brought together using external information. After all, we do not drive a car on the basis of dashboard information alone. The Board of Directors of an academic medical centre might for instance worry about a decrease in the number of students enrolling for a degree program in medicine and for that reason consult the ministry of education. Today, ‘decisions’ that are taken in the outside world can have enormous impact on organizations. Which external information is of importance exactly, largely depends on the environment in which the organization operates.
7. Qualitative and quantitative information about resources
Indicators do not spontaneously indicate whether production processes and production resources are in step with each other. Poor maintenance on machines for instance may result in the sudden failure of machinery, which in turn may cause substantial deterioration of performance. For an organization to ‘survive’ in the long term, both aspects of business – production and means of production – must be managed well. A fighter pilot, who wants to make sure he reaches his destination, will also regularly check the fuel gauge and the oil pressure gauge.
When indicators are surrounded by and enriched with a variety of contextual information – as shown above – a complete performance management model is created. This enables us to both display and manage all sorts of key connections within business operations. Consequently, we can strengthen or weaken these connections or linkages – through taking measures and actions – until an optimum is reached.
Organizations and corporations can use several performance management models (or control models) that already exist, such as the well-known Balanced Scorecard and the lesser known INK model and Six Sigma model. The first two models mainly work in accordance with a vertical business-driven approach, whereas the latter model primarily uses a horizontal business-driven approach and focusses mainly on the quality aspect within organizations argued from the perspective of customer needs. In principle the various performance management models aim for the same thing: an efficient and reliable control mechanism to align the performance of an organization with the goals of the organization so to increase and develop the quality of that organization.