Performance Dashboard Design
7. Conclusions and Recommendations
The study revealed that managers are becoming increasingly dependent on dashboards; however, not all organisations are able to develop them successfully. The tangible benefits of dashboards are evident, even when businesses are in the early stages of BI.
Unfortunately, there is no generic template for dashboards, as the development of successful dashboards can be a detailed and complex task. The literature review yielded best-practice guidelines for developing performance dashboards. The identified research gap was that there are minimal case studies on the successful use of dashboards in manufacturing firms that would help to determine the criteria for designing dashboards as a tool to achieve organisational strategic goals. This research focused on operational dashboards to ensure the communication of strategic goals on every organisational level. The research gap was closed by acquiring theoretical design criteria for dashboards, and integrating the criteria with best practices based on research conducted within a manufacturing firm.
The primary objective of the study was to explore the design of performance dashboards in relation to achieving organisational strategic goals, and to observe dashboards in an organisation in comparison with theoretical criteria for designing dashboards. The theoretical criteria for dashboard design focused on the areas of dashboard content, analysis, visual effects, functionality, and platforms. The secondary research objectives were to determine whether the dashboard platform could be Microsoft Excel or web-based; to assess the BI maturity, generation of dashboards and evolution of reporting in the organisation; and to assess the success of dashboards in an integrated and complex manufacturing firm.
Although many business units in the organisation still have dashboards in Microsoft Excel, they are evolving towards web-based programs. Platforms must be user-friendly and accurate; must ensure improved governance; must have historical, predictive, and statistical analysis capabilities; and must enable standardisation in the organisation. The research revealed that Microsoft Excel will always have a role, mostly as a tool that feeds into web-based dashboards. Additionally, in a highly complex environment, Microsoft Excel may be the most appropriate dashboard tool to prevent translation issues between Microsoft Excel and web-based platforms. The ability to access dashboards via mobile devices should be available; however, it must be managed correctly so as not to impact the work/life balance or disempower subordinates.
Using the BI maturity model as a tool to assess the maturity of the organisation, most business units were found to be between the 'infant' and 'teenager' stages. The organisation mostly uses simple monitoring tools, with some having progressed to using drill-down functionality. The organisation uses scorecards to communicate the strategy and dashboards to improve business performance on a more frequent basis.
The organisation has reaped the benefits of dashboards, which can be substantially quantified via energy efficiency, cost reduction, loss reduction, adherence to plans, and logistic cost savings. Dashboards were mentioned as a contributing factor to the record production achieved by the organisation in the previous financial year. The visibility of performance that is achieved by dashboards has improved efficiency in the organisation and in the way that concerns are managed before they escalate into a bigger problem.
The theoretical criteria for dashboard design focused on the areas of dashboard content, analysis, visual effects, functionality, and platforms. During the research, the additional criteria that emerged were:
- Monetary impact of poor performance
- Dashboard intelligence and fault trees
- Business culture for dashboards
- Dashboard maintenance
Dashboard content should be the key parameter and source of information that ties into the KPIs of the business unit and the organisation's strategy. Enhancing the dashboard by including the monetary losses associated with poor performance can be powerful, as users understand its impact and the urgent need to fix it. Information in dashboards must be accurate and reliable, and relevant; and this is achieved by ensuring that baselines and relationships are incorporated into the dashboard. The research results indicated that information in the form of day averages is adequate. When developing dashboards, collaboration with stakeholders is important. While the drive for dashboards can be from the top management downwards, the development of dashboards must occur in a bottom-up approach to ensure that all stakeholders agree on the detail, including the parameters to monitor, the definitions, and the ranges.
Dashboards must have drill-down capabilities, the extent of which must be agreed among the stakeholders when developing the dashboards. All drill-down layers must cascade up to the main page of the dashboard in one tool. Dashboards must have the ability to provide historical trends. A built-in predictive capability is crucial to allow the user to be proactive. To support predictability, statistical analysis capability is important. Fault trees and expert knowledge should be captured in dashboards so that they can provide preliminary diagnostics for problem areas. Dashboards must also create synergy between stakeholders so that they are empowered to optimise performance with the information that is presented.
Dashboards must be well-presented to receive the right attention. The dashboard must be a single page or screen. Simplicity and standardisation are important. The standard robot status - using red, amber, and green colours to represent the status of parameters - is ideal, along with globally understood measurement scales. Simple graphs with individual trends and tables are a preferred way of representing information. While the literature suggests the use of blinkers, flashers, and alarms, these were not seen during the research, and were mostly discouraged.
The study addressed the use of Microsoft Excel versus web-based platforms, as well as the accessibility of dashboards from mobile devices, as a secondary objective. To build on this, organisations are willing to pay for platforms that can serve their requirements. Ideally, users want dashboards to be fully automated; however, in complex and integrated organisations, users accept that in some circumstances information must be captured manually; and they can manage this accordingly. Organisations that are in the early stages of BI can develop their dashboards in Microsoft Excel, which can serve as a trial for establishing their requirements to develop successful dashboards.
Dashboards should be standard, and user customisation should be limited. Interaction with stakeholders or employees should be by means of traditional communication methods such as phone calls or personal messages. While online user interaction might have some benefits, it is more futuristic, and is not necessarily a priority at present. Similarly, the incorporation of process context, performance, and knowledge information in a dashboard is not mandatory at present.
The development of successful dashboards can be time-consuming, especially when behavioural adaptations are required. Exposure and assigning accountability helps to drive the correct behaviour. Behavioural changes may include upskilling and software training. Dashboards do not necessarily require executive sponsorship. Dashboards can emerge when management identifies the need to improve business performance and thus to introduce dashboards into their business.
Where manual inputs are required, stakeholders need to ensure that this is well-managed. Dashboards are living tools that must be adapted and updated as required. Revision control, management of change, and maintenance of dashboards must be included in the business process. Dashboards will require technical and IT maintenance, for which resources must be considered.
The theoretical criteria were consolidated with best practice guidelines to establish a tool that organisations can use to implement and review their dashboards. Figure 2 provides a diagrammatic representation of the guidelines.
Dashboards form part of a business process. Their development and implementation can be a detailed and complex task, and should not be undertaken in haste. The business process involves developing an understanding of the process and operating relationships, together with the creation of fault trees and decision trees. While dashboards can be driven from the top down, development must be from the bottom-up. Also, senior managers and area managers can initiate the use of dashboards without executive sponsorship.
While the research gap was closed, there is research on the practical application of dashboards in industry. Future research is recommended in other manufacturing firms, in organisations of advanced BI maturity, and in organisations of a more creative typology.
This research provides a consolidated set of mandatory and optional design criteria for operational dashboards that organisations can use when implementing or reviewing their dashboards. The approach is structured and simplified, providing a business with adequate background for implementing dashboards as a business enablement tool to achieve strategic goals.