Performance Dashboard Design

Dashboards should provide managers with key information to monitor their areas of responsibility. Problems can be identified easily, and action can be taken to improve organisational performance.

The research discovered the following categories that encompass all common aspects relevant to the design of performance dashboards: dashboard content; dashboard data analysis; dashboard visual effects; dashboard functionality; and dashboard platforms.

In addition to these categories, Person argues that a facilitator or consultant is vital when creating a dashboard, to guide the organisation's leadership team and put forward difficult and probing questions.

2.5.1 Dashboard content

Rasmussen and Malik support the value of displaying critical key performance indicators (which are aligned with strategic objectives) in dashboards for effective decision-making.

Malik adds that personnel with decision-making authority require timely, relevant, accurate, and actionable information, in a form that is not static and overwhelming but rather dynamic and interactive. Malik also stresses the complexity of developing dashboards, mainly because they must be customised for each organisation. While Rasmussen argues that real-time information should be available in a dashboard, Stiffler argues forcefully that 'real-time' is a much overused 'buzzword'. His view is that data in most organisations does not change by the minute, and that the inconsistency between real-time reporting and infrequent objective setting reveals an obvious flaw in organisational reporting. Eckerson proposes the use of the term 'right time' as opposed to 'real time', meaning that the system should be able to deliver the right information to the right people at the right time to enable optimal business decisions.

Bremser and Wagner must be commended for including 'seek user input' in the dashboard development process, to discover user needs and get feedback on the dashboards implemented.

The studies of Groger, Hillmann, Hahn, Mitschang, and Westkamper discuss the importance of 'process context' (the context of a certain process step in the value chain), 'process performance' (information about technical and managerial performance), and 'process knowledge' (information for learning and continuous improvement) in dashboards. While Groger et al. propose principles that are likely to enhance efficiency and effectiveness in business performance, establishing such a system may be rather ambitious for organisations with a lower BI maturity.

2.5.2 Dashboard data analysis

Rasmussen, Malik, Bremser and Wagner, and Cokins endorse the use of analytical tools, such as heat maps, drill downs, statistical analysis, data mining, and predictive analysis, to empower users for further insight. Bremser and Wagner discuss predictive analysis, what-if analysis, and the incorporation of a statistical model. Few is the exception, contesting the use of dashboards for analysis, as this task should be left to the business analyst with the required expertise. Eckerson references a data management company that supports this view.

While the degree of analysis and the underlying details in a dashboard are debatable, a reasonable conclusion is that data analysis and predictive analysis should be included in dashboards, ensuring that they can be managed by the user and do not result in information overload for the user. The developer must also ensure that all users will be competent to use the analytical data provided.

2.5.3 Dashboard visual effects

Malik indicates that dashboards must be visually attractive, and the design must allow the user to synergise various pieces of information via a single screen. Eckerson, Rasmussen, Syrett, and Few agree that a dashboard should be one well-designed screen, enabling the user to get the critical information they need at a glance. Bremser and Wagner imply that there are user biases when viewing dashboards, and hence the developers must arrange elements correctly. They also endorse the use of graphs, charts, backgrounds, and colours in dashboards. Eckerson suggests that dashboards contain tabs and filters to manage information. Butler suggests that well-designed dashboards will be a combination of the sensible use of bar, line, and pie charts with numerical data tables. Dashboards must be programmed to signal critical concerns to users via alarms, blinkers, emails, sounds, traffic lights, dials, sliders, etc. to enable users to respond and act timeously. Colour coding is often used in dashboards, where red indicates poor performance, amber is borderline, and green is good.

2.5.4 Dashboard functionality

Malik proposes that dashboards have the functionality to be personalised for each user's requirements and situation. This subtly contradicts the dashboard benefit of ensuring a consistent view of the business that is common to all stakeholders. The degree of customisation allowed on a dashboard is therefore debatable.

Based on an analysis of the literature, Groger et al. emphasise that 'process communication' is a basic need for the development of a concrete dashboard service. Dashboard offerings must include an interactive method of sending messages, and of support between employees, as well as a way for leaders or managers to assign tasks to workers. Once again, these requirements may be ambitious, and the developer and stakeholders must evaluate their organisational maturity and culture before embarking on such dashboard functionalities.

2.5.5 Dashboard platforms

A popular platform for dashboards is an enterprise portal, which is generally a single location for accessing dashboards, documents, presentations, and other applications. Dashboards must be supported by an underlying software architecture that ensures that the correct information is available at the required frequency. Bremser and Wagner state that the software must have the capability to access the organisation's data warehouse for information. Some software may have higher costs; however, these costs may be justifiable. Platforms or software must be supported by the organisation's information technology (IT) policies and expertise. Malik adds that all leading dashboard software programs are web-based. Interestingly, Bremser and Wagner endorse Microsoft Excel as a dashboard tool, mainly because of its cost effectiveness and ease of use. They report that some companies pilot dashboards using Microsoft Excel, and then move on to more sophisticated software; however, the downside is having to customise program design elements that may be pre-packaged in specialised applications, as well as governance. Syrett and Person reaffirm the use of Microsoft Excel as an option for dashboards. Few argues that dashboards must often be available, even when one is not at a computer. Bremser and Wagner agree that developing dashboards for mobile devices may be attractive, but assert that this option may not be cost-effective, and do not strongly advocate the need for dashboards to be supported by mobile devices. Malik specifies that good dashboard software must have a fast response time; be intuitive, secure, scalable, and industry compliant; use open technology; and be supportable and cost-effective.