I recently had a chance to sit down with the guys from Blue Granite a leader in implementing performance management dashboards and had a chance to ask them their opinions on the world of dashboards, here’s what they had to say:
What is the future of dashboarding?
Dashboards will improve significantly on three fronts 1) user customization, 2) mobile access, 3) visual design
1. Customization – for the majority of users in the organization, dashboards are “what you see is what you get”, I imagine the next round of innovations from dashboard providers will include drag & drop “widgets” like you see in portals such as Yahoo, Microsoft Live, Google, ESPN, and others. In those sites users can create custom home pages for the weather in their city, news filters by subject or geography, sports teams, etc. Dashboards are somewhat flexible today but still require the knowledge of behind-the-scenes design tools to create custom pages. The flexibility for users to drag and drop from a menu of dashboard elements on-the-fly will become much more accessible and easy to use.
2. Mobile Access – Many business users who have a significant need dashboards (executives, managers) are not sitting at their desk all day, and are typically traveling or sitting in meetings. The ability to quickly check indicators and trends using a mobile device is greatly needed and will become mainstream soon. Some provides like Cognos are rapidly entering this area, and I imagine others such as Microsoft will catch up soon with their next wave of releases. Mobile devices should be able to easily sync on a scheduled basis with business intelligence systems, just as they do with email systems today. There are ways to get dashboards on mobile devices now, but the access is sometimes clunky and we should see some major improves in the coming 12-18 months.
3. Visual Design – Most dashboard platforms today are severely lacking in the design department. I’m in the Stephen Few camp, which is anti-gauges, anti-3D, anti-pie charts. Most dashboards today go for “show” instead of cognitive understanding of the data. The focus should be on the message, and what actions the message is trying to convey to the users. At one recent client meeting the CFO described 15 reports that are emailed to the entire management team on a daily and weekly basis. The reports provide just about every metric available. The trouble is the users are not understanding exactly what the reports are trying to tell them. Sales trends are down at this company right now, yet half of the managers walk the halls feeling upbeat and unaware of the issues. The CFO is perplexed and is looking for help. The answer is to move the IT team away from a “data first” to a “design first” mentality. Just because we throw some numbers into a report doesn’t mean the users will get it. Dashboard tools of the future will help business analysts by providing a much more advanced set of design tools and display options. Think of Visio stencils for dashboarding.
What is the difference between a personal dashboard and a corporate dashboard?
Dashboards function at every level in the organization, from corporate to department to function, and on down to the individual worker. The scope and function of the dashboard is determined by the objective(s) it supports. In the case of a corporate dashboard, the focus is often on business strategy and execution, which may include a balanced scorecard and/or strategy map that represent key objectives and the metrics that quantify performance in those areas. It applies to the all of the workers in the organization at a general level.
A personal dashboard is focused around the needs of a particular individual. It is likely more tactical or operational in nature, representing metrics associated to the specific responsibilities related to that worker, which ultimately roll up to broader department, unit, and corporate goals. For example, the personal dashboard for a warehouse manager might include metrics around shipping, error rates, worker safety and so on. These metrics are within the span of control of the warehouse manager, and they roll up to the broader strategic corporate objectives around productivity, quality, and safety.
What are some general tips you should think about when designing a dashboard?
When designing a dashboard, ensure that you have a good understanding of your audience, and create a design that meets their needs. Typically, a mix of tabular and graphical data will be ideal, but the needs of the audience will determine where the focus will lie. For example, an enterprise dashboard aimed at top executives may make heavy use of visual indicators such as stoplights (red/yellow/green) to communicate performance, whereas a department dashboard for the purchasing team may rely heavily on tabular data. Visual indicators support quick insight, while tabular data facilitates exploration of the data, two very separate needs.
Next, balance style and substance. Focus on the data first, and the aesthetics second. Flashy gauges and dials do not deliver insight, well-designed tables and graphs do. Create meaningful tables and graphs and augment them with color to highlight variances and trends and let the numbers do the talking.
Also, ensure that the metrics and design of your dashboard support the transition from insight to action. In other words, if a metric is off-target, in a 'warning' status, what should be done? Who should notified? Including metadata for each metric helps facilitate action. For example, include the name of the individual who owns the metric and provide a facility for annotation of metric performance. Another way to support the transition from insight to action is to provide a shared collaboration workspace for the dashboard. As performance issues are identified, the collaboration workspace provides capabilities such as task and issue lists, discussion threads, blogs, wikis, and document libraries to support the work activities associated with the metric/performance in question.
Finally, strive for a design that incorporates data dense visualizations. A primary challenge in designing a dashboard is finding enough screen real estate for the contextual data that provides insight into focus area of the dashboard. Examples of data dense visualizations include sparklines, invented by Edward Tufte, and bullet graphs, invented by Stephen Few. Sparklines are mini-charts that illustrate the trend of a particular metric over time. Bullet graphs are a good replacement for gauges, as they illustrate the actual, target, and threshold values for a particular metric at a point in time.
What are some of the challenges to be aware of when designing and implementing a dashboard?
A key challenge we have run across is making sure the dashboard is “actionable”. A dashboard can quickly become a novelty item unless serious consideration was given to business needs, mainly the desired activity outcomes (objectives) that the project sponsors hoped for when they funded the dashboard project. Our rule of thumb is that these activity outcomes should align to short and long-term strategic vision for the organization.
It is also important to create the dashboard with strategic, tactical, and operational layers in mind. A dashboard that might be actionable for one group of users may be worthless for another. When designing and implementing the dashboard, focus on meeting the needs of each user group and each objective. The more strategic, high-level users will want to see results across departments and teams that align to the CEO’s vision, while the day-to-day operational users will want to know if they are succeeding with their specific job function or team. A “dashboard” can be a small collection of screens that align to these various business needs and user groups in order to maximize value for the organization.
What’s the difference between a dashboard and a scorecard?
I like the definition of a dashboard put forth by Stephen Few which states, "A dashboard is a visual display of the most important information needed to achieve one or more objectives; consolidated and arranged on a single screen so the information can be monitored at a glance." A dashboard may contain tabular data, graphs, and other visual indicators to represent a snapshot of current business performance. It is a broad summary of performance in a particular business area. A common example is a sales dashboard, which often displays sales numbers from multiple perspectives, such as sales by product, geography, and customer.
A scorecard, on the other hand, is a specialized tool designed to communicate performance versus pre-determined targets, where performance is represented by important metrics or Key Performance Indicators (KPI's). A scorecard typically displays the KPI actual and target values, status, and trend. A scorecard is often embedded within a dashboard. Building on the sales dashboard example, a sales scorecard might include key metrics around dollars, units, market share, and profit.
Both dashboards and scorecards can be static, printed reports, or highly interactive end-user business intelligence applications. The defining feature of a scorecard is the use of KPI’s or key metrics, whereas a dashboard may include scorecards, tabular data, and graphs.
Where does a dashboard project fit in the overall scope for a customer’s vision for BI?
A dashboard is the centerpiece for BI systems within an organization. When thinking about home value, people thing about the kitchen and bathrooms. When thinking about the value of BI, people think about the dashboard. A dashboard view is the culmination of all of the hard work by the data teams and analyst teams to create a solution for the organization that can affect change, which is what BI is all about.
Given that, a dashboard within the underlying reporting, analytics, and planning tools is not going to fulfill the promise of BI’s potential. The dashboard should have a strong underlying set of tools that help users dig through the information, easily uncover “the answer” they are looking for, and put teams and decisions into motion. It’s best if the dashboard and underlying tools are one integrated package to give the business users the most efficient and productive environment possible.