Dashboard popularity is continuing to grow as organizations look for easier ways to access information within large data sets and to analyze that information in hopes of managing performance, making better decisions and lowering overall costs. However, the development and maintenance of these dashboards is not always easy. Organizations can choose charts, graphs, maps and almost any visualization imaginable to use within their dashboards. Unfortunately, more choice does not guarantee more effective use of visualizations, which may lead to companies being unable to identify the best way to build and monitor their dashboards.
In general, businesses require some guidelines to help them identify the essential components of a dashboard and how to develop their own dashboards to suit specific needs, all the while taking into account general rules that will help ensure the proper information is highlighted on the dashboard itself. This article discusses the information that should be shown within a dashboard to make it effective. This includes metrics identification and development, the types of visualizations to be considered, and overall dashboard management.
Identifying the dashboard’s purpose and general considerations
The one thing that is important to realize when building and designing a dashboard is that each organization will have its own business needs and corporate culture, meaning that what is effective for one organization may not work in another. However, companies can use guidelines that apply to overall dashboard use as a start when building initial dashboards or when looking to optimize current performance metrics. In either case, identifying the proper metrics, providing the right level of detail to end users, and making sure the information is timely are all aspects that should be considered irrespective of the type of dashboard being deployed.
Breaking down metrics
In addition to identifying the general purpose of a dashboard, looking at metrics is one of the first steps to building an effective dashboard. So the questions for organizations to start with is how to define effective metrics and how to identify what metrics are essential and which ones are just nice to have? Before defining what metrics to use, the definition of metrics should be identified. Metrics, or key performance indicators (KPIs), are basically the measures of success tied to various business drivers.
And to link goals to strategy, organizations should focus on specific KPIs instead of trying to measure everything that can be thought of. Therefore, companies should look at the top five to 10 items that are important to measure on a regular basis. Whether sales performance, sales pipelines, customer satisfaction, sales by region, employee compensation management or employee performance goals, etc., high-level metrics will provide a general view of what is important to monitor on a regular basis. The goal of any dashboard should not be to look at everything available, but to gain insights into what is happening within the business unit to make better decisions and anticipate issues that can be acted upon quickly.
Using the proper visualizations
Once metrics have been defined, organizations have to consider how to visualize the information. Overall, there is continual debate within the data visualization world about the proper types of visualizations that are most effective. Some state that simple spark lines and line graphs provide the best way to identify how organizations are performing against targets, while others prefer more visually pleasing ways of looking at information. Either way, the goal of visualizations is to make analyzing information easier, meaning that the way information is viewed will differ based on individual preferences.
The important thing to remember is the purpose of specific visualizations:
- Gauges and bullet charts (with targets identified) can be a good way of identifying how well organizations are performing against set targets because goals can be set and then the gauge dial will change based on performance - or the range market will move towards the target or threshold.
- Bar charts are a good way to identify overall sales based on products or services or to compare individual performance (whether individual, service or product based).
- Stacked bar charts and pie graphs let companies compare different measures such as products against one another. A pie graph can show percentages of sales in comparison with others or identify regional performance. In addition, drill-through capabilities let decision makers choose what information might be worth looking at more in depth. Also, some pie graphs can automatically group information together if it falls under a certain percentile to focus on the larger numbers.
- Heat maps can be used to identify general trends and product popularity as well as sales comparisons.
- Geographic maps can serve many purposes from identifying different gas prices in various regions, comparing or contrasting election results, looking at discrepancies in sales or student scoring, etc.
- Embedded reports are used in some dashboards to combine reporting and dashboards into a single interface.
- Charts and graphs provide organizations with a general view of sales and time-series type of information.
- And the list goes on…
One or more of these visualizations can be displayed within a dashboard. The more simply the information is displayed, the better - as it is important to balance the concept of analyzing information and overall data visibility with providing instant access to the most important and immediate metrics required for regular monitoring.
Many solutions providers offer dashboard templates that can be used to drag and drop visualizations into general dashboards with two or four blocks, etc. This idea provides the basis that organizations should use when designing their dashboards. Even though business users will be comparing multiple types of information in one centralized area, in reality, the less and more direct information viewed (that can be drilled through to gain deeper insights) is the way to go. Therefore, it may be best to limit dashboard to six key areas that encompass the 10 metrics, leaving room to connect to other visualizations, reports or data sources based on in-depth analyses that might be required.
Dashboard management and takeaways
Overall, with a focus on front-end interactions, dashboards have become the first line access for deeper analytics. Organizations should identify the types of metrics that are essential, limiting their scope to the top 10. Visualizations can then be combined or separated based on how the metrics are interrelated. Also, choosing appropriate ways to view information can be as important as the data being monitored, so it is important to match the types of visualizations with their purpose to get the most effective design.
In closing, even though dashboards expand the general usage of business intelligence, businesses cannot overlook the back-end requirements as well. Here are some general aspects organizations should think about when considering the dashboard infrastructure:
Data architecture and what the company already has in-house.
- Data latency and how often data updates are required. Although this may not affect the actual visualization components, it will affect the type of deployment and how often data is updated.
- Current BI, reporting and analytics that may be redundant or complementary to the dashboard initiative.
- The type of users who will be interacting with the information, as data visualizations and overall design and functionality will differ based on the skill set of end users.
- Future goals and anticipated uses of the dashboard.
About the Author
Lyndsay Wise is an industry analyst for business intelligence. For over seven years, she has assisted clients in business systems analysis, software selection and implementation of enterprise applications. Lyndsay is the channel expert for BI for the Mid-Market at B-eye-Network and conducts research of leading technologies, products and vendors in business intelligence, marketing performance management, master data management, and unstructured data. She can be reached at firstname.lastname@example.org. And please visit Lyndsay's blog at myblog.wiseanalytics.com.
(Copyright 2010 - Dashboard Insight - All rights reserved.)