Building a Dashboard in 5 Easy Steps
Dashboards help users understand if a process or project is running smoothly. Typically, metrics are centered on some performance aspect, such as tactical objectives, time lines, or quality result. Keep in mind, the information does not present the next level of detail, which is what or why events are happening.
Dashboards be divided into two categories: Management and Operational. These categories are based on the information presented and the target audience.
- Operational dashboards are near real time focused and aimed at front-line staff. In Profiling Operational Dashboards, Eckerman notes two operational dashboard types: Detect and Respond and Incent and Motivate. The huge wall of monitors that NASA staff use during a space shuttle mission is a good example of a Detect and Respond dashboard. Alternately, call centers use Incent and Motivate dashboards, generally displayed on a wall, to encourage friendly competition or to allow managers know who needs assistance or training.
- Management dashboards are used by different levels of management and are based on the end goal. You can divide these dashboards into two broad categories: strategic and tactical. Strategic dashboards align with the organization’s strategic objectives while tactical dashboards address more localized areas of concern. These dashboards use historical data rather than real time data for trend analysis.
Strategic Dashboard Example
This dashboard is geared more toward executives. It lists the KPIs and provides some supporting information.
Five Steps to Building Useful Dashboards
Use these five steps as a guideline when you create your dashboard. In this example, we are creating a tactical dashboard for the SAS administrators and SAS business users. The end result was created in SAS Visual Analytics Designer.
Determine what the Users Wants to Measure
When starting a dashboard design, you might find it most useful to work with the users in a brainstorming session around a white board. This session might cover topics such as goals, pain points, and any existing metrics. Encourage the users to describe things that could cause issues or what success looks like. If you are talking to SAS administrators and business users, you might hear examples like the following:
- Can we identify memory usage on a server that is consistently swapping when accessing a report?
- How can we ensure that the system is being used and adopted?
- When is the best time to apply hot fixes?
What you learn is that users want to be proactive about the health of the environment and ensure that it can meet the organization’s goals. Probably before the session began you could have guessed that was a goal – but you may not know what it looked like to the users or how they think about it.
How do you Measure a SAS Environment?
By the end of the brainstorming session, you notice patterns beginning to form and you can identify measurements to support the organization’s goals. You might see three main concerns: users, data, and system performance. Granted it will be different for your situation, but I’m encouraging you to look for the larger areas of concern and simplify it. Too many areas of concern just turn into clutter.
The following table lists three major areas of concern and suggested measurements. There may be overlap in these categories, so you will need to determine which goal is the best match your needs. At the end of the discovery process, you should have a list of goals that allow you to find the data sources.
Here’s where the real work enters – it’s hard to select the metrics to measure. Don’t take this step too lightly – the business needs to consider the impact of what is measured and how it is communicated.
Create Sample Layouts and Build Prototypes
As the goal process continues, you can start thinking of the dashboard layout. Often dashboard designers find the layout of the indicators challenging. Using a tool like Microsoft Visio, you can play with indicator layout to consider how to group and display the data. Then you can present these ideas to the team for discussion.
Often when users see the indicators together, they start to realize what may not make sense. Some goals do not really belong on the dashboard. For instance, the administrators want to perform maintenance activities during non-active times. It is not really clear what goal we could use on the dashboard or how we position the measurement to add a conclusion. Perhaps the user activity really is information that belongs in a web report that allows the user to explore at will. In this case, the data is gathered but presented in a different report section.
All of the indicators from the SAS Visual Analytics Designer are box-shaped even if it is a round pie chart. In the following figure you can see a sample layout with suggested exploration paths. Remember the “01 area” is where people look first – so your most important indicator goes there. For layouts, you generally want the most important indicators larger while the lesser indicators are smaller. As you review the figure, you can see what elements are considered the most important, how indicators are linked and where the user would go for the supporting information.
Collect the Supporting Data
After determining the goals for the Admin dashboard, the data planning, collection and processing step begins. You should not discount goals where the data is hard to find. Instead, determine if the goal is stated correctly. It could be that the user needs to redefine the goal to better support the data. It’s also possible that your current environment doesn’t collect the information and you might want to consider alternate sources for the data, such as manually collecting data with a spreadsheet or a third party tool.
Once you have collected and formatted your data, you can load it into a tool. With the tables converted into SAS data sets, you can load them into SAS Visual Analytics in a variety of ways: Interactive load, autoload, local import, remote import, and data query. For ad-hoc data loads, the local import method might be suitable.
However as we want to ensure our data is loaded after the batch jobs are complete we use the Autoload facility to also accommodate reloading the table too. The Autoload data directory becomes the synchronization point for the source tables.Check the Administrator’s manual for more details about where to place your data to autoload into the tool.
Create the Final Dashboard
At this point, you should have your goals established, some dashboard layouts, and the associated data loaded into the system. The SAS Visual Analytics Designer is a great way to create dashboards and interactive reports. The tool has indicators, graphic objects, and interactivity that are useful in building a dashboard.
Working with Indicators
Indicators are used to represent metrics. It’s not always clear the best way to present the data and you have to try several methods. It’s easy with SAS Visual Analytics because you can quickly change the data from one visualization to another.
It helps when you understand the purpose of the metric. For instance, you could determine how many users there are and show how many unique users there was each week. However if you plot this information across a time line you really only understand the login trend. You learn that more users access the system at the beginning of the month and during the holiday shutdown, almost no one used the system. These are all facts, but not conclusions.
You need to understand the purpose of the metric. For example, many executives want access to the system but then rarely use it. Is it likely we are going to remove their login? Maybe the team wants to ensure user adoption. With lack of training, some users revert to storing data in Excel or something similar. The intent of the goal is to ensure 80% of the users are active each month. In this case, you would construct the goal as a percentage (active users this month/total users registered). You could also present the data with an indicator as shown in the following figure. Then you could link the indicator to the list of inactive users for follow-up.
Shepherd the User Adoption Process
As stated the dashboard design is an interactive process. One advantage of the SAS Visual Analytics Designer is that it is very easy to make changes. With the user present, you can create new categories and calculations, place new indicators and rearrange the layout quickly. With the user seeing the live results, it is easy to agree on a final layout and dashboard.
The Visual Analytics Designer also offers more interactive features that may be new to some users. Stephen Harris, who has implemented dashboards for large organizations, noted that user adoption is often an obstacle. He suggests that you work closely with new users during the transition phase by ensuring the user understands how the drop-down filters work and how to click to link to other report sections. He further suggests that you attend meetings where the dashboard is used to observe how the team interacts with it. “This is why your interaction is so vital. If you take the time, you will increase user adoption,” Harris said.
A dashboard is an investment. By working together, the two groups are able to monitor activity and resources and become more proactive in identifying issues quickly, recognizing patterns for potential end-user training and identifying opportunities for resource maintenance, effectively reducing an organizations business risk.
When SAS administrators understand how the tool actually provides value to the organization and the business users understand the background processes required to support the tool, they start to form a common language that makes other projects easier to implement.
Need Help Getting Started?
Zencos can assist your team with dashboards. We can help from the beginning to end by leading your team through the metric selection process to training you to use SAS Visual Analytics. Call us or contact us here.
Latest posts by Tricia Aanderud (see all)
- Use Network Analysis to Understand Your Customers with SAS Visual Analytics 8.2 - 2018-01-21
- My 7 Favorite Features in SAS Visual Analytics 8.2 on Viya - 2018-01-14
- Creating a Web Analytics Report in SAS Visual Analytics 8.1 - 2017-06-19
- Designing Dashboards: Sending Your Style Vibe - 2017-01-21
- SAS Visual Analytics: Design Versus Reality - 2016-10-05