Have you ever seen or built a dashboard that was deemed important for an organization but ultimately received little to no use from the users? Most of us – if not all of us – have been there. It can be frustrating and if you’ve ever been there, you have to be wondering what went wrong.
“Dashboard” has become a buzzword in the Business Intelligence (BI) community within the last decade. Recently, it seems that many companies want dashboards that look “cool” and make a big splash – without realizing that a dashboard is really another tool or method for delivering information to the users so that they can make well-informed decisions. The keyword in that sentence is “information,” and the delivery of accurate information should always be the underlying purpose of any BI / Data Warehouse (DW) project.
Don’t worry – there is hope. In this post, I will share what I believe are the five keys to building a successful dashboard that your customers will love.
Many people are creating dashboards that look great and serve the ultimate purpose of informing and educating their users. I recently completed a BI / DW project (using Agile) that involved creating new infrastructure and building a dashboard for a client in the greater Washington, D.C. area. The project was a success and we currently have about 800 users – that’s a lot! – who are actively using the tool.
Looking back, I can pinpoint several key factors that contributed to the success of this project.
1. Know Your Audience
It all starts with the end-users. Knowing your customers will help you understand what information (Key Performance Indicators or KPIs) matters the most to them. Keep in mind that the dashboard you are building intends to help make their lives easier by providing insight into information at the right level of detail in the right way.
So if you are building a dashboard for senior management or the C-level folks, then you might want to keep things pretty high-level (with the metrics that are of interest to them), whereas administrators would most likely appreciate high-level dashboards with drill-down capabilities so they can access the most detailed information.
2. Ensure Data Quality
Having good data quality is also important. After all, your customers intend to do something with the information you provide. If the data quality is not there, then both the product, and your team, can lose credibility quickly. I learned from my past experiences that the first impression is more important than people realize, and it is almost impossible to regain the users’ trust if your product gets a bad reputation for not being a reliable source of information.
To prevent this from happening, I recommend conducting rigorous testing on a regular basis (you will generally not have this problem if you are using Scrum) or setting up automated tests so that errors are identified and fixed in a timely manner before they get compounded into bigger problems.
3. Promote Frequent User Feedback
It is important to keep the users involved throughout the process. Scrum is a methodology that ensures early and often delivery to reduce risks since users are constantly testing the product and providing feedback as the product gets built. Late changes are generally expensive, so engaging your customers early in the process will eliminate surprises and help ensure a smooth product roll-out. Here are some tips for how to get started using one of the Agile methodologies for your next BI / DW project.
4. Make it Interactive
While delivering accurate information to the right audience is important, we must not ignore this other important aspect – user experience. An interactive dashboard with filters and parameters allows users to view different “slices” of the data in real-time, and providing instructions such as “click here to drill down into details” or “select the desired chart type” help make the dashboard more user-friendly and easier to navigate. Remember, if your customers do not like or know how to use your product, then your project is headed in the wrong direction.
5. Keep it Simple
Don’t try to fit too much information into the dashboard, it will only make it unreadable and can adversely affect the users’ impression of your product. Use a simple color scheme and charts that are suitable for the data. Don’t use certain charts just because they look cool – selecting the right charts to display your data in the right way is just as important as the data itself.
Lastly, there is no one-size-fits-all solution, so don’t expect one dashboard to work for the entire organization. Create separate dashboards that target different audiences and their unique needs.
Final Thoughts
Lastly, when building a dashboard for your users, always consider the following:
- Who am I building this dashboard for?
- What information do they need?
- Is the information I’m working with accurate? Is it a true reflection of the systems of record?
- Does the current software development framework allow for continuous user feedback? Who can I engage to gather feedback?
- What information filters or parameters should I include to provide users with additional ways to view the data?
Have a story from creating your own dashboard? Want to recommend your favorite tool for creating dashboards? Please share your BI / DW project experiences with us!