Congratulations on your promotion with a window office! You move into your office and just stand looking out the window at a beautiful park. The flowers are in bloom and you are able to see the city skyscrapers in the distance. Wow!
A couple months go by and you realize that you have not looked out the window since that first day. So you take a moment to look out the window. The flowers are still in bloom and you still are able to see the city skyscrapers in the distance. You shrug and mumble, “Well so much for the window office, the view hasn’t changed at all. There is nothing new to see here.”
Ad-hoc analyses show views of the customer profile, product usage, predictive model performance, or customer proximity to stores. These views are critical to understand at a point in time. Like that first look out your office window.
But you do not refresh the ad-hoc views again until much time has passed or major changes in customer make-up or product offering occur. These views are critical for a foundational understanding of a customer base or category performance but what is missing is how to make this actionable on a repeatable basis.
You want to give your dashboard users something new to see. Avoid the “nothing new to see” reaction. The consumers of your dashboards need to see changes and the context to those changes if they are going to regularly view your dashboards as a management tool.
Here are some key tips to remember as you are creating dashboards for consumption:
1. Know your audience
Picture the primary user in your mind. Focus on what their needs are. Understand what their goals are. Build the dashboard to answer their questions but also help them answer the questions of their boss. Give them a reason to anxiously anticipate the updated dashboards by providing context to how what they are seeing compares to the past or to different audiences. Is the mix of products for this segment indexed higher or lower than the average customer?
Show how the information is changing and if that change is significant. Should they worry or celebrate the change?
2. Tell a story
Users of dashboards and reports get frustrated if all you are serving up are facts. They want a tour of their business and how it is performing. They want access to data that provides critical insights on how their customers are responding to marketing stimuli and a view into their entire business performance.
Part of knowing your audience is structuring your message in a way that they can quickly digest and tell others about. The information needs to be actionable.
Think about how you would take your dashboard and present it to a business leader. Is what you are displaying distracting from the story or enhancing it? Is there a better way to articulate what is happening?
A former boss of mine used honor the late, Paul Harvey, by asking “what is the rest of the story?” There is always more to the story. Sometimes we do not realize what it is or sometimes we do not think it is relevant. Challenge yourself to find what is not being displayed on your dashboard that could provide necessary context or background to the story your dashboard is telling. Maybe this context just needs to be written in a couple paragraphs of background about that product launch or campaign strategy.
3. Balance data with visuals
People process information differently. Some users just want data tables and others want more visual information. Create dashboards that address the needs of both of these cognitive styles.
You could create a summary bar of key stats like total sales, sales per customer, sales per visit, or average categories purchased with period over period changes. This resonates with the data person, but it also is a quick snapshot of current performance -good for the elevator speech to the boss. Graphs and data tables that best reflect the objective of the dashboard are also displayed. Avoid unnecessary visualization.
4. Allow for flexibility
Even if you are the best at understanding your audience and building your dashboard to tell the story, the user still needs some flexibility to explore different cuts of the data. Some cuts of the data could be time, customer segments, customer geography, region-store hierarchy, department-category-product hierarchy, and marketing campaign hierarchy.
Anticipate what the most likely cuts are but refrain from opening up the flood gates of allowing the user to pick from dozens of cuts and even more measures. Most consumers of your dashboards are not going to be a power user. Even if they were, they likely will want to build their own dashboards and not rely on you to build it for them.