Looking at the quantity of data collected by businesses each day it is impossible to even picture-making any sense of it without proper application of business intelligence. Businesses rely on information based on this data to tell them a story of how well they are doing or help them make a change. How to get the story across to people busy running businesses? The answer lies in BI dashboards.
Unlike reports containing copious information, BI dashboards are straight to the point. They are meant to be the easiest way to absorb information and the quickest way to paint a picture of whatever it is that one wants to know more about. This is only the case when they are built right, of course.
Choose the right tool
Unless you already have a reporting tool you use, you will need to select one. There are ways of building dashboard-like reports with unspecialized tools, but they will as a rule be time-consuming, quick to become obsolete, and with limited options.
Your reporting needs depend on many things including the industry you are in, the nature of your business, the size of your business, the size of the markets you are in, etc. Try considering them all before making the decision.
Even the reporting tools which may be considered leaders in their respect may prove not to work well with the way you collect data or be too pricey and exceed your reporting needs. It’s not always about what is best out there but what is best for your business.
One important feature of any BI tool is that it is easily accessible to end-users. Choose a web-based solution and, if possible, make sure that it is optimized for multiple devices (most are nowadays). Also, if it is not too much to ask, choose one with a mobile app.
Without easy access, you will be defeating the purpose of a dashboard which is empowering those making decisions about a business.
Get the requirements right
If you need to be concise, you must know how to convey a lot with as little as possible. This is exactly why before building a dashboard, you need to know its exact requirements. Speak to end-users of your dashboard and make sure you understand its purpose.
Every dashboard is meant to be telling a story by answering many questions. List those questions and they will tell you which metrics to use.
It is perfectly natural to revise a dashboard after users start using it. However, to save everyone’s time down the track, invest some extra time at the beginning stages. Speak to each of the users and understand their expectations.
Be the storyteller
Order of events
Every process has its beginning and its end. Even when there is no chronological order, there is still a logical sequence of metrics. Have that in mind and help the decision-making process by following that logic. Make sure to accentuate key points so they are easy to spot.
Set the context
You set the context based on the purpose, the users, and the frequency at which a dashboard will be used, and the frequency at which certain trends change. As an example, if you expect a dashboard to be visited daily or weekly, there is little purpose in presenting trends by calendar month or year. In this case, you would be better off by observing shorter periods giving perspective against the current state, and even at rolling periods such as the last 30 days, rather than the previous month.
Give context
Make sure your dashboard communicates the context of its visuals. This is as simple as giving descriptive titles and labels. Depending on a visual, a short title such as “Conversion Rates” could lead to a misunderstanding, or at least require some extra time for understanding what’s being represented in the visual. Is it a monthly conversion rate or a yearly conversion rate? Is it by sales rep, by market, by product, etc.? Let the users know in advance what they are looking at.
Use common language
When labeling visuals, make sure you use the jargon all users can understand. You cannot tell a story if you do not speak the same language. There is a difference between how a certain piece of information is recorded and labeled in your database versus what it represents in your visual. What’s more, it could be a simple matter of formatting that affects readability e.g. “date_app1” vs. “1st Appointment date”.
Less is more
Remember that a dashboard should be concise. Cluttered dashboards are hard to read and generally avoided by users. That is how dashboards become obsolete. Stay on track. Remember, this is not a report. Avoid a pitfall of getting into further analysis of each metric. Once a trend is recognized as worthy of additional attention, you can work on further analysis. In the meantime, all that it causes is clutter.
Visual Representation
Design
Speaking of clutter and the way your dashboard looks we’ve come to the point of its visual appeal. Apart from knowing that our mind dubs clutter as confusing (this being overcrowded dashboards or the use of too many colors), there is a set of principles outlining how the design of certain elements affects our perception, that is the way we see and understand things. These are called Gestalt principles and are widely used in the design.
Type of visuals to use
Not all charts and graphs are equally effective in showing different types of data. A dashboard based on financial data will likely be different to that representing operational information across different markets. Select the right type of visual depending on what it is you are trying to represent (proportions, trends, etc.) and how many dimensions there are.
Empower your end-users
Once you empower your end-users your dashboards can achieve their full potential. The users will be able to access dashboards anytime they need answers. However, to truly empower them would be to allow them to drill down data and ask questions. Leave room on your dashboard for filters. Some reporting tools have elements that allow you to ask questions about your data.
This type of setup reduces the need for clunky visuals and multiple similar visuals while answering any questions your users might have.
Finally, BI dashboards are an ultimate summary providing answers to key questions required for decision making. Once they stop being relevant, they become obsolete. Stay on top of your dashboards. Review the metrics in them and make sure to remove the ones which are no longer relevant and add new ones as they come up.
Create powerful data analysis solutions by adhering to good practices of building BI dashboards.
