As companies are moving ahead with the transition to being data-driven, new BI tools and visual analytics are in the spotlight now more than ever.
Traditional BI is taking a back-seat as the new wave of self-service BI tools is empowering business users to ask and answer questions about their business without the heavy involvement of IT. This means it’s crucial for anyone analysing their data to be able to trust and understand the data in their BI tool (their visual analytics tool). And since we rely so heavily on visual presentation of our data , misrepresentation of that (data ) can lead to bad business decisions.
One particularly important question regarding the choice of tool is “Can you answer questions from your data?”
A typical use case for visualizing data is product profitability: “What are my Top 10 most profitable products? How about by product group? What about the least profitable?” Answering questions visually with a simple Top/bottom N analysis or using ( a second) multiple business views aka. dimensions should be a feature in every single visualization tool. Unfortunately this isn’t the case.
Building on the previous question, the next step could be to analyse the products that are at the other end of the profitability spectrum: “Where are these products being sold?” The best way to visually represent where your products are being sold is on a map. However, with mapping data there is always the pitfall of incorrect geocoding. Does your tool know the difference between Paris, France and Paris, Texas? I hope so.
“How long have these unprofitable products been unprofitable? Is this a reoccurring problem?” To answer this question, we need to be able to keep these values while excluding the rest. We also want to do this without having to create additional filters or reload the data, for the analysis to be truly self-service. For analysing whether our unprofitable products are under-performing from one year to another, (our ) the analytics tool has to have the ability to switch between date parts. Without these features we’re not able to answer our questions and therefore cannot make informed decisions.
One of my favourite data visualizations is the scatter plot. It’s a great chart for showing the relationship between sales and profit, and plotting lots of data points onto a canvas at once. Visualizing your product portfolio in this way is a visually compelling way of seeing all your products at once. However, what if your data visualization tool only showed you only a part of the data without you knowing that the first 1000 datapoints? That would lead to some seriously misconstrued answers. This is unfortunately not at all an uncommon ‘feature’, as it gives to appearance of quick load times, but at the expense of the user.
So more often than not, visualization tools are able to create compelling dashboards but miss the mark when additional analysis is required. And if you’ve been reading this and have realized that your current visual analytics tool is not up to par, then click the link below to download our free trial version of Tableau, and experience true visual analytics for yourself, without any of these crippling restrictions.
Stay tuned for information on our upcoming Tableau Breakfast Sessions in May!