Warning, Smart Data Discovery Tools Could Make You (Look) Dumb
January 26, 2015 Leave a comment
• The current crop of data discovery and visualization tools is getting smarter, requiring less analytical expertise to shepherd the business user quickly from question to insight.
• However, without some tutelage and guidance, advancements such as guided discovery and recommended visualizations could ultimately lead to less informed business decisions.
Earlier this month, TIBCO updated cloud-borne business intelligence (BI) software, adding Recommendations to TIBCO Spotfire Cloud. The goal is to make it easier and faster for a broader swath of users to dive for valuable business insights within the dark waters of big data. With the addition of a built-in analytics intelligence wizard (TIBCO’s words), Spotfire Cloud with Recommendations will now automatically suggest visualizations based upon the data selected for analysis. That’s a good thing, right?
It should be a good thing. After all, the goal of lightweight data visualization and discovery tools is to democratize data, making it accessible to as many users as possible. The trouble is, very few people know the difference between a box and whisker plot, heat map, and candlestick chart. Only those with some significant training in the data sciences would know which of these is the best way to make sense of a particular data set.
That’s why solutions like Tableau Desktop/Server, IBM Watson Analytics, Qlik Sense, and now TIBCO Spotfire Cloud have prioritized this problem, combining heuristics, machine learning, and best practices to literally handhold users through the process of selecting the most appropriate way to visualize a given data set or answer a particular question. Rather than wading through the complexities that lurk within the underlying data itself.
These solutions are making terrific inroads in not just identifying the right way to view data but whether or not that underlying data is itself correct, useful and appropriate to the task at hand. Products like IBM Watson Analytics can give users, for example, a dashboard indicating just how much trust they should place in the insights they’re about to uncover.
But what then? What if the data itself or the suggested visualization could potentially mislead the business user. That’s the missing step, the empty seat in an otherwise full auditorium within today’s concept of a data-savvy organization. Companies must not look at the intelligence that’s being built into these products as a reason to forego an investment in people, particularly those who can assist everyday business users, helping them avoid the dangers hidden within a false sense of security engendered by these smart visualization tools.