Fundamental Challenges in Mobile Data Visualization

1. Location. Location. Location.

We are not talking about ordinary real estate here. Mobile realestate is scarce. Finding the right medium on minimal real estate and designing visual components to tell the story is essentially the hardest part of this endeavour.

##2. Aligning the message Data is fine. Essentially Data visualization is 70% Data and 30% Visualization.  So you would think that if we have the data you would have probably solved 70% of the problem.  In some cases its true and I would agree. I have worked for medium to large enterprises and I see that the larger the enterprise gets, proportionally it gets more challenging to get quality data.  But what is most important is the messaging which goes along with the visualization needs to be aligned.

The picture above is a little fuzzy (but thanks to Stephen Few), the concept is clear - the message on the top is not really aligned with the graph below.  Now this could be because some additional information was missed or there was not enough space to explain this in a presentation (kind of goes back to point #1). But you would ask how is this problem really related to Mobile Data Visualization ?  The answer is - drawing on mobile devices ( iPads, iPhones, Androids and Blackberry OS’s) is challenging from a technical standpoint and what almost all vendors do is to start with the basic chart types which might not help in getting the message across.

##3. Just enough data Going mobile presents a challenge for visualaization in terms of securing data at all stages of the analytic process, primarily for data transmission and device access. For this reason many vendors are opting for a mobile product that provides functionality via server-side architecture with the use of a Web infrastructure like HTML5. Information then can be stored, managed, and provided using a mobile BI server, where the risk of exposing sensitive data is lower than storing it on a mobile device that could be stolen.  The other aspect of this is performance of the visualization tool itself.  If there are datapoints which are not relevant to the visualization medium then there is no point sending that data to the device itself.