A new era of open source dashboards

Dashboards as designed by traditional softwares (such as Qlik or Spotfire) are good at providing visualizations and some rudimentary statistics to customers with no technical skills. The visualizations can be designed without much programming as long as you stay within the limits of what the software has made available to you per design.These softwares come with a price. A price in money obviously but also a price in terms of freedom. Going down the path of using such softwares means that you either accept the limitations inherent to these softwares or that you are ready to spend some significant money on IT projects to have the softwares been upgraded for some in house functionalities.


These licensed softwares evolve over time. Indeed, potential new functionalities for licensed softwares are evaluated against multiple criteria that ultimately will decide if they get implemented or not: cost, feasibility, overall customer interest, competitiveness, etc. Only the functionalities ranking high in the evaluation grid will prioritized for implementation and the others probably never will. It is perfectly legitimate and understandable for the software industry to work that way. But more and more the contrast with open source technology is vivid.In the open source world of R, and R shiny to be more specific, new functionalities are continuously released almost independently from the evaluation grid aforementioned. Indeed, releases are not driven by profitability but rather by passion and/or utility. One might think that this could have a detrimental impact on quality but with the source code of the functionalities being published publicly on Github, bugs can be reported and collectively addressed by the community. Most popular packages are consequently also most likely to be extensively reviewed by the community and see their potential bugs corrected. An obvious comparison can be made with Wikipedia where overall quality was proven to better than traditional encyclopedia thanks to the extensive review.


To the defense of the software industry, their product is more likely to be used without technical expertise and that might well be one of their last competitive argument to resist the wave of open source technology in that field. But it is noted that the amount of expertise needed to build R shiny visualization keeps decreasing over time with packages simplifying use of HTML, javascript, etc. Plus, the requested expertise can nowadays be acquired easily with the boom in the last decade of quality MOOC (Massive Open Online Courses) for R. Another difference with the software industry where courses, educational materials are sometimes proprietary and sold at a high price making it difficult to build a community of experts.


Regarding cost, the comparison seems relatively straightforward with open source and free technology on one side (R shiny) and expensive software on the other hand (Qlik & Spotfire). However, as discussed above the expertise required for open source technology might inflate the salary budget to attract talents with required expertise. Or expertise can be built internally thanks to MOOC or other internal training strategy. Some extra cost could also be induced by the need of ensure quality control at the level of the IT infrastructure which is less heavy when taking a software off the shelf.