There are two big determining factors: 1.) Is your analysis fixed, or are you interested in expanding your analysis over time? 2.) Is your analysis method sec

There are two big determining factors: 1.) Is your analysis fixed, or are you interested in expanding your analysis over time? 2.) Is your analysis method secret sauce? Additionally, your scalability needs are also a factor. If you plan on expanding or significantly modifying your analysis over time, the breadth of analytical tools available on the back-end is much larger. Working on the front-end is going to constrain your access to analytic libraries. You could of course do some analysis on the front end and some on the back-end but this is a hassle and why bother if you don't need to. If your analysis is secret sauce, and you are worried about someone ripping off your site, you definitely want it on the back-end. Scalability wise, moving computation to the front end is almost always a good idea. You'd be surprised at the numbers you can pull with well written javascript, and you can do even better if you can figure out a way to do your calculations using GLSL. In my opinion you're better off prototyping and getting a product started with calculations on the back-end then migrating work to the front-end as needed in most cases.