A BIG thanks to @karthikvempy for volunteering to help coordinating a redesign project! He in fact has co-owned the project -- starting from helping me structure my ideas, to coming up with user profiles for interviewing and even ensuring things are in motion even when I'm not able to contribute.
When we redesigned the Reports section at Recruiterbox, one of the things we wanted to do was to make them look "nice", especially the charts. Until now, we've been mostly using the default colors provided by Google's charting library, which looks so BAD! Ok, "bad" is subjective, it at least doesn't go well with our color palette.
But how do you come up with so many colors that look nice on charts, while being consistent with your color palette?
When I started researching, the first thing I learned was that there are three different types of charts and there are corresponding ways of choosing color palettes:
Sequential and diverging palettes are used to represent values that vary continuously, from low to high or high to low. E.g. population density (sequential), altitude (diverging). However, these are mostly relevant when the values are represented on a map, and cannot be conveyed by something like the size of a bar in a bar chart.
Categorical palettes are the ones that are used to separate items into distinct groups. This is the kind of data we primarily represent in our reports. As such groups have no inherent order, categorical palettes typically contain colors of different hues.
So, these are the two criteria I derived for designing a categorical palette:
Alright, moving from the funda part, the next challenge was to come up with enough number of colors required for charts that satisfy the above criteria. I'm happy that I came across this article from a company called Viget, about how they solved exactly the same problem.
Following the same process, we came up with our new palette after multiple iterations between me and Sneha (our visual designer). And it was fun to watch people going crazy looking at the iterations.
For people who are curious about the iterations, here you go: https://www.dropbox.com/sh/zmz7zy1781zfbe1/AABwJL0wUmHQOKtNzBMsoiega?dl=0
So after all these tweakings, this is the new color palette we have come up with. I'm super happy about the result. We're yet to put it out on charts with real data and see how it works though. http://i2.wp.com/viky.in/blog/wp-content/uploads/2016/07/color-palette-all.png
Just dumping my thoughts from a recent discussion...
You have something that you want your users to do. It could be signing up for your product, giving you their contact number, getting more productive using your product, or clicking on a particular button inside your app. How do you make them do this?
I'd like to quote a favorite statesman of mine in the domain of public policy design: If you want people to do good, you need to make it easy to do good, and extremely hard to do bad things. The idea is that most people tend to do what is convenient for them at the moment, without objectively weighing the pros and cons. So if you make it convenient for them to do good, most of them will do it. You can deal with the outliers by imposing penalties.
Now coming back to the original question, UX design is through which you make the users do what is desired. Applying the public-policy funda to UX design, a good user experience is something which makes it easy for your users to do what is desired.
All the aspects of good UX design – being intuitive, less number of clicks, etc – can be derived from this.