When data hurts design
An underexamined issue in the industry is that under certain circumstances, data can have harmful effects on design. Don't get me wrong, I am obsessed with data. At its best, data can reveal unrealised facts about user behaviours and drive better design decisions. However, at its worst, data can unintentionally lead to disastrous consequences for user experience. Without an easy way to measure emotional impact and user sentiment, these decisions have negative effect that often remains unnoticed.
Data, the design ally
At one of the previous companies I worked for, there was an application that was in severe need of a redesign. In the past, this application had not had any design thought bestowed onto it. It was ladened with bandaid features and developer-driven design decisions.
Unfortunately, redesigning this application right would not be a small undertaking, and this company was already strapped for developer bandwidth to work on company priorities. There was also a general impression amongst the product team that customers rarely interacted with this application.
So I turned to data. Not having the luxury of being able to run reports or look at analytics on this stuff, I sat down with one of the developers who painstakingly queried system logs to gather the raw data I needed.
I parsed this data through Refine, and finally I got the insights I was looking for.
So it turns out, customers were interacting with this application. There were users who active on there many times a day. These customers were also ones who were driving significant revenue for the company. How did they put up with this poor experience for so long?
Data had been my friend all along in this journey. I not only had concrete evidence that this design project could no longer be ignored, but I also had data on customer usage that would form the basis of how I prioritised goals and tasks for the interface redesign.
When data turns to the dark side
In 2009, Google's top designer left because of their obsessive testing culture. Google is infamous for testing 41 shades of blue to see which one performs the best.
...data eventually becomes a crutch for every decision, paralyzing the company and preventing it from making any daring design decisions.
This type of design, design by robots, devalues the designer's role. It breaks down what designers do as just the end result - just the form.
Design by SEO
The other instance of data being detrimental to design is when you design for robots. Design by SEO often results in sacrificing usability and the enjoyability of your users for the sake of increasing rankings on Google. Examples of this include having unnecessary copy that doesn't make much sense to humans, but instead is written for Google's crawlers in hopes that this will increase 'relevancy'.
A happy middle
Data should inform design, not drive it. Google's data-driven methods just test the symptoms, not the causes. These tests only reveal the what, not the why. In the pre-singularity era, design by robots is not the way. Human intuition and the understanding of context and motivations of humans, by humans, is still more powerful than the ability of machines.
The best product is one where both design and engineering play nice. Where the product is driven by logic and emotion. To use Apple as an example is bit cliché at this point, but they exemplify why you can't neglect engineering for design, or vice versa.
Apple is notorious for avoiding market research, preferring instead to design by intuition.
You can't just ask customers what they want then try to give that to them. By the time you get it built, they'll want something new.
At the end of the day, process depends on the project/product. What makes sense for some companies, may not work for others. Designing thoughtfully and being able to back up results with data will always need to be balanced with having the freedom to create work that you are proud of.