Some Issues That Sometimes Happen When Using Data Analytics To Understand Fashion Trends and Brands

Unpacking some flaws of data science, and how they show up when analyzing fashion, too.

In the years I’ve spent using data and analytics to measure fashion trends and brands, it’s been an interesting challenge to apply something as numerical as data science to an industry that is incredibly artistic and subjective.

However, while data analytics is certainly a more objective, numbers-based space than fashion and design, there are actually many important subjective considerations that come into play. And, there is still a layer of human involvement necessary to make sense of all the numbers, and understand what the data is telling us in the first place.

Plus, as applications of technology—and specifically AI—become more widespread, it’s important to recognize how actually not-objective and biased data science can be. Or, as I’ve previously written, how computers are often unintelligent, too. And analyzing fashion trends and brands is no exception.

@databutmakeitfashion

DATA, BUT MAKE IT FASHION on Instagram: "some thoughts 💭 on th…

So, below are three flaws or areas of bias in fashion data science I frequently run into in my work. Not only to increase transparency into what the intersection between technology and fashion can look like, but also to highlight considerations that ensure the fashion-data is being used in the right way. Let’s get into it.