Digitalization Post Covid Part 1: the Fashion Industry

NovelWang
3 min readMay 13, 2021

On one day in April 2020, I read the news that Neiman Marcus is filing for bankruptcy. It does not come as surprising at all, especially during covid. I may feel a bit nostalgic for some of the stores I have often been to, but overall I read that news with a sense of reassurance. I am all for the acceleration of fashion eCommerce.

Did I ever love Neiman Marcus? Yes, absolutely. They carry high-end clothes and are usually located in a good area (like Stanford shopping center at Palo Alto and fashion island at Newport beach), and they have really nice bathrooms, sometimes good restaurants nearby as well. But as for shopping, I am completely fine with searching and buying online. In the worst case when the size does not fit, I would just do a free return. The only thing that seems like a hassle to me today (living in the US), is that you need to print the return label and bring it to UPS for a return. Needless to say, fashion department stores would either digitalize themselves or go bankrupt.

I see three big opportunities in fashion eCommerce digitalization.

  1. Imitate online store UX features from Amazon. I started to use the online platform to buy clothes like Calvin Klein, Nike, and etc a few years ago. Of course, they have an eCommerce website or the app, but oh my god, the user experience is terrible. Either the login is extremely hard, or the check-out and payment have too much friction. All they need is a team of PM, UX designers, engineers, and data scientists to COPY amazon! It’s almost guaranteed that their sales would increase a lot via very small optimizations. However, nobody is doing that. Why? I will explain later in this article.
  2. Invent new AI technology to enable and stimulate the fitting room experience digitally. Imagine you can try out clothes and shoes at home with your cell phone? Why would you drive 20 mins to a shopping mall and then spend 10 mins finding parking only to do the same thing on that second-floor fitting room of bloomingdales?
  3. Recommendation. Fashion taste recommendation is not an easy ML problem to solve, but it’s something that must be done for fashion digitalization. Amazon is doing it, Alibaba is doing it, Stitchfix is doing it, so brand retailers will also have to do it, even though (sadly) they are not doing as much as of now yet.

Interestingly enough, these brand retailers are not doing much of these things at all. For years, the website is still hard to use, it’s hard to know which size would fit the best, and the store does not treat me like as I am an actual human being who has my own taste even though I had visited the Calvin Klein website a thousand times, not even to mention that they just do not have a size for an item that I want to buy, but does not even bother to give me the option to tell them about that important demand-and-supply information. But why is that? Why do these brand Ecommerce would rather leave money on the table than digitalizing themselves better through better UX, better AI technology, and better recommendation algorithm?

Organizational behavior seems to be the answer. Most such retailers like CK, Nike, and even Walmart are well-established old corporations that have a somewhat more bureaucratic organizational structure than tech companies, which makes them less agile, goal-oriented, and innovative. They are less likely to make disruptive changes that could mean short-term cost but long-term benefit. The leaderships are often too short-sighted to act aggressively towards a full revolution. After all, the CEOs are already paid a lot even though such a company may die in 10 years. A report from Accenture shows that more than 90% of CEOs would rather choose legacy between legacy and legend. Today we often think of Jeff Bezos and Elon Musk even when talking about great CEOs, but they are the very few legendary ones in the tech industry. When you look at all CEOs across all industries, most of them are not.

Given that, what does it mean for startups interested in this field?
B-to-B solutions to package AI technology seem to be a good way to go. One example I found while I was researching this topic is a very interesting startup Syte trying to build the AI piece for fashion eCommerce. They have modeled various shopping scenarios and devised solutions to each and packaged these into a platform that can be directly used by brand e-commerce.

--

--