Did you know that there are 12 factors to be considered while acquiring data from the web? If no, fret not! Download our free guide on web data acquisition to get started!
“Fast fashion is a contemporary term used by fashion retailers to express that designs move from catwalk quickly to capture current fashion trends.” – That is what Wikipedia tells us about fast fashion. However, today, reality is far from it. Although at first instance, you might think that companies must be hiring a truck-load of designers to capture the fast fashion market and sell more and more products to the evolving urban crowd, the truth is far from it. In fact, companies like Myntra are hiring more data scientists than designers to make the decisions and find the latest trends in the market.
Teams like the Rapid innovations lab at Myntra, have started to take business decisions related to fashion, using data. Whenever they have a fashion idea, with enough data to back it, and it looks like it has some potential, they run the data through some loops and iterations to see if they can spot a trend. Simply put, if your designer comes up to you, with ten designs and ten background colors for a new t-shirt line that you want your customers to buy this summer, then you can build an intelligent system that can actually cluster these designs and colors to tell you which one would sell the most. You can then put the design and color combination to test and see how the item sells. If it does sell well, then your algorithm’s decision was right and you can retrain it with this information (also called reinforcement learning). If it was wrong, you can retrain it with this information as well. With time, this real-time data will help you make more and more intelligent decisions.
Where there is data, there are attributes, and hence there is a possibility of creating some intelligence. The process of gathering this intelligence is what has changed with time. From spreadsheets to complex and complicated algorithms, and now we have algorithms that try to get better with time, with the support of data.
What began with exploring how one could automate the tedious process of spotting best-sellers in fashion and manually clustering them to determine their success, has ended up in this stage today. Even the definition of a “top-seller” is different today for these tech-fashion companies. You can no longer compare two products solely on the basis of the number of pieces sold. An item with ten thousand views and only fifty purchases is actually worse off than one with twenty views and ten sold. Who knows maybe the first one was being advertised on Google and hence many bought it? Also, the percentage of discount given, whether the item was bought during the sale, the number of returns and a lot of things actually go into declaring an item a best-seller. Declaring an item, a best-seller is no longer an easy task in the twenty-first-century fashion world.
And at this point in the fashion world, it is no longer a simple trend finding algorithm that would work. You need to find the color, material, stripe, color of the stripe, what that t-shirt goes with, and a huge number of attributes that are used. When a person is buying that one t-shirt that you have specifically designed to be the best-seller, he should also be suggested the ripped jeans that would go best with it and the bandana that would spring up the bling. Only if you can utilize collected data in this manner to capture the complete attention of the customer, will you be able to sell fast fashion in the competitive market.
You might go on to argue that it is a field of creativity and art. No-one is arguing that. In case of fashion and accessories, there’s a huge margin, because everyone knows that a big part of the stocks will be going unsold. However, as people become more price-conscious and the middle class is exposed to fast fashion, the one with the best prices will be the first person that people will flock to. So, to reduce your unsold-items and cut costs, you will have to make precise decisions when going to market. This way, you will be able to decrease your market prices, and more and more people will have access to your inventory, thereby making more customers happy. In the long run, it will benefit your company and will help make fashionable clothes accessible to all.
However, the human element cannot be removed completely. End decisions and tie-breakers still have to be taken care of by humans. For example, if there is a particular dress that goes with a particular kind of shoes, it is a fashion expert who would be better at making this decision. A balance has to be attained. And experts from the data realm as well as the fashion world would have to pool their brains to make the decisions.
You have to identify which problem needs a human touch and which one can rely on the data that it is backed by. Every problem needs a different approach. What works well today, gives excellent results, and increases your revenue might not work tomorrow. However, if you start integrating data and using data to support your decisions, you will at least have a fighting chance for the coming times.
Every season needs a new fashion line and no more do people ape whatever celebrities wear, blindly. Things become a “rave” all of a sudden and to know from before what a customer will like, based on his or her previous purchase history is something that everyone is trying to do today. It is somewhat similar to what Netflix does – keeping a track of what a user watches, what micro-genres he watches the most, and then tries to recommend him something that he will definitely watch. So, if you can do exactly that, track shopping data of different people, different age groups, and make calculations based on the many other attributes, and then suggest the perfect item, the customer will buy it, and you would have married data and fashion under the watchful eyes of your brand.
Your email address will not be published. Required fields are marked *
Save my name, email, and website in this browser for the next time I comment.