“The “New” in Retail is Available Now Pt. 2”
How retailers actually benefit from Computer Vision and Machine Learning.
The fashion world is not only fascinating but also extremely attractive from a business standpoint, having caught the attention of many large companies. More recently, we have seen Amazon’s substantial effort to position itself more “effectively” in that industry.
In last week’s article (“The “New” in Retail is Available Now Pt. 1”) I wrote about the impact technology has had on the fashion industry. Companies such as Neiman Marcus, Walmart, and Macy’s have been invested in their eCommerce and IT Departments.
Today, millions are spent on rolling out products that try to please consumers, with different models, colors, and materials, being that in some cases, decisions are made from focus group results, which don’t always produce the expected financial results.
The problem is that when footwear manufacturers don’t have in house designers, they analyze information that is obtained from their own professional team, and use such information to decide what they believe will be trending in the following seasons. That comes with a risk that inventory may not be sold, albeit a reduced one if focus group research is done.
In reality, focus groups are used as a laboratory for these companies, with the intention of better understanding its consumers and consumption expectations. With that type of research, no one really knows what will sell, since the information on hand is what people are searching, and through social media behaviors, such as comments, shares, and pictures posted by users.
This technology is a “Game changer”
Many companies continue trying to understand trends and its customers’ behaviors through focus groups in order to target its products.
In that sense, a technology that has been gaining importance in the market is Machine Learning (a type of artificial intelligence (AI) that provides computers with the ability to learn without being explicitly programmed. Machine learning focuses on the development of computer programs that can change when exposed to new data.)
ShoesDsire, the startup I founded a couple of years ago, in addition to providing image recognition focused on shoes (basically, you snap a picture of any shoe, be it on a magazine, TV show or even from Internet, and the app brings results with similar shoes), has taken a step further in enhancing its technology with Machine Learning capabilities, adding more value to both users and retailers.
Machine learning, coupled with computer vision, can draw a very precise profile of its users, analyzing what is being posted, shared, commented at many different levels, including geographical location.
This technology brings information that in addition to reducing production risks as to which product will be a hit, will also lower inventory costs, guiding retailers as to the appropriate product distribution based on each region’s expectations.
By analyzing human behavior, Google and Facebook have become a reference as to the direction in which things are going, delivering exactly what its users want and like in terms of content.
The Boston Consulting Group recently examined how companies with advanced digital strategies are performing better relative to their peers. In that survey, 25 fashion retailers based in Europe and North America, companies that are leaders of digital implementation in stores, showed stronger EBIT (Earnings Before Interest and Taxes) growth than digital followers. (https://www.weforum.org/agenda/2015/05/how-digital-technologies-are-transforming-retailers/)
Even initially, being observed from a distance and with some mistrust, a few technologies were tested. But it was in the early 2000s that a certain technology was understood and implemented by retailers — RFID(1) tags to improve inventory accuracy. (http://losspreventionmedia.com/insider/loss-prevention-technology/the-roi-for-rfid-technology-in-retail/).
I still remember (back in 2003) when Walmart announced that it would require its suppliers to use RFID, which created positive results in a short period of time. RFID enables cycle counts to be completed about 25 times faster than traditional manual bar code scanning.
As I said before, companies that don’t assimilate and adopt new technologies will be left behind, and it seems there is no turning back, especially when the “New” in retail is available now.
(1)RFID (Radio Frequency Identification) was invented in 1948 by the scientist and inventor Harry Stockman. (http://www.nfcnearfieldcommunication.org/timeline.html)