How Digitization is Changing the Game of Sustainable Fashion (Part 2)

Artificial Intelligence (AI), Augmented Reality (AR), Machine Learning (ML) were once considered trendy, click-baity terms that graced the pages of articles across the board. However, in the blink of an eye, evidence of such technologies has permeated and redefined our everyday lives. From the moment we wake up, to the moment we hit the sack, we are constantly bombarded with ads, commercials, updates, sales, and so on.

Rampant consumerism has increasingly become a norm in our generation, thanks to the constant alerts and notifications taking over our devices, convincing us the need to buy the latest gadget or pair of shoes. The fashion industry realizes this trait and has taken advantage of our hunger for constant novelty, which has become the very basis of their business model. Most fashion brands create an entire season’s worth of clothing (which, by the way, are on a weekly basis) all at once, making calculated guesses about which items will be popular. They then sell as much as they can at full price, before discounting the rest or selling it to outlets. The remaining, unsold merchandise is thrown out to be incinerated or dumped in landfills.

The question then for us becomes, how can we utilize digitization in the fashion industry? Specifically, how can digitization help with making fashion clean up its act? Since AI has the potential to drive a seismic shift for the benefit of not only the fashion industry but for our environment as well. In Part 1 of this topic, we’ve talked about ways in which digitization could help drive greater sustainability in fashion through design, manufacturing and logistics. Here we’ll explore the role of digitization in retail and e-commerce, how it could change and enhance the experience of consumers.

AI & AR (AUGMENTED REALITY) IN RETAIL

Virtual Personal Stylists

AI has already found a home in customer support and experience roles, so it is only apt that they would be utilized in advisory roles in fashion, specifically as a virtual personal stylist. AI algorithms will be able to recommend the best items to customers based on their height, weight, shape, size, and preferences. Attributes like, types of occasion and weather could be even categorized and tailored for each customer. Customers are then given accurate sizing based on the details they have inputted into the system, enabling them to find those elusive perfect outfits that both fit well and suit their fashion preferences. On average 40% of online purchases are ultimately returned due to sizing inaccuracy. And what happens after a piece of garment is returned? They are simply destroyed or discarded instead of resold. The use of AI in customer styling and advisory is crucial as it allows for greater customer satisfaction and a reduction in the number of returns, reducing fashion waste as a result.

iLuk, an AI-powered personal stylist, comes in the form of a pod which will be placed at retail outlets. So how does it work? The customer will enter the pod, and gets photographed and scanned from different angles. The scanned images of the customer are then passed through custom 3D reconstruction algorithms, to develop a 3D avatar of the customers. The measurement data are then fed into the software to analyze the customer’s details, such as body shape, size, etc. Customers would then get personalized styling suggestions based on the details and fashion preferences.

Now, will human stylists be obsolete in the years to come? The answer remains to be seen, as the job of a human stylist consisted of understanding the customer as an individual, rather than just a momentary input of data sets. Thus for now, the best outcome would be for the role of human personal stylist to be complemented with AI algorithms. The styling and fitting process could be done with human stylists in the loop, such as picking the final suggested products gathered by AI virtual stylists, suggesting extra tips and tricks to the customer, or adding a final touch to the outfit with an accessory to give it a personal, human touch.

AR Virtual Try-on & Smart Mirrors

Augmented Reality (AR) has the potential to change a lot of aspects of the fashion industry. Essentially, AR is an interactive experience of a real-world environment, where the objects in our environment are enhanced with the integration of digital elements in real-time. The AR innovation that has the most potential to impact sustainability is virtual try-on of clothes and accessories, which has found a lot of traction in the fashion landscape. Brands such as Dior to Jacquemus have developed Instagram filters that let users try on sunglasses, bags, etc. Rebecca Minkoff even partnered with the app Zeekit, to allow customers to virtually try on clothing from anywhere in the world. Similarly, Zara has rolled out its own AR experience across 120 stores all over the world. Zara’s shoppers could hold up their smartphones to store windows to see AR models come to life on the screens of their devices, walking and posing around, showing how the garment looks like on a person and not just on mannequins.

(Left) Dior’s Instagram filter. (Right) H&M’s interactive AR app experience.

Likewise, Smart Mirrors which are powered by AI systems could also transform and simplify the shopping experience, with virtual visualisation of clothes and accessories. Consumers could see how the garments look like on their bodies without actually putting them on. The AI smart mirrors are complemented by RFID-enabled and Bluetooth chips attached clothing racks to relay information on the mirror’s touch screen interface, so that whatever article of clothing the customer brought into the fitting room will be shown on the mirror automatically. Customers could even look at different sizes and colour options, plus receiving personalized mix-and-match options to complete the outfit.

Smart mirrors at work at Rebecca Minkoff’s flagship store in NYC.

Rebecca Minkoff, ever the tech-forward brand, embraced the smart mirror at its flagship store in New York City. The store features a large interactive mirror which shows of the latest brand content, which the shoppers could also use to browse various other items and add them to their fitting room to try on. Smart mirrors installed in the fitting rooms also give shoppers a myriad of options, such as contacting a stylist, changing the lighting, getting product recommendations, and even sending the desired product directly from the fitting room to checkout to complete their shopping. Shoppers are also able to use the mirror to save the items that they have tried on during the visit to a personal profile, so that they can be accessed during future visits to the store.

In a similar fashion, Van Heusen created a retail environment with technology in mind, complete with a “Virtual Trial” mirror and a “Style Simulator” room. Clothes can be virtually previewed on a customer on the “Virtual Trial” mirror. As for the “Style Simulator” room, which sounds like something out of a science fiction novel, allows customers to try on outfits for the day, night, or evening wear, with even an option to turn on music that matches the occasion to create the right mood.

The most impactful aspect of AR virtual try-ons and smart mirrors are fewer returns. Consumers will be less inclined to buy pieces that don’t suit them, when they will be able to preview how it looks on them before purchasing the clothing, especially from online retailers. Less product returns result in less waste in landfills and carbon dioxide generated, thanks to AR technologies.

AI-backed Retail Inventory

AI can also be used for practicality and productivity in brick-and-mortar stores and online retail. Through recording sales, returns, and logging online purchases, retailers can keep track of stock and gauge which stores need which products. And visual perception based AI solutions also help store owners to keep a check on records of inventory levels and also categorize items in-store According to a survey by Capgemini, artificial intelligence could help retailers save $340 billion annually by the year 2022, by enabling efficiency in several processes and operations.

At H&M, they are combining both AI analytics with human intelligence to use what is dubbed as “Amplified Intelligence”. AI and advanced analytics have been used to improve their business overall, leading to a circular fashion system. H&M is improving the way they spot trends, plan logistics, and reduce the number of discounted sales, and masses of unsold stock by those technologies. Supply and demand are examined in advance, before allocating a sufficient number of products to each store, once again, reducing the number of wasted garments and resources.

AI and ML (machine learning) systems in retail are also providing an automated solution to monitor the customer’s activities while shopping, as well as visualize their sentiments to know what kind of products are popular and preferred and what products are overlooked. AI can also track footfalls in retail stores or record the shopping experience of the customers, to obtain feedback on how their experience was while shopping, so that retailers could improve their services.

AI & MACHINE LEARNING(ML) IN E-COMMERCE

Personalized Shopping Experience

Similarly, like retail stores, AI has been making its way into e-commerce and mobile apps. Fashion brands at large leverage the power of AI and ML algorithms to create a personalized shopping experience for customers, as driving conversions is their number one goal in this highly competitive industry. Those technologies collect real-time data such as location, clicks, searches, actions, purchase history, time spent on each page and so on, so that brands would be able to have a better understanding of their audiences. Oftentimes, the more a shopper interacts with on the app or website, the greater the accuracy of the personalized experience.

Through understanding and processing of those data, the shopping platform would be able to recommend similar products tailored to a shopper’s color preference, budget, and other attributes. Besides, stores can provide a tailored shopping experience by also automatically changing interface language and currency, offering time limits and discounts on potentially desired items, alerts on sales, and dynamic content that changes according to a specific shopper’s behavior; thus increasing the overall customer satisfaction and return rate.

Chatbots & Automated Customer Services

How many of you remember the good ol’ days of Clippy? Clippy was an early version of an AI chatbot in Microsoft Office, ostensibly designed to make writing easier, that is, till it was completely removed in 2007. To begin with, what are chatbots? A chatbot is simply a computer program that simulates conversation with human users to complete some sort of service. Sounds familiar? Well, Clippy’s legacy lives on in voice assistants such as Apple’s Siri, Google Home, Amazon’s Alexa, and Microsoft’s Cortana. Through such chatbots and conversational assistants, fashion brands could gather information by asking customers questions, understanding customer behaviors and desires, diving deeper into their shopping patterns, and suggested related add-on items to increase conversion. When a customer needs new shoes or a dress, instead of interacting with a website or mobile app, they can simply have a conversation with an intelligent conversational agent. Through back and forth dialog, the customer can find the optimal fashion product or accessory. This interaction provides a win-win situation where there is greater satisfaction for the customer, and much more valuable consumer information for the fashion brand.

A great example would be H&M’s interactive chatbot on the Kik messaging app. It begins by asking users to select photos of clothing that they like. Then it asks them to pick their personal style, such as Casual, Classic, Boho, Grunge, etc. With this information, the bot creates a comprehensive fashion profile for the user, consisting of items that they think the user might like. Users can also create their own outfits, browse and vote for outfits created by others on the bot for an interactive yet personalized shopping experience.

In addition to services mentioned above, AI-powered chatbots could revolutionize the customer support space. Today’s e-commerce platforms have to be available 24/7 and on multiple channels to boot. Automating customer support can save eCommerce businesses time, money, and operational resources. Besides, freeing staff from answering repetitive questions allows them to focus on more challenging requests. Round-the-clock chatbots can easily carry the burden of answering basic, routine queries.

This goes to show that chatbots are more than just rigid computer programs in shopping platforms. They’re a way to create enjoyable yet helpful shopping experiences for everyone. More specifically, chatbots can improve response rates, improve customer service, automate online purchases, and provide better communication; that is, on top of reducing operational costs, improving sales funnel , and encouraging interaction with shoppers. With the rise of messaging apps, users are more ready and willing than ever to shop online with bots.

Visual Search

In addition to conversational bots, brands are incorporating image-based search into their e-commerce platforms. Customers are now able to take pictures of clothing they like or styles they want to imitate, and smart image recognition systems can match the photos to real life items available for sale. On top of that, AI-enabled shopping apps allow customers to take screenshots of clothes they see online, identify shoppable apparels and accessories in that photo, and then find the same outfit and shop for similar styles. If customers were to search using a text query alone, the amount of suggestions or related items would be limited, and perhaps slightly inaccurate. Visual search algorithms are fed with a tremendous amount of data sets that contain the annotated images, enabling the content and context of these images to be understood and recognized, then a list of related results would be generated. Additionally, ML can be trained to recognize separate objects within a picture. This allows retailers to cross-sell items to consumers, giving customers the opportunity to complete an outfit that somewhat resembles the image query that they’ve uploaded.

OTHER POTENTIAL USES OF AI & ML

As discussed in Part 1 of the topic, the use of AI & ML in fashion is invariably correlated with the potential to better predict trends, and guiding human beings on the design process. While it is certainly true that AI can be very effective at providing such kinds of insights; helping designers choose fabrics, patterns, and colors that will resonate with their target audience, therefore making the life of a human designer easier, productive, and more efficient. The thing is, the immense potential of such technologies could have been maximized to address the bigger issues facing the fashion industry.

Eco-friendly Fashion

Environmentally-friendly fashion has seized the limelight in recent years, and for a good reason. Green consumerism is slowly gaining momentum among consumers, especially among the millennial age group (aged 22-35), as they are more likely than other generations to say that they would pay extra for eco-friendly or sustainable products. The incorporation of AI and blockchain into fashion production could help to reduce the environmental impact of fashion. Blockchain refers to a decentralized, distributed ledger that stores a complete list of transactions sequentially. In many cases, suppliers, retailers, and even manufacturers have no clue about the source of their goods or raw materials.

The lack of transparency and credibility is one of the reasons why environmentally-conscious consumers stay away from certain brands that don't disclose such information. The use of AI and blockchain, however, can rectify that; playing a part in creating a circular economy (also known as closed loop system) for every item. By revealing the entire lifecycle of a garment, from the source of raw materials, manufacturing techniques used, associated logistics, all the way up to finished products which are then distributed to retailers, businesses are empowered to create more efficient, sustainable, and transparent supply chains. The origin of each item can be tracked, brands could identify weak links in their operations, and substandard or faulty materials can be rooted out at the source.

The success of Depop, a peer-to-peer shopping app with 11 million users, has shown that there is an appetite for clothing rental and resale. AI can help to interpret the data and maximize the sustainability potential of this emerging space; thereby could bring about changes or addition to existing business models, which leads us to the next point.

AI informed ethics

One of the biggest challenges pervading the fashion industry is the innumerable counterfeit products flooding the market. The damage done to established and reputable brands are severe; loss of sales revenue and profit margin, reputation damage, are just a few to begin with. The cumulative effect of these devastating impacts poses a significant (if not fatal) risk to brands who are targeted by counterfeiters. At some point, spotting fake goods required the expertly trained eye of specialized customers or other enforcement officers. Now, AI systems could sift through databases of images online, and spot fakes that look increasingly similar to genuine products with machine learning algorithms. Blockchain could then solve the issue by tracing the counterfeit item from the retail and distribution stage back to its origin point of plagiarism.

AI also has the capability to avert brands away from the reputation damaging faux pas that has been making the news lately. Brands here and there have been accused of plagiarism, or worse, cultural appropriation and insensitivity in recent times. While the fashion industry is fundamentally rooted in individual uniqueness, creativity, and originality, it’s important to remind ourselves that nothing is truly new. Ideas and designs are constantly revised, recycled, and reimagined for a new generation. Still, where does inspiration stop and plagiarism begin?

The same abilities that AI and machine learning could do with analyzing and sorting images, could be used to review designers’ work alongside massive archives of designs and imagery, flagging similarities when identified. Designers would then be able to make revisions in advance to avoid inadvertent plagiarism or appropriation.

As consumer behavior and expectations change, the fashion industry is gradually catching onto the fact that the use of technology is endless and full of potential. The use of data analysis, animation, interactivity, and projections in fashion has proved the creative potential of the fusion of technology with fashion. Thus the alliance of fashion and technology speaks to the growing need for fashion brands to integrate digitization in order to capture consumer attention, especially in a market that is overrun with competitors who are constantly coming with new and exciting products, faster than ever before.

Throughout the course of helping fashion brands to address major industry challenges, AI will reshape the entire sector. While there is lots of cacophony about jobs being replaced by machines, the truth is that it will create and augment more roles than it eliminates. More value-enhancing and creative tasks will be given to employees, rather than routine, stagnant tasks of which machines could handle. Digitization could even evolve the business model that the current fashion industry employs, by empowering manufacturers and brands to redefine how they would engage, interact, and sell to their customers. We are now seeing that such technologies can add value in every segment of the fashion industry, from the design and manufacturing process, all the way to marketing and sales of the end product.

The limitless capabilities of AI, AR, and so on have yet to be fully explored, as even the IT sector itself is constantly evolving. Ultimately, those technologies will do just more than pushing the boundaries of what we understand fashion to be. The road to low-impact, sustainable fashion may be slow, but surely achievable through the deployment of digitization and AI in the fashion industry.

Like what you’ve just read? Share it with your friends, and subscribe to us here if you haven’t already!