Advancing AI in the ‘More-of-Less’ Era of Apparel Retailing
The complexity of consumer demand and the strain it puts on the end-to-end supply chain make profitability all but impossible without effective use of artificial intelligence.
Artificial intelligence (AI) and machine learning (ML) have become invaluable in the apparel industry for analyzing massive data sets, via algorithms, and culling meaningful insights from that data.
AI makes it possible for machines to learn from experience. Using deep learning and natural language processing (NLP), technologies using artificial intelligence can accomplish specific tasks by not only processing large amounts of data, but by also recognizing patterns in that data, and then using that knowledge to further refine outputs as more data is provided.
One of the key points about AI that makes it so incredibly valuable has to do with scale. It’s not that humans are unable to understand data or recognize patterns; it’s that they cannot approach the level of data collection, storage, parsing, understanding and evaluation that can be performed via AI. In practice, what that capability allows businesses to do is to increase the accuracy and speed of decision-making, with confidence.
In the past couple of years, AI has become one of the top technology investment priorities for retailers. When it comes to confidence in data, one of the problems that retailers suffer from is that even though they continue to invest in a great deal of technology, they often still do not understand the customer. “We don’t know what’s going on [with the customer], and so the technology doesn’t have that much of an effect,” says Brendan Witcher, vice president, principal analyst, digital business strategy, Forrester.
Consider the owners of a hypothetical mom-and-pop store in a small town. Each day, dozens of familiar faces file through the store. The owners have been proprietors of the store for decades, and they have come to know the purchasing habits and personal preferences of each of their customers. The owners greet their customers by name. They keep product on hand that they know their customers want, and will try out new products that they believe their customers would be interested in, based not only on past purchases, but on particular interests or beliefs they know those customers hold. This is knowledge accumulated over years of conversations and community connection. You can imagine that the possibilities for conversion that could occur in an environment like this would be virtually limitless.
Perhaps a customer who comes in to the store on a regular basis for coffee and a newspaper shares that he’s nervous about an upcoming week-long visit from his granddaughter. The store owner might suggest some local attractions, such as an amusement park, and then direct the customer’s attention to a new display of sun hats for kids. The next week, the customer returns with his granddaughter, bedecked in her new sun hat, and the owner shows them that a new bathing suit to match has just arrived. He also gives the granddaughter a piece of her favorite candy, which her grandfather had stocked up on during his previous visit.
Now, imagine a retailer with multiple brands, thousands of constantly changing SKUs and hundreds or even thousands of stores, spread over a broad geographical region. Getting a granular level of knowledge of the customer to be able to predict what they want and what will sell and where, across a vast enterprise, is virtually impossible for humans to do at that scale. On the back end, building a supply chain that can respond to that level of granularity is an equally, if not more difficult, endeavor.
In a recent presentation, Witcher identified AI as No. 4 in a list of top 10 tech investments that retail professionals are most likely to make this year, based on a Forrester survey and report. That’s telling in and of itself, but where AI stacks up on the list is almost a moot point, because AI is a key to unlocking greater value of the majority of other technologies that companies require to succeed in today’s competitive marketplace. Retail executives are starting to realize this.
“Two years ago, AI scared people. They were intimidated by it. But now there is an overwhelming realization that we have to harness AI to augment the entire apparel supply chain,” says Karin Bursa, executive vice president, Logility. “You cannot keep pace without AI. Marketing, innovation, everything is moving so fast. With AI, you can literally evaluate hundreds of scenarios rapidly. That opportunity is pushing companies to make a move [when it comes to investing in technology],” she says. In other words, the ability to harness AI throughout the supply chain has become a compelling reason to move their technology platforms forward.
“The number of mentions of AI on company earnings calls is growing exponentially, with executives talking about how they are harnessing AI to move faster, gain precision and gain new insights into their business — and not just in the supply chain, although that’s a big part of it.”
Why AI is more necessary than ever before
If we step back and think of apparel in general, the industry has been moving with increasing complexity for a number of years. There are more seasons per year, there are fewer numbers of more SKUs, there are more channels to buy and receive product and more trends coming from more media and other sources. “It gets to a point where it’s more of less,” says Bursa. “Smaller selling seasons, more volatility in the market. What that means from a supply chain perspective is more variability in the market.”
With more variability come more chances of making the wrong choices, whether in raw materials, design, assortments, pricing, inventory or identifying trends, to name a few of the points where decisions can go astray. This is where all of the data the retailers have accumulated, combined with data from outside the four walls of an organization, can prove invaluable.
One area where AI is seeing a lot of demand is in forecasting and demand of product — that age-old retail saw of determining the right product, right price and right place at the right time. The increasing variability possible within this framework has increased exponentially in today’s environment, says JoAnn Martin, vice president, industry strategy and market development, JDA Software.
Why is that? Omnichannel is one contributor. Today’s consumers are able to shop anytime and across multiple channels via digital devices, and can have their products delivered to just about any location. With more channels in the mix, retailer and brands must not only serve customers across a wider spectrum of places — their brick-and-mortar stores, their e-commerce sites, catalogues, third-party marketplaces such as Amazon, and sometimes across both wholesale and retail locations — they are often also creating separate product by channel. SKU counts are rising rapidly.
Personalization is another contributing factor. Interest in personalization — whether in product or marketing — has been on the rise since 2008, when the market dipped, says Martin. The reasons for that are manifold. Younger generations show preference for product customized to their individual tastes. But also, at a time when consumers were holding their wallets close, retailers were seeking new avenues to get them to spend. Nevertheless, at that time, the technology was not advanced enough to make a significant play in that area, she says. Today, AI and ML have made the near impossible a reality.
When you want to pull data from multiple sources — sales and customer data down to size and SKU, store and channel data, weather, news, trends, and so forth — things become “very complex, very quickly,” says Martin. In the past, companies would have made forecasts at an aggregate level, but AI allows them to marry internal and external data, and then to drill down to a very granular level of that combined data.
Here’s a specific example. Perhaps there’s a particular athletic shoe that a retailer has been selling in the U.S. midwest. The retailer has historical sales data for that region, but cannot extrapolate from that where else in the country the same shoe might achieve similar or greater popularity, says Martin. Today, on top of its customer profile data it could overlay fitness indices to see where those types of customers are most concentrated in other regions of the country.
When plugging this type of data into algorithms, AI also can spit out the type of detail that companies need to understand how to allocate and assort down to the size and style level, by store, so that product is delivered in consumer-centric assortments, and inventory is managed holistically across the entire enterprise. By sifting through millions of pieces of data, AI can begin to reveal insights at incredibly detailed and nuanced levels that would be impossible to capture otherwise: you may learn that, for example, across a five-year period, every blouse that sold at full price in the U.S. Southeast featured lace trim around the sleeves and fabric in a 60/40 cotton/poly blend.
AI can provide guidance in pricing, allowing companies to localize and maximize profits across channels, by indicating when to change price, when to mark down, and so forth. “There’s an opportunity cost to holding markdowns,” says Martin. “New product can’t get to the floor.” Being able to optimize inventory flow to maximize profits is incredibly valuable to retailers, she says.
Today, apparel companies need to be able to respond to swift changes in demand that originate from a variety of sources. Demand is dictating price in a more flexible way, she says, with data feeding in from a lot of different places. “It’s the ‘Uberization’ of pricing.” If there is a big football game, for example, it may be possible to increase prices on team apparel because everyone wants to show their fandom during the game. There are a lot of syndicated data sources that provide information on a wide range of factors. That data can be overlaid with a company’s internal data to fine-tune decision-making across the enterprise. Today, via AI, a host of factors can be taken into account with much greater ease than was previously possible.
Preparing your enterprise to take the AI plunge
While it’s true that AI is a self-learning tool and that the data AI pumps out improves over time, as it learns, that’s only true insofar as the data flowing in from the start is good and clean, and that there is an understanding of the components that are needed for the AI to produce meaningful results.
You’re probably familiar with psychologist Abraham Maslow’s hierarchy of needs, a pyramid structure whose base is comprised of basic needs (foods, shelter, clothing), with successive layers moving increasingly through more psychological and emotional needs before capping off at the tip with self-actualization.
Like building a house and starting with the foundation, just about any endeavor has a natural progression whose first steps must be completed before advancing to the next. This is true for AI as well and there is actually an “AI Hierarchy of Needs,” created by data scientist and AI advisor Monica Rogati, that can serve as a good model for understanding how to ready your organization to move logically toward effective use of AI within it.
The first step on the march toward AI is collecting data. Today, most apparel companies have this in spades, ranging from product lifecycle data to inventory data and sourcing to logistics data to unstructured data from social media sites. From there, companies must have way to flow and store data, and the third step is to transform that data by cleaning it and identifying any anomalies therein. Ascending, the fourth step on the pyramid requires aggregating data into segments and features that make sense for your business. In other words, what do you want to measure? What pieces of data would relate to that knowledge? What analysis will produce meaningful results? Only once data is clean and correctly identified can companies move to step five where experimentation can begin via A/B testing or algorithms to start to build knowledge that can then be used to begin to teach AI.
Ascending to the tip of the pyramid requires a unified data strategy that not only manages and integrates data across technology platforms such as enterprise resource planning (ERP), product lifecycle management (PLM), sourcing, supply chain and POS (point of sale), so that there is one master data source, but it requires that the data is the right shape and type, long before it gets pulled into any of your plans, says Bursa. “You have to be able to tap into details such as product attributes. “Why did this style sell better in this region? What did the actual selling size curve look like at full price?” To understand trends among product attributes (say, boot leg vs. straight leg) and pricing (full price vs. mark down), requires that data to be available and mapped into the system in a useable way. Many companies are not harnessing their rich data at this level, she says.
One of the main culprits preventing companies from getting to this stage comes from the absence of that master data source, and that can be attributed to the information siloes that still rein inside a majority of apparel businesses. The inability of that data to flow seamlessly in real-time across the supply chain is a huge hurdle. “Those gaps are creating latency in the decision process,” says Bursa. For example, in the product design stage, you may create a design, but have to wait two weeks before you get your sourcing bids back. You’re not giving that data to your supply chain team to looks at raw materials in the process. "We see these gaps that are three, four, 10 days long," she says.
There are opportunities to take those gaps out of the process and to gain visibility not just across the enterprise but across multi-enterprise operations that may be spread geographically around the world — to collapse those cycles and get business to move faster. That reduces risk because decisions can be deferred closer to market — often several months nearer when you eliminate all of that lost time.
The chicken and the egg of automation
Becoming AI-ready requires automating your processes. Automating your processes to allow AI to be more effective allows you to automate more processes.
There are several things to consider in automation. One has to do with the actual move from manual to automated processes. The other has to do with trusting the automation to function as it is set up to do. It sounds like a no-brainer, but letting go of tasks, especially for people who have often relied on their ‘gut’ to make decisions in planning, can be difficult. As with most new endeavors, the best thing to do is to take the first step, and apply AI to some area of your business. “You’ll pick up momentum. Your confidence level will increase. You’ll see how it works in one area of the supply chain, and from there you’ll seek opportunities to use AI to make better decisions elsewhere,” says Bursa.
Companies need to figure out how they can automate as many processes as possible, with humans stepping in on an exceptions-only basis. “With the current state of employment in the marketplace, we need to help people be more efficient and effective every day. We need to use that talent to do the hard stuff — the cognitive things where humans can add extraordinary value. Completing computations or looking at worksheets is not the best use of their time or talents. We need humans to deal with the outcomes of that data.”
Let’s look at the task of replenishment, which deals with moving product to the store that has already been selling at the store. Imagine the time gained when the process of figuring out how many of a certain item have sold, in which sizes, and reordering those items is taken off the table. Consider that to optimize this process requires sifting through a trough of variables that at a point surpass human capabilities — when is each individual store across a vast network likely to run out of a particular style in a particular color in a particular size? “Some retailers are now automating 85 percent to 90 percent of their replenishment, getting style-color-size right down to the peak days of the store — one retailer we spoke with found that Monday-Wednesday-Thursday was peak time for city locations and Friday-Saturday-Sunday in rural areas,” says Bursa. That level of granularity allows companies to optimize deliveries and also provides further insights.
The number of factors to consider in today’s omnichannel environment is exponentially larger than it was in the past. Consider men’s wear brand Haggar. “[Its] biggest challenge in omnichannel is that, when you’re in a traditional retail store, you’re going to put a smaller set of merchandise in the physical store because you’ve got to have all of your sizes available. You’re not going to offer as many options,” says Bursa. In e-commerce, that product portfolio is broader, serving a wider spectrum of customers, but Haggar may sell lower volumes across that spectrum. On e-commerce, instead of selling four styles of slacks, maybe it will sell 20 styles, and instead of three colors of khakis, five. You don’t have space to do that in a physical retail store, she says. Volumetrically, Haggar may sell more product through brick and mortar, but to be included in e-commerce channels, it needs a broad portfolio. Within that scenario, each store or e-comm channel has its own size profiles. “The average girth of a man varies from region to region. There is varying demographic data. Being able to see those differences in the data and provide that information to a design team is valuable,” says Bursa.
“When you augment human capabilities in these processes and let the AI use machine learning to tune parameters with each planning cycle — it may not sound sexy but the results are amazing,” says Bursa. The network is staying in step with everything you’re doing and keeping you in tune with what’s happening in the marketplace. It’s making small incremental changes that guide the business in the right direction.
And all of those small changes add up to big change in the end.
Jordan K. Speer is editor in chief of Apparel. She can be reached at [email protected].
Tech Exec Q&A
Apparel asked the technology community to weigh in on AI. Here's what they had to say.
JoAnn Martin, Vice President, Industry Strategy and Market Development, JDA Software
Apparel: What is the key area in which you see artificial intelligence (AI) changing the way that consumers behave, and what impact is this having on the way retailers respond to them?
Martin: Artificial intelligence is not changing the way consumers behave, but it is informing how retailers respond to them. Customers expect a personalized experience whenever and wherever they buy, and retailers are increasingly turning to AI and ML to satisfy these expectations.
For instance, retailers are turning to AI to optimize their forecasts and pricing. They’re leveraging predictive capabilities that come from the combination of historical sales data and external data signals, such as social media, news, events, and weather (SNEW). This level of analytical insight allows retailers to predict real-time demand and deliver a hyper-personalized experience to their customers.
Apparel: What do you see as the best opportunity for apparel businesses to use AI to improve the way they operate their supply chains?
Martin: Retail has a problem with e-commerce returns. Research shows that online purchases are returned at a higher rate compared to in-store purchases and this creates incremental supply chain costs. As e-commerce continues to dominate, returns are becoming more of an issue and AI can help.
Retailers can use AI to make sure they’re offering competitive pricing and markdowns to lower their return rate. While retailers must think about what price would make an item competitive in the marketplace, they also need to consider the price that would lower the return rate. AI can (and should) be used to help predict this.
Apparel: What's the one thing you'd like people to understand about your company's technology in the age of AI in apparel?
Martin: Apparel is a sector that is rife with uncertainty. The battle between brick-and-mortar channels and e-commerce means that there is no margin for error. Automatic replenishments, the correct pricing of products, seasonal trends and consumer trends are all drivers that can make the difference between making a profit or not. And AI makes these drivers very predictable.
The JDA family of Luminate solutions embraces leading AI and ML technologies and enables retailers to connect their supply chains and stores from end to end, delivering complete supply chain visibility and predictive and prescriptive recommendations that power more profitable business decisions.
Karin Bursa, Executive Vice President, Logility
A: What is the key area in which you see artificial intelligence (AI) changing the way that consumers behave, and what impact is this having on the way retailers respond to them?
Bursa: Consumers interact with AI-powered personal assistants such as those offered by Amazon and Google to receive answers and information at a moment’s notice. In retail, consumers now expect merchandise to be available just as quickly — when and where they want it. As the pace of product introductions increases, the number of seasons grows and the lengths of seasons shrink, retailers must turn to advanced technologies powered by AI and machine learning (ML) to keep pace. Through the use of AI, Logility helps to accelerate the product-concept-to-customer delivery cycle to ensure retailers and brand owners are able to satisfy the new demands of today’s tech-savvy consumers.
A: What do you see as the best opportunity for apparel businesses to use AI to improve the way they operate their supply chains?
Bursa: Many apparel and fashion supply chains are still very manual and labor intensive perpetuating long product design, sourcing and distribution lead times. And, for most brand owners, time equals risk. To accelerate time to market, current manual processes need to be automated through the use of AI enabled solutions. Early examples of automating product development, vendor selection and capacity reservation, and merchandizing and allocation have helped slash cycle times by as much as 50 percent. Automation and augmentation through AI empowered solutions speeds up processes and frees talent to work on more value-added activities that lead to further breakthroughs and a delighted customer.
A: What's the one thing you'd like people to understand about your company's technology in the age of AI in apparel?
Bursa: Logility Voyager Solutions leverage advanced AI and machine learning algorithms to transform the vast wealth of data available to help predict customer needs, identify trends, address disruptions and deliver a more synchronized and responsive supply chain from product concept to customer delivery. For example, Voyager Solutions utilizes AI to analyze future demand signals and actual performance to proactively react to changes. This innovative approach creates a more precise supply chain plan and increases forecast accuracy by 20 percent to 50 percent.
Stephen Sze, Director of Software Engineering, Alvanon
A: What is the key area in which you see artificial intelligence (AI) changing the way that consumers behave, and what impact is this having on the way retailers respond to them?
Sze: With the help of AI, consumers are sold a more personalized and unique experience. In apparel, they expect that algorithms can accurately determine their size and fit, or tell them what styles will flatter their body shape.
Retailers, in this omnichannel world, will have to cater to these new expectations through adopting technologies that can cater to these expectations while constantly finding fresh ideas to gain consumer retention.
A: What do you see as the best opportunity for apparel businesses to use AI to improve the way they operate their supply chains?
Sze: The low-hanging fruit lies in automation of certain operational processes. AI makes it possible to automate lots of manual tasks in the supply chain, which allows for shorter lead times and much faster response times. In today’s world, agility and flexibility is probably the most important aspect of any apparel business that wants to continue to grow and thrive.
A: What's the one thing you'd like people to understand about your company's technology in the age of AI in apparel?
Sze: Alvanon uses the latest AI technologies and Machine Learning techniques to develop our body technology. Our team of data scientists, AI/Machine learning technologists and developers work alongside our consulting group to develop innovative yet practical solutions for the evolving apparel supply chain. As a result, our body technology is currently used in more than 600 brands and their supply chains globally.
Nathan Pieri, Chief Product Officer, Amber Road
A: What is the key area in which you see artificial intelligence (AI) changing the way that consumers behave, and what impact is this having on the way retailers respond to them?
Pieri: Amber Road’s AI focus is on machine learning (ML) and ways we can help brand companies to optimize their supply chain further. As consumer signals relating to trends, pricing and material composition are received, our investment in ML is designed to help accelerate an effective response through sourcing, quality, production, and delivery.
A: What do you see as the best opportunity for apparel businesses to use AI to improve the way they operate their supply chains?
Pieri: We have many reference applications that demonstrate the power of ML for brand companies. The first is related to fast fashion, where based on the critical attributes of the product design you can more quickly source the right set of suppliers based on their competencies and quality metrics. Another area is in securing raw materials and projecting whether an inbound shipment is early or late. By analyzing historical delivery performance by type of product across your supply base, a model can be developed to accurately identify risky shipments and proactively mitigate.
A: What's the one thing you'd like people to understand about your company's technology in the age of AI in apparel?
Pieri: AI technologies such as ML require massive amounts of data to be effective. So first things first, you need to digitize your processes and create the "fuel" that can power ML algorithms. In this way, you get the initial benefit of process automation and follow up with another wave of benefits related to further ML-driven supply chain optimization.
Michele Salerno, Director of Marketing, Assistant Vice President, Celerant Technology Corp.
A: What is the key area in which you see artificial intelligence (AI) changing the way that consumers behave, and what impact is this having on the way retailers respond to them?
Salerno: Where we are seeing the most traction is through email marketing automation. With complete integration of your CRM (Customer Relationship Management) and sales database with your email marketing platform, your promotional marketing can automatically respond to your customers based on their prior purchases and/or preferences. To put this into perspective, a retailer can set up their promotional emails and set automated workflows based on specific behavior. Then, once their customers reach that behavior, they will be added into that segment and sent the specific promotional emails based on that specific behavior. For example, let’s say your customer purchases a toddler bicycle, either in your store or on your website. The next day an automated email can be sent thanking them for their purchase and suggesting related items that can complement the bike. Then, two weeks later an automated email can go out asking them to review the bike, linking to the specific product page for that bike on your website. Several months later, an automated email can go out in the spring offering a discount to come in and have the bike tuned up. Perhaps two years later an automated email can go out encouraging the customer to trade in their toddler bike for a youth bike. The beauty of the automation is that you, the retailer, did nothing. What sparked all of these actions was your customer simply purchasing that toddler bike initially; the software recognized that purchase, segmented the customer, and triggered the series of promotions based on that action.
A: What do you see as the best opportunity for apparel businesses to use AI to improve the way they operate their supply chains?
Salerno: One area that we are helping our retailers fulfill their orders is through automation. Our software enables retailers to utilize each store location as its own 'mini' warehouse, for optimized order fulfillment. The retailer can pre-set logic, in which the software can then automatically choose the location best suited to fulfill the order, based on a range of variables such as location, cost, quantity on-hand, style, and more. Retailers no longer need to have a person dedicated to sifting through online orders, manually printing them, searching for product availability and allocating items to shipping facilities. This entire process can now be done through automation, resulting in decreased labor expenses and fewer out-of-stock issues, with improved efficiency and faster fulfillment.
A: What's the one thing you'd like people to understand about your company's technology in the age of AI in apparel?
Salerno: At Celerant, our goal is to maximize our clients’ growth and efficiency through our innovative technology. We aim to not just keep up new technology trends, but in many cases- set the trends. As the industry and technology changes, we will be there every step of the way to help our clients benefit from the latest forms of technology.
Ketty Pillet, Vice President of Marketing, Gerber Technology
A: What is the key area in which you see artificial intelligence (AI) changing the way that consumers behave, and what impact is this having on the way retailers respond to them?
Pillet: As AI continues to grow and develop, brands and retailers are already able to leverage data to better understand consumer preferences including what they like, when they buy and even what they are willing to spend on certain items. As AI develops, retailers can use that data to develop the products they know the consumers will buy. With this information, companies are able to create a personalized experience for each of their consumers, suggesting the products they know the consumer will like and offering the right promotion at the right time. As AI develops, it will improve the relationship with consumers by better understanding their specific needs and offering a more intimate experience. Retailers who understand that data is key to offering a great customer experience will see an increase in brand loyalty.
A: What do you see as the best opportunity for apparel businesses to use AI to improve the way they operate their supply chains?
Pillet: In order to improve their supply chains, apparel businesses need to become proactive instead of reactive. Often times, apparel businesses produce a large inventory with only an educated guess on whether or not it will all end up in the consumers’ wardrobes. In order to reduce risk, companies need to leverage consumer data to gain better insight into what the consumer actually wants, allowing them to reduce inventory and risk. However, this also means they must produce styles quickly and deliver them to market at the right time.
For companies to create quickly they need to implement a process that allows them to seamlessly pass data from design to manufacturing. A fully integrated digital end-to-end process will allow apparel businesses to automate both their development process and their production by being able to pass data from design through production. An end-to-end solution helps reduce lead time while increasing profitability and efficiency.
A: What's the one thing you'd like people to understand about your company's technology in the age of AI in apparel?
Pillet: Data is crucial in today’s world. Gerber Technology’s end-to-end digitally integrated process helps to improve the global workflow from design to the final product by ensuring all data can go from one step to another, accurately and seamlessly. The unique end-to-end solution completely connects the entire supply chain through a full suite of software and hardware solutions that allow data to be passed seamlessly from design to print to cut, enabling brands and retailers to move quickly and efficiently.
Sean Elliott, Chief Technology Officer, HighJump
A: What is the key area in which you see artificial intelligence (AI) changing the way that consumers behave, and what impact is this having on the way retailers respond to them?
Elliott: Artificial Intelligence is providing brands and retailers the ability to tailor the whole commerce experience in ways that connect the consumer more intimately to the purchase. This includes voice commerce (Alexa), hyper personalization of offering, custom SKUs, and AI-powered customer service channels post purchase. The power of AI enables retailers to offer more differentiated offerings but also raises the expectation level of the consumer.
A: What do you see as the best opportunity for apparel businesses to use AI to improve the way they operate their supply chains?
Elliott: Supply chain operations contribute to the total value chain for a consumer, offering at a basic level exceptional service levels but also creating opportunities to differentiate experience through the fulfillment and reverse logistics processes. AI provides opportunities for supply chain to better leverage data in the decision making process, consuming AI driven predictions to drive task management within the warehouse, carrier selection and execution, or drop ship selection — driving more powerful consumer experiences. AI also drives automation opportunities, simplifying operational practices by automating routine decisions and directing facility functions on a more lights-out basis.
A: What's the one thing you'd like people to understand about your company's technology in the age of AI in apparel?
Elliott: HighJump’s solutions for supply chain management provide the industry’s best platform for the agile supply chain. Adoption of artificial intelligence benefits from a total capability to adjust business processes based on AI data and guidance. Customers can seamlessly manage change and adapt their supply chain over time to meet demands. As a result, HighJump empowers customers of all sizes to leverage modern technologies and optimize business process for today and beyond.
Will Yester, Senior Manager at Kalypso.
A: What is the key area in which you see artificial intelligence (AI) changing the way that consumers behave, and what impact is this having on the way retailers respond to them?
Yester: AI’s ability to detect patterns and then make accurate predictions will have a huge impact to how consumers shop. Subscription services like Stitch Fix use AI-driven algorithms that evolve based on customer feedback and returns, ensuring that future recommendations constantly improve and delight their members. As retailers improve their capability to translate omnichannel data into predictions that influence product decisions, AI has the potential to unlock new business models. Successful retailers might soon push recommended, customized apparel to microtargeted shoppers, continuing the push for convenience and a shifting away from the traditional retail model.
A: What do you see as the best opportunity for apparel businesses to use AI to improve the way they operate their supply chains?
Yester: Most retailers are not yet ready to compete in data-driven business models. The disrupters are data companies that sell apparel; their analytics backbone gives them the ability to service customers in new and exciting ways. Retailers will need to take three steps to contend: 1) Align on a predictive analytics strategy to define the goals behind the data and an investment approach, 2) Harness siloed data from across the enterprise and make it available for analytics through a shared platform, and 3) Start a program of incremental experimentation that starts to address big, ugly business problems and gain momentum.
A: What's the one thing you'd like people to understand about your company's technology in the age of AI in apparel?
Yester: As a professional services firm and strategic partner for our clients, our objective is to democratize AI and to help our clients understand real opportunities for value, understand how to make it possible, and build and adopt these capabilities on their own to achieve AI self-sufficiency. Advanced analytics and AI will be a critical capability for companies, so our role is to help our clients build these skill sets and tools rather than utilizing off-the-shelf black-boxes. We focus on the product side of the retail business, integrating data about the product across the lifecycle — including demand and use by consumers — to strengthen future product development for our clients.
Mark Burstein, President, NGC
A: What is the key area in which you see artificial intelligence (AI) changing the way that consumers behave, and what impact is this having on the way retailers respond to them?
Burstein: The future of retail depends on predicting each individual consumer’s impending purchases. Artificial Intelligence relies on big data, so as more consumer data becomes available, AI will allow more retailers to become hyper-accurate when marketing products to each consumer. Advertising provides a negative experience when the marketed products are not relevant to the individual. I do not have a cat or a baby in my household, so advertisements for cat food and diapers waste my time. An example of a positive experience occurred when I booked a ski trip to Colorado for my family, and the ski shop knew that I would arrive on a certain day, stay at a specific hotel, and that I did not own skis. After booking the trip, I immediately received an email stating “Mr. Burstein, we see that you arrive to Aspen on February 15 and plan to ski for seven days. We know that you like 178cm K2s and wear a size 29 Nordica boot. Please click here and we will have your equipment waiting for you at your hotel.” The retailers that offer positive and personalized experiences to individual consumers will thrive. The ones that don’t will cease to exist.
A: What do you see as the best opportunity for apparel businesses to use AI to improve the way they operate their supply chains?
Burstein: As AI, blockchain, and IoT develop and converge, predicting an individual’s future purchases will become much easier. I am looking forward to the day when I receive an email from Timberland stating, “Mr. Burstein, we see that you have worn your favorite hiking boots about four days a week over the past three years. You have walked almost 3,000 miles and we believe they are nearing end-of-life. We would like to replace them with our latest version, at a significant discount, if you return your current hiking boots to us so we can recycle them.” Supply chain planning would predict each consumer’s future purchases and aggregate that demand to a top-line forecast. AI would enable companies to quickly determine requirements for distribution, manufacturing, and product development, in that order. This is completely opposite of the current product lifecycle process. This model cannot scale without AI as there are just too many options and routes for manual analysis and decision making, regardless of the number of employees working in the supply chain area. With a cognitive digital platform supporting the business, supply chains will truly become demand-driven and will execute decisions in a much faster and more efficient manner.
A: What's the one thing you'd like people to understand about your company's technology in the age of AI in apparel?
Burstein: NGC is uniquely positioned as our Andromeda Cloud Platform integrates planning, execution and AI from concept to customer. No two retailers are exactly alike, so there is a real opportunity for each brand to craft and configure the solution to meet the unique value proposition they offer to their customers. The industry is just starting to understand the potential of AI in the supply chain. The future will be very exciting indeed!
Matthew Rhodus, Director and Industry Principal – Apparel, Footwear & Accessories, Oracle NetSuite
A: What is the key area in which you see artificial intelligence (AI) changing the way that consumers behave, and what impact is this having on the way retailers respond to them?
Rhodus: While AI holds tremendous promise in offering the consumer a host of benefits, most current use cases that have a strong business ROI are largely benign to the consumer. In many cases, the consumer might not be aware that AI is playing a role in the transaction. Whether that’s through chatbots, product recommendations, real-time offers to help conversion, reduce friction in the supply chain or guarantee on-time delivery. Currently, one of the strongest areas of consumer-facing use cases that could influence customer behavior would be in cross-device, channel and platform consistency in marketing to drive inspiration with shoppers.
A: What do you see as the best opportunity for apparel businesses to use AI to improve the way they operate their supply chains?
Rhodus: Brands are finding it harder and harder to grab the attention of shoppers and breakthrough the noise to reach their consumer. Influencer marketing has been one of the more successful tactics to grab the attention of newer shoppers and keep pace with changing consumer expectations. When multiple devices and social media platforms are leveraged to discover new influencers, brands and inspiration that drive purchase decisions, AI could help make sense of the tremendous amounts of unstructured data to ensure a brand resonate with the right audiences in a crowded marketplace.
A: What's the one thing you'd like people to understand about your company's technology in the age of AI in apparel?
Rhodus: AI is only as good as the foundation of information you give it access to. The technology has the ability to consolidate consumers’ basic information and to manage quick service requests, but without full access to a singular data source and transactional system, AI can’t help identify the next steps needed without further guidance or instruction. In most cases, human interaction is still required, if not highly preferred to address more complex inquiries or service requests. Oracle NetSuite provides that singular foundation for both data and execution, allowing AI to more effectively provide both information and direction on how to solve challenges for brands and their consumers.
Robert Fabrizio, Vice President of Business Intelligence and Analytics, Reflexis
A: What is the key area in which you see artificial intelligence (AI) changing the way that consumers behave, and what impact is this having on the way retailers respond to them?
Fabrizio: Consumers expect retailers to use AI to create more accuracy in all things – from product offering to service. Retailers can be proactive and do just that by more accurately forecasting store associate schedules, for example, reaping the benefits of increased sales and operational efficiencies, as optimal staff levels ensure consumers are attended to on the sales floor. Reflexis Systems influences the customer experience by streamlining corporate-store communications and labor planning using AI. This results in a great customer experience.
A: What do you see as the best opportunity for apparel businesses to use AI to improve the way they operate their supply chains?
Fabrizio: Using AI to determine best shipping routes given all external factors plays an important part in getting product to stores to satisfy customer expectations. The supply of goods along with forecast of shopper traffic and needed store staff together result in a better customer experience.
A: What's the one thing you'd like people to understand about your company's technology in the age of AI in apparel?
Fabrizio: Reflexis Systems’ artificial intelligence products rely on multiple mathematical models, historical data and information to complement our company’s history of industry acumen. With this complete understanding retailers can get the right associates in the right place doing the right tasks to make every shopper’s experience delightful.
Raj Badarinath, Vice President of Marketing & Ecosystems, RichRelevance
A: What is they key area in which you see artificial intelligence (AI) changing the way that consumers behave, and what impact is this having on the way retailers respond to them?
Badarinath: Shoppers have a split personality when it comes to AI. On the one hand, they expect retailers to cut through the noise of a crowded marketplace with relevant content and offers, remember their preferences from prior visits and purchases, and make apt recommendations while they browse online. As they move fluidly from mobile phone research to in-store evaluation to social media sharing, they expect brands to keep up and deliver consistently memorable experiences.
But on the other hand when asked about specific personalization techniques, consumers become more cagey.
Theoretically, at least, shoppers understand that such a rich shopping environment can only exist by processing data about their online and offline behaviors. More than half of U.S. consumers are willing to share data in order to receive a better shopping experience, according to a RichRelevance survey.
Just 32 percent of shoppers are somewhat positive about AI as a general concept, and fully 69 percent said using AI to select and automatically purchase items on their behalf would be “creepy.” Four out of five consumers believe companies should be obligated to disclose how AI is using their data — a sentiment reflected in recently-enacted privacy legislation in the U.S. and Europe.
To deliver relevant experiences that stay on the right side of the cool/creepy line, merchants must enable both the real-time intelligence they need and the transparency shoppers demand. Personalization platforms that rely on open and explainable AI can provide clear insight into how and why decisions are being made, and enable retailers to control data input and algorithm logic.
With that transparency as a foundation, merchants can employ AI to process large volumes of data in real-time to scale personalized experiences to every shopper everywhere they use a screen. AI can also enable companies to be more responsive to changing trends through better planning, merchandising, and inventory management.
A: What do you see as the best opportunity for apparel businesses to use AI to improve the way they operate their supply chains?
Badarinath: One key way that apparel retailers can use AI is to help master logistics and supply chain planning. For example, merchants can better understand regional and seasonal purchasing patterns and anticipate inventory demand. AI can also help predict which warehouses, distribution centers, or stores will be called on to deliver last-mile fulfillment and enable retailers to ramp up support as needed so they can meet customer expectations for swift, low-cost delivery.
Furthermore, integrated AI-driven experiences for consumers on the front end can result in better merchandising, a reduction in returns, helping drive down reverse logistics costs on the back end. What returns do occur can be optimized for efficiency, and shoppers returning items can be enticed with relevant alternatives to encourage additional purchases and continued engagement with brands.
A:What's the one thing you'd like people to understand about your company's technology in the age of AI in apparel?
Badarinath: Shopping for apparel online is particularly challenging. Shoppers are often looking for very specific attributes such as “soft,” “flowy” or “made in USA.” These aspects are critical to shoppers when looking for merchandise, but not readily available in the commerce experience given that most catalogs tend to be quite sparse.
RichRelevance’s natural language processing (NLP) technology solves this by identifying and extracting insights from unstructured text, product descriptions, user reviews, partner data, user-generated content, shopper activity and behavioral data to give merchants a more accurate picture of the purchase intent and interests of any given customer. This not only improves the customer experience by surfacing merchandise that is much more relevant, it also increases the number of items that shoppers purchase.
This exclusive NLP functionality is especially important for apparel brands because it delivers on implied as well as explicit needs in real time and it helps surface new items, highly-seasonal merchandise, and fast fashion that have accumulated little click-trail data yet — or never will, due to fast inventory turnover.
Rebecca Call, Solution Delivery Manager, VARGO®
A: What is the key area in which you see artificial intelligence (AI) changing the way that consumers behave, and what impact is this having on the way retailers respond to them?
Call: Next-generation retailers see AI as an opportunity to leverage the changing customer profile to result in conversions and customer loyalty. By reliably predicting customer demand and serving up relevant purchase opportunities in the right medium and at the right time, retailers begin relationships with new customers and maintain relationships with existing customers. Only by meeting or exceeding customer expectations in this Amazon Prime era, can retailers hope to maintain and grow those relationships.
A: What do you see as the best opportunity for apparel businesses to use AI to improve the way they operate their supply chains?
Call: AI in both of its forms — augmentation and automation — benefits retailers by predicting demand, stocking the right products in the right areas at the right time, supplementing labor during on- and off-peak times, and delivering orders in record-breaking time. From coupling AI with analysts who select and maintain appropriate product lines, to integrating AI and robotics into distribution networks, AI is seen as the next generation of retail fulfillment.
A: What's the one thing you'd like people to understand about your company's technology in the age of AI in apparel?
Call:AI allows retailers to keep up with ever-changing customer demands, but implementing technology into their supply chains does not need to be complicated. Some retailers incorporate individual AI components but fail to look at the impact to their entire supply chain. By selecting systems like Vargo’s COFE® Warehouse Execution System, apparel retailers can realize not only individual component efficiencies, but also integrate multiple technologies into a single layer of their operation. COFE® manages the myriad of robotics and AI components, human workers, and all other resources to work in concert — and in real time — to improve individual and overall efficiency, reduce cost and reliance on labor, and consistently exceed customer expectations.