《The AI Industry Series_ Top Retail AI Trends To Watch.pdf》由会员分享,可在线阅读,更多相关《The AI Industry Series_ Top Retail AI Trends To Watch.pdf(40页珍藏版)》请在得力文库 - 分享文档赚钱的网站上搜索。
1、IAI INDUSTRY SERIES Top Retail AI Trends To WatchIIAI vs. super fakes. The future of shoplifting. Robots in your supermarket. We look at the top artificial intelligence trends reshaping the retail industry.Many of the AI trends emerging in retail are the result of a wake up call from Amazon, as the
2、tech giant forces retailers to reconsider their e-commerce strategies and innovate in the brick-and-motor space to stay competitive.In our second industry deep dive, we use the CB Insights database to unearth the top AI trends transforming the retail industry.(Check out our first report in the AI In
3、dustry Series: Top Healthcare AI Trends To Watch.)IIIWhy now for AI in retail?Amazon has driven many of the leading trends in retail AI.The companys expansion into grocery and promises of 1-hour grocery delivery have pushed supermarkets to experiment with AI-run “micro-fulfillment” solutions to stay
4、 profitable.Amazons introduction of grab-and-go cashierless stores has spurred an unmanned store frenzy in China, and smaller startups are emerging in the US to service other retailers who want in on the cashierless trend.Other emerging applications of AI have been driven by a desire for more visibi
5、lity and transparency such as computer vision for in-store inventory monitoring, or neural networks for tracking goods through the supply chain.But retailers hoping to capitalize on these technologies to keep up with tech giants like Amazon still have a ways to go.In an analysis of 1,600+ earnings c
6、alls from more than 50 publicly traded US retailers, only 9 retailers had begun discussing AI-related strategies for their websites or physical stores as of January this year.However, we are seeing increased investment in retail AI. Retail AI startups raised $1.8B across 374 deals from Q113 Q318. (T
7、his excludes sales, marketing, and advertising startups providing AI solutions for clients across different industries.)IVCompetition from Amazon, demands for transparency, and increasing investment are all transforming the retail AI landscape.Here are the top retail AI trends to watch as the indust
8、ry takes shape. VBrands pay big to be on supermarket shelves AI is making sure theyre visibleAI vs. super fake goodsAR & computer vision make beauty brands data richMicro-fulfillment centers help grocers connect with online shoppersThe voice shopping revolution that never wasWalmart goes all in on r
9、obotics R&DShoplifting in the age of grab-and-goChinas unmanned retail frenzyFood space goes driverlessRise of the AI stylist141114182023262931Table of ContentsVIAt CB Insights, we believe the most complex strategic business questions are best answered with facts. We are a machine intelligence compa
10、ny that synthesizes, analyzes and visualizes millions of documents to give our clients fast, fact-based insights. From Cisco to Citi to Castrol to IBM and hundreds of others, we give companies the power to make better decisions, take control of their own future, and capitalize on change.VIIThe CB In
11、sights platform has the underlying data included in this reportWHERE IS ALL THIS DATA FROM?CLICK HERE TO SIGN UP FOR FREEVIIIBeti Cung,CORPORATE STRATEGY, MICROSOFT“ We use CB Insights to find emerging trends and interesting companies that might signal a shift in technology or require us to realloca
12、te resources.”TRUSTED BY THE WORLDS LEADING COMPANIES1Brands pay big to be on supermarket shelves AI is making sure theyre visibleShelf space is limited, and brands pay thousands of dollars to compete for it. Beyond that, they have little visibility into what goes on in the store. AI is changing tha
13、t.“Slotting fees” are not a new concept in retail.Apple & Eve spent around $150,000 to secure freezer space for their fruit punch product in a few stores, while Frito-Lay paid an average $100,000 per supermarket chain to introduce a new product, according to a 2001 study published in the Journal of
14、Law and Commerce.Earlier this year, Whole Foods considered charging its top vendors around $300,000 for several weeks of prime shelf space.But once brands sign a contract with the supermarket, they have little visibility into what happens on the supermarket shelf and whether their products are displ
15、ayed as promised.12Startups are capitalizing on this by selling real-time store data.Computer vision platform Trax Retail, for instance, uses images captured by in-store cameras, robots, or mobile phones to stitch together a digital version of the physical store.Trax said 95% of its business is with
16、 manufacturers (like Coca- Cola and Henkel), in an interview with Computer Business Review.Traditionally, brands send salespeople or auditors to manually check product placement in stores. While Traxs product still requires people to visit the supermarket, it is attempting to auto- mate the tracking
17、 of metrics like share-of-shelf and distribution.Supermarkets like Walmart are looking into selling this data directly to manufacturers. Below is an excerpt from Walmarts recent patent application.3Retailers need in-store data to track items and manage inventory, among other things.Walmart partnered
18、 with Bossa Nova Robotics to monitor misplaced price tags and missing items on shelves in 50 of its stores.But the above patent hints at a more futuristic plan, where in-store retail robots receive requests for real-time data from an external vendor (or a CPG company like Coca-Cola).The system then
19、charges the vendor for the task before the autonomous robots complete it.Startups could also potentially partner with both brands and supermarkets as Trax has already begun to do and monetize by selling the same computer vision application for different use cases.But the retail store environment sti
20、ll poses unique challenges for computer vision algorithms. Two CPG brands may have very similar packaging, items stacked behind one another may not be visible, or items thought to be missing could be tucked away or in a different aisle. Startups are acquiring other startups to improve their technolo
21、gy and add additional datasets.In July 2018, Bossa Nova Robotics acquired Hawxeye, an AI company developing face detection and object recognition technology. Trax acquired retail intelligence company Quri in January, and Nielsens Store Observation unit last year.4AI vs. super fake goodsFakes are get
22、ting harder to spot. Online shopping is making it easier than ever to buy fake goods from luxury handbags to watches and cosmetics on the internet, forcing brands and pawnbrokers to experiment with AI.From drugs to handbags to smartphones, counterfeiting is a problem that affects all types of retail
23、.Some product imitations look so authentic that they are classified as “super fakes.”A simple keyword search on the CB Insights platform shows that discussions on counterfeits are trending up.25Chinas rapidly growing e-commerce platform Pinduoduo mentioned “counterfeit” 11 times in last quarters ear
24、nings call, describing “a very hard fight against counterfeit goods and problematic merchants.”“In 2017, weproactively removed a total of 10.7 million problematic products and blocked 40 million links thatraised infringement issuesWe have also partnered with over 400 brands to work together on comba
25、ting counterfeit.” COLIN HUANG, CEO OF PINDUODUOPatent applications, including those for anti-counterfeit tech and developing products that are difficult to counterfeit, have been on the rise in the last 5 years.6Top patent applicants here include SICPA (which develops security inks and anti-counter
26、feiting solutions for pharma, banking, and oil & gas clients), Amazon, Merck, and Samsung Electronics.BATTLING COUNTERFEIT GOODSBrands are fighting the war against fakes on two fronts:1 In the online world, identifying and removing online listings that infringe on brand trademarks like brand name, l
27、ogo, and slogans.2 In the physical world, identifying fake goods like luxury handbags that are rip offs.7Online counterfeiting is vast and complex in scope and scale.E-commerce giant Alibaba, which has been under some fire for not doing enough to counter fake goods on its sites, reported that its us
28、ing deep learning to continuously scan its platform for IP infringements. It uses image recognition to identify characters in images, coupled with semantic recognition, possibly to monitor brand names or slogans in images of products listed on its sites.Counterfeiters use keywords and images very si
29、milar to the original brand listing to sell fake goods on fake websites, fake goods on legitimate marketplaces, and promote fake goods on social media sites like Instagram.When one listing is taken down, counterfeiters may repost the same fake product with a different string of keywords.Barcelona-ba
30、sed startup Red Points is using machine learning to scan websites for potential infringements and find patterns in the choice of keywords counterfeiters use. It boasts clients in the cosmetics, luxury watch, home goods, and apparel industries, including MVMT, DOPE, and Paul Hewitt.Spotting fakes is
31、trickier and more manual in the physical world.For instance, when a seller posts a second-hand luxury handbag for sale, or goes to a pawnbroker to trade it, the verifica- tion process usually involves an authentication expert physically examining the bag, including the make, material, and stitching
32、pattern.8Heres how much eBay charges to authenticate one luxury handbag using identification experts:But with the rise of “super fakes” or “triple-A fakes,” its becom- ing nearly impossible to tell the difference with the naked eye.Building a database of fake and authentic goods, extracting their fe
33、atures, and training an AI algorithm to tell the difference is a cumbersome process.For instance, startup Entrupy had to buy handbags, including from family and friends who owned luxury products. They shopped for genuine and counterfeit handbags, and worked with authentication experts to build a dat
34、abase for training the algorithms for 2 years.The process is harder for rare vintage luxury goods.Entrupy developed a portable microscope that attaches to a smartphone. When users take and upload a picture of the prod- uct (handbag, watch, etc), AI algorithms analyze microscopic signatures that are
35、unique to each product, and verify it against a database of known and authentic products.After a one-time fee of $299 for device set up, Entrupy offers dif- ferent packages, ranging from $99/month for 5 authentications to $999/month for 100 authentications.The database is growing, but there isnt a c
36、omplete set products out in the market.A paper published by Entrupty highlights some other operating assumptions and limitations.9The key idea is that objects manufactured using standard or prescribed methods will have visually similar characteristics, compared to the manufacturing process a counter
37、feiter would use (non-standardized, inexpensive mass production).Secondly, the tech may not work for things like electronic chips that are nano-fabricated (variations at a scale that Entrupys microscope cannot detect).Cypheme is taking a different approach.Its ink-based technology can be used as a s
38、ticker on the product, or directly printed onto labels and packaging.Nikkei Asian Review detailed the tech in an interview with the CEO a random pattern is generated from a drop of ink, the pattern is surrounded by another circle of orange ink that Cypheme claims is proprietary to the company and im
39、possible to replicate, then each unique pattern is associated with a specific product on a database.It uses a smartphone camera and neural networks for pattern recognition to verify the ink pattern for the specific product against its database.This means Cypheme has to work directly with brand manuf
40、ac- turers to make sure products are shipped with the tracing ink.It recently entered into a partnership with AR Packaging, a leading packaging company in Europe working with food brands like Unilever and Nestle.While printing ink on packaging is efficient for tracking an item from the manufacturing
41、 plant and along the distribution chain, the tech doesnt work for secondhand purchase authentication.For instance, a buyer may remove Cyphemes sticker from the packaging of a luxury watch, and decide to resell it at a broker shop or online. In this case, verifying authenticity is not possible unless
42、 the printing is part of the product itself.10Cypheme claims the entire ink print is around 12mm, making it a viable solution for printing directly on products, like the inside of handbags or shoes.The solution for luxury brands and other high-stake retailers, moving forward, may be to identify or a
43、dd unique fingerprints to physical goods at the site of manufacturing and track it through the supply chain.11AR & computer vision make beauty brands data richVirtual try-ons serve a dual purpose in beauty retail: to solve beauty shoppers pain points and, more importantly, to collect data on consume
44、r and product preferences for retailers.Beauty is one industry where augmented reality applications have already seen success.Augmented reality became mainstream across the space in 2017. Perfect Corp and Modiface, which both offer virtual try-on technology for beauty brands, established themselves
45、as the go-to providers of augmented reality for the beauty industry and have worked with major corporates including Sephora, Estee Lauder, LOral, and others.Both Perfect Corp and Modiface combine augmented reality and computer vision to let shoppers virtually try on different looks, while simultaneo
46、usly collecting behavioral data for brands.Modifaces tech collects a variety of data points around facial characteristics, including face shape, skin tone, wrinkles, and more.312This can help retailers determine how people with specific facial characteristics may be more likely to purchase certain t
47、ypes of products, thereby potentially predicting inventory with greater accuracy.Initially, the technology was used to solve the beauty shopper pain point of trying on makeup easily, without the mess.LOrals acquisition of Modiface earlier this year has helped the company launch a variety of AR-power
48、ed beauty experiences for LOrals beauty brands.It recently launched a long-term partnership with Facebook to create AR beauty experiences for its portfolio brands on the social networks platform. Within Facebook, users can virtually try-on products using a smartphone camera, and then be seamlessly r
49、edirected to parent sites to make a purchase.LOral is also rolling out Modiface-powered web-based try-ons, as seen with the LOral Paris brand.There was some question of whether Sephora and other beauty brands would cease working with Modiface after its acquisition by LOral, according to WWD. However, Sephora still plans to use Modifa