Decoding Four Key Technologies That Are Reshaping Digital Shopper Behavior

How AI, VR, AR, and IoT are rapidly changing the way we shop online

Mobile shopping is but the tip of the iceberg that is ecommerce innovation

Throughout the shopper journey, technology is disrupting the way we find, examine, compare, and order products online. New value drivers emerge every year in retail innovation, but the following four key technologies have proven to be crucial in the shifting consumer behavior. With ecommerce now accounting for about 13% of total retail sales with a healthy y-o-y growth rate of 16%, according to the latest data from the U.S. Commerce Department, what happens in ecommerce is bound to see a ripple effect throughout the retail ecosystem. Let’s take an in-depth look at four key value drivers in ecommerce one by one.

AI Drives Product Discovery

Online shopping starts with discovery. Even on the rare occasions where shoppers know exactly what they want to buy, the task of discovering where and which specific brand to buy still remains. At the moment, Google remains a popular starting point for consumers to search for product ideas and information. According to a 2017 survey by marketing platform Kenshoo, 85% of online shoppers would turn to Google at some point of their purchase planning.

However, Amazon is increasingly starting to overtake Google as the front door for online shopping in the U.S., as recent surveys show somewhere between 49% to 55% of online shoppers start their product searches on Amazon, not to mention the 40% of U.S. consumers with Prime membership that are even more likely to start their shopping journey on Amazon. And this is where things get interesting. Amazon is an open ecommerce marketplace, but not every brand is on Amazon. In fact, many big-name brands have decided against selling on Amazon for fear of relinquishing too much control to the ecommerce giant. For those brands, they will be missing out on roughly half of the online shoppers from the get-go¹.

Amazon is increasingly starting to overtake Google as the front door for online shopping for U.S. consumers

Regardless of which of the two sites they start on, the beginning of online shopping journey is heavily influenced by algorithms. Google’s algorithms rank search results based on a mix of keywords relevance, Google Shopping ads, and other SEM buys, while Amazon has its own algorithm-driven results where its own private label products usually get favorable placements if no brand name is specified in the inquiry.

Beyond search results, Amazon also employs an AI-driven recommendation engine to surface an individualized selection of products users may be interested in based on the data it collects. Of course, algorithm-driven product recommendation has been around for a while now, — remember that unfortunate incident when Target figured out a teen is pregnant and sent her congratulatory coupons before her own family found out? — but what Amazon does is considerably more sophisticated. Beyond purchase history and product page visit, Amazon’s recommendation engine also gauges shopper interest based on factors like items saved to wish lists and shopping carts, items they have rated and liked, as well as what other customers have viewed and purchased (aka the “people who bought this also bought…” suggestions). It is not an understatement to say that a significant amount of Amazon’s revenue growth can be attributed to it successfully integrating recommendations across the buying experience — from product discovery to checkout. In fact, 35% of all Amazon sales are estimated to be generated by its recommendation engine.

The outsized influence that AI-driven algorithms have in product discovery online is but a preview of artificial intelligence’s growing influence in ecommerce. With voice shopping and visual search poised to take off in the next few years, AI-driven solutions will grow even more prominent in determining product recommendations and discovery for online shoppers.

The outsized influence that AI-driven algorithms have in product discovery online is but a preview of artificial intelligence’s growing influence in ecommerce.

By the end of 2022, 55% of U.S. households will own at least one smart speaker, and voice commerce will be a $40 billion industry. Amazon, the current leader in the U.S. smart speaker space with over 70% of market share, is a main driving force in this emerging sector and stands to become a major beneficiary if voice shopping truly breaks into mainstream. Only one or two top results will be served to consumers when they shop via voice, determined by, of course, the underlying algorithm powering the AI assistant in use. When shoppers use Alexa to search for an unbranded item they haven’t previously purchased, they tend to be steered toward products under the Amazon’s Choice label, which Amazon’s algorithm selects based on a variety of criteria such as “low return rate” and being “highly rated”.

Only one or two top results will be served to consumers when they shop via voice, determined by the algorithms powering the AI assistant.

While some brick-and-mortar retailers have been making strides in building out their online sales channels and services, they lack the technical know-how and platform when it comes to voice shopping. Naturally, retailers including Walmart, Target, and Carrefour have rallied around Google to catch up with Amazon in this regard. Doing its part to extend the duopoly in online product search into voice, Google also recently launched the Shopping Actions program to unify its ecommerce offerings for retail partners across organic search, Google Assistant, and its shopping service Google Express.

Beyond voice shopping, Google is also actively exploring visual search — another domain made possible by advances in machine learning and computer vision — and its potential in upending how people discover products. As we laid out in our recent piece on the rollout of Google Lens, the rise of visual search will provide a convenient and seamless way for consumer to identify a product and, once the ecommerce component of visual search becomes better integrated, add it to their shopping cart for purchase. In fact, with 75% of U.S. consumers already searching for visual content before making a purchase, one can easily see how visual search would be quickly adopted by shoppers as a new form of product discovery. Amazon, ahead of the curve, added a product recognition feature into its main iOS app back in 2014 to enable visual search for mobile shoppers.

Pinterest is another key player in visual search, and in a way, its Pinterest Lens product is even more optimized for ecommerce. Compared to Google’s all-encompassing general purpose approach to Google Lens, Pinterest Lens is built to specifically recognize fashion and home decor products — two of the most popular categories on Pinterest — from its inception, and that is certainly the angle that Pinterest has been pushing.

In September, Pinterest added a “Shop the Look” feature that lets people tap products featured in a photo to reveal a list of links to similar products available for purchase. Around the same time, the company also announced a partnership with Target to extend this feature to allow customers to snap product photos and browse similar items for sale at the retailer. Last month, Home Depot announced a similar partnership to facilitate visual search for its products. As far as retailer partnerships go, Pinterest is handily winning over Google at the moment.

Of course, visual search integrated with product recognition is not exactly a new thing for retailers. In 2015, JCPenney worked with computer vision startup Slyce to do something similar to what Pinterest Lens is allowing Target shoppers to do. Last year, eBay also launched similar visual search tools last year to let customers find what they want using photos from your phone or web. However, those tools were rarely utilized due to low awareness among shoppers as well as lack of deep ecommerce integration. Now, with the wide rollout of Google Lens to newer Android phones as part of Google Assistant, plus the partnerships Pinterest has been striking with major retailers, those tools could finally break into the mainstream and become part of everyday shoppers’ mobile shopping habit.

With recent wide release of new tools, visual search may finally break into the mainstream and become part of everyday shoppers’ mobile shopping habit.

In addition, AI’s impact on what we discover when we shop online goes even before we start our product search. Many retailers are now starting to recognize the power of machine learning in optimizing their supply chain and logistics management. Zara, for example, uses machine learning to analyze product sales and customer preference data to gauge fashion trends and make its supply chain more efficient and responsive. JDA Software, a leading software provider for supply chain and retail solutions, has been applying machine learning to forecasting demand in order to help retailers with production planning. Even the very design of those products may be actually shaped by AI, as is the case with Stitch Fix’s popular Hybrid Designs line. Given the entrenched way algorithms are driving product discovery today, it is not a stretch to say that AI is quickly becoming the biggest influencer in online shopping.

VR & AR To Transform Shopping Experiences

For all the convenience that ecommerce has to offer, one persisting caveat of shopping online is that it lacks the kind of in-person, high-touch experience that real-life shopping offers². While some online apparel retailers such as Stitch Fix and Trunk Club have tried to address this problem by going for the “try-at-home” model, the applications of virtual reality (VR) and augmented reality (AR) in ecommerce may finally creates shopping experiences that transcend the ones physical stores is capable of.

VR stores have been envisioned as a viable way to digitally recreate the kind of high-touch brick-and-mortar experience in a virtual environment. During the holiday season of 2016, Alibaba worked with Macy’s to transport Chinese shoppers to a virtual replica of the retailer’s NYC flagship store, where they can look around, checking out featured items, and making purchases. Walmart is also reportedly actively exploring VR shopping with its in-house innovation team, Store №8, looking into issues like creating a 3D full-body avatar for virtual try-ons and creating virtual sales assistants to help out VR shoppers.

Besides recreating the physical shopping experience, VR can also be a valuable lead-generation tool for retailers. For example, Lowe’s recently created a branded VR experience to teach its customers to renovate their bathrooms, in hope of inspiring their DIY spirit to take on home renovation projects and thus purchase some tools and materials from Lowe’s. Similarly, IKEA worked with HTC to let users design their own kitchens in VR.

That being said, low consumer adoption of VR headsets (overall U.S. user adoption of VR headsets below 5%) remains a big roadblock for realizing this vision of VR shopping. Right now, it makes more sense for retailers to explore VR as an experiential tool for lead generation or customer service rather than a direct sales channel.

In contrast, AR is much better positioned to become a transforming force in retail in the near future. By overlaying virtual objects on top of real-world environment, AR enables retailers to effortlessly blend the line between the digital and the physical, bringing the products into the homes of online shoppers. For example, IKEA launched the IKEA Place app that leverages Apple’s AR platform to allow users to drop a scaled 3D model of the furniture item of their choosing in their home to see how it looks and fits. Similarly, Topology Eyewear lets people try on glasses a variety of frame styles and colors in its AR app.

If VR aims to digitally transport you to the store, then AR is all about bringing the products to you. Both can be optimized with personal data to provide a customized experience, and they each correspond to the two strategic paths forward (experience vs. convenience) we laid out for retailers in our 2018 Outlook.

If VR aims to digitally transport you to the store, then AR is all about bringing the products to you.

Between the two, VR is naturally suited for enabling the kind of powerful immersive experiences that can do wonders for conversion rates. In fact, bringing customers into virtual stores for browsing is but the first step for VR shopping experiences, which can and should go far and beyond to transport shoppers into various scenarios so as to better showcase the products and push for sale. For instance, if you are selling camping equipment in VR, then why not transport your shoppers into the woods and walk them through how useful your products would be in that setting? And if you are a fashion retailer experimenting with VR, why not make the shopping experience into a fabulous runway show? After all, VR provides the best kind of show-and-tell experience that transcends any in-store sales pitch.

AR, on the other hand, is more tipped towards providing shoppers with the convenience and ease they need to properly examine a product in a real-world context, as an AR experience, by definition, is tethered to the user’s surrounding environment. At the moment, AR edges out VR both in terms of consumer adoption, thanks to the rollout of mobile VR support by all major tech platforms last year, but its full potential in bringing the products to shoppers will only be realized when AR glasses becomes a viable mainstream product.

IoT Commerce To Reform Shopping Habits

The rise of mobile shopping has expanded people’s idea of online shopping. Messaging and social platforms were not originally conceived to be sales channels, yet in recent years, most have grown increasingly ecommerce-friendly as they ramp up monetization efforts. The rapid adoption of smart speakers is already giving rise to voice shopping, but beyond those voice-activated speakers, IoT devices and smart appliances at home hold even greater potential in making ecommerce more seamless and frictionless.

As more and more devices and appliances become connected, suddenly everything in your house becomes a potential ecommerce channel that can fulfill orders and, with some help of AI, even automate the orders for you. In the near future, the deployment of 5G networks will power more connected devices and enable them to carry out shopping capabilities. The future of ecommerce may not be about digitally recreating the brick-and-mortar experience at all, but rather integrating online shopping into every facet of our increasingly digitized life.

To be clear, we are not talking about sticking a tablet to your refrigerator door so you can shop for groceries on it. Rather, the rise of IoT commerce hinges on a deep integration of ecommerce capability, especially digital payments, into the connected home devices and the ability to leverage the data those devices collect to better understand needs. Your smart washing machine should know you are running low on detergent, just as your connected coffee maker should know you’re running out of coffee beans. And in the near future, they should be able to order replenishments for you, or at least prompt you to do so.

Over 20 billion connected devices are estimated to be powering the global economy by 2020, and that presents a big growth opportunity for ecommerce. To that end, some companies are already starting to explore IoT-enabled ecommerce. IBM, for example, is partnering with Visa to integrate digital payment into future connected devices running on IBM’s Watson IoT platform to enable ecommerce capability.

Whirlpool’s latest kitchen appliance lineup gives another example of how IoT-powered ecommerce may work. By integrating with curated cooking experiences provided by recipe site Yummly, their smart oven can auto-adjust temperatures according to the recipe input and ensure the dish come out just right. From there, it’s not hard to see that they are one step away from bringing grocery shopping or meal kit delivery into the fold to help users order those ingredients based on the recipe of their choosing.

Of course, we still have a long way to go to get to that future. At the moment, IoT-powered ecommerce mainly manifest in terms of reordering previously purchased products. Case in point, consider the Amazon Dash Buttons. Wirelessly connected to your Amazon app, these branded buttons can help you reorder household products with just one single push. Dash Buttons are powered by Amazon’s Dash Replenishment Service, which is open to third-party manufacturers. Major home appliance makers including GE, Samsung, and BSH Hausgeräte are already using the service to leverage Amazon’s retail platform to build automatic reordering experiences into their IoT products.

Taking things one step further, Amazon also created the Dash Wand, an Alexa-enabled connected barcode scanner that enables users to scan CPG products for quick reordering. Then there is also the fashion-oriented Echo Look, an Alexa-enabled smart selfie camera that holds strong potential to be integrated Amazon’s “try-at-home” apparel service, Prime Wardrobe, which has recently started rolling out to more customers.

Even our cars, as they become increasingly connected, will also change the way we shop and order stuff. The mobile retail concepts powered by autonomous vehicles that Toyota and Ford showcased at this year’s CES heralds a futuristic vision of a “car-as-a-store.” Buoyed by partnerships with Pizza Hut and Amazon, Toyota’s e-Pallette fleet is set to explore the new ways for retailers to bring the orders and retail experience to the doorstep of online shoppers. When whatever you order online can be delivered to you, wherever you are and whenever it is, via a self-driving vehicle within hours, the ecommerce experience inevitably becomes far more flexible and versatile.

In the end, IoT commerce is not about the setting up yet another online retail channel, and more about how quickly the combination of shopper data, AI-driven systems, and connected devices are going to fundamentally shift our expectations as to where we shop and how we order. When the act of making a purchase become automated, retailers may risk getting cut out as more brands go directly to consumers.

The combination of shopper data, AI-driven systems, and connected devices are going to fundamentally shift our expectations as to where we shop and how we order.

If it comes to be, this new ecommerce reality would create the kind of brand lock-in that completely eliminates the consideration phrase for consumers, making it even harder for new brands to compete with the default choice. To avoid such a fate, retailers will need to work hard at either partnering with the IoT platform owners to supply the goods, or creating a hyper-personalized, high-touch shopping experience that will lure back the shoppers.


  1. Some may argue that selling on Amazon is unnecessary if your brand has a name recognition among shoppers to the point that they are searching by your brand. While that is partially true, it is also important to take into account the sales lost to a competitor brand that is available on Amazon. No wonder we are seeing more and more name brands such as Nike and Calvin Klein embracing Amazon as an ecommerce partner. Plus, even if brands don’t want to be on Amazon, many are still present there through third party sellers, who buy at retail and then sell at a higher than MSRP price to gain margin, in order to capture the audience on Amazon.
  2. While this may not have been a big issue in certain product categories such as books and electronics, this problem has been holding ecommerce back in key retail categories such as apparel and grocery, where it is a must for many shoppers to try on or closely examine each individual item before purchasing.