AI in eCommerce: Trends Redefining Shopping Experiences
ChatGPT has witnessed a decrease in recent months, but the enduring presence of AI remains unwavering. In response, eCommerce platforms are diligently forging their own AI solutions, whether through in-house development or strategic acquisitions, to amplify their product listings, refine search functionalities, and elevate the level of personalization offered to their customers.
The use of AI in professional settings is also on the rise, with a massive increase projected. At present, over 58% of users employing generative AI technologies are applying them in their jobs.
To illustrate the impact of this trend, here are several instances of how leading eCommerce platforms are embracing generative AI solutions.
AI-Generated Summaries of Customer Reviews
In August, Amazon introduced a major improvement to its customer reviews system through the implementation of generative AI technology. The primary goal of these brief review summaries is to provide shoppers with a quick overview of a product’s standout features and potential drawbacks, streamlining the decision-making process.
Each review summary spotlights commonly referenced product attributes and customer feedback extracted from written reviews. In addition, predefined keywords associated with specific product attributes enable customers to easily delve into more comprehensive details with a simple tap.
Suppose a customer seeks to gauge the product’s user-friendliness, a simple click on the “ease of use” attribute grants access to review highlights from other shoppers pertaining to that specific quality.
It’s important to highlight that the review summaries are generated solely from verified purchases found within Amazon’s review database. This method guarantees that customers can promptly assess collective feedback with a single glance, establishing a trustworthy benchmark for informed decision-making.
AI for Product Listing Creation
In September, Amazon unveiled a tool harnessing the power of generative AI for the creation of product listings. This technology is designed to assist sellers in rapidly generating and optimizing their listings on a larger scale.
The tool works by requesting sellers to input a selection of keywords or prompts to describe their product. It then generates a variety of content options for sellers to incorporate into their listings, such as product titles, bullet points, and detailed descriptions. Sellers may use this tool for crafting new listings from scratch or for fine-tuning and enhancing their current ones.
While enhancing the volume of high-quality listings represents a major application of Amazon’s AI tool, sellers must exercise due diligence by thoroughly reviewing their listings to eliminate any inaccuracies or the use of off-brand language.
In a demonstration for TechCrunch, a Walmart representative explained the company’s ongoing exploration of generative AI to enhance the shopper’s journey at every step, ranging from the initial page browsing and search phase to the final purchase decision.
Three upcoming features on the horizon encompass a shopping assistant, generative AI-driven search capabilities, and an interior design functionality.
AI Shopping Assistant
Set to debut in the weeks ahead, the shopping assistant offers an improved, interactive, and conversational engagement for customers. It possesses the capability to address inquiries, offer tailored product recommendations, and provide comprehensive information about particular products. For example, it can assist shoppers by suggesting Christmas gift ideas suitable for Christmas parties.
AI-Driven Product Recommendations
In the near future, shoppers will have the ability to input contextual inquiries directly into the search bar. Leveraging advanced generative AI technology, Walmart’s search tool possesses the capability to grasp context and create a curated selection of products that match a single query.
For instance, if a shopper is looking to organize a unicorn-themed birthday party, the AI will showcase a diverse assortment of items, including balloons, paper napkins, streamers, and more. This innovative feature not only streamlines the process but also eliminates the need for customers to conduct multiple separate searches, thus enhancing efficiency and saving valuable time.
According to Walmart execs, this latest tech initiative aims to give customers the means to discover a broader range of products and brands. This vision is enabled by the platform’s advanced personalized features, starting with an improved search function that effectively organizes and categorizes products.
“Now, we can answer more complex and ‘mission-based’ customer questions, like giving lots of party planning ideas, instead of just returning a single item, like balloons. This means a wider range of products and brands can be seen in response to single queries,” Jon Alferness, Walmart US Chief Product Officer, told PYMNTS.
AI Interior Design Assistant
In addition to a generalist shopping chatbot, Walmart is also actively working on an interior design assistant to aid customers in transforming their living spaces. Alongside harnessing generative AI capabilities, this feature seamlessly integrates augmented reality (AR) technology into the process.
To get started, customers are required to upload a room photograph, after which the system captures images of each item within the space. Subsequently, they can engage the chat assistant for redecoration advice, and the AI will recommend item placements within the room.
Customers retain the freedom to express their preferences regarding which items to retain or purchase. Furthermore, the AI inquires about a designated budget, enabling it to suggest affordable items that align with the customer’s financial constraints.
The forthcoming AI tools are set to be released a few months following the company’s initial foray into generative AI.
Back in August, Walmart unveiled a specialized AI tool designed to cater to the requirements of approximately 50,000 corporate employees. This tool, integrated into the recently launched Me@Campus super app, empowers employees with the means to efficiently manage various aspects of their Walmart work experience.
During this launch, the company expressed its intentions to simplify tasks like document summarization, meeting readiness, and project advancement, all while receiving favorable responses from users.
Back in February, Shopify launched Shopify Magic, an AI tool for generating product descriptions for sellers. Now, the company is extending its capabilities with the introduction of a new AI technology, building upon the foundation laid by Shopify Magic.
Called Sidekick, the new AI-driven assistant is tailored to enhance your eComm operations by providing personalized assistance rooted in your Shopify data, streamlining your daily tasks for optimal efficiency and effectiveness.
Sidekick will be seamlessly integrated as a button within Shopify, offering quick responses to seller inquiries, which covers a wide range of information, such as sales trends and more.
The assistant is also capable of aiding you in keeping your Shopify stores up to date. For instance, it can swiftly implement site-wide discounts upon request, ensuring all items on the website reflect the desired price adjustments.
eBay has recently introduced a new AI-powered listing creation tool that uses images provided by sellers to automatically populate additional product details.
This new feature is poised to streamline the selling process, catering particularly to beginner sellers who might find themselves daunted by the exhaustive information required to create a listing.
It uses the photos uploaded by sellers to autonomously fill out essential listing details, such as product titles, descriptions, item categories, and other pertinent information crucial for potential buyers.
However, eBay’s AI direction seems to have left many long-time sellers dissatisfied with the platform’s recent developments.
The official eBay community forum frequently visited by sellers is increasingly inundated with grievances about the subpar performance of eBay’s listing generator. This tool, which has been undergoing limited testing, has raised concerns, with one community member, vssoutlet, asserting that the AI-generated content is not only misleading but, in some cases, entirely inaccurate.
Vssoutlet highlights an example involving a Pentax SLR camera listing, where eBay’s AI erroneously claimed the camera came with a lens kit—a glaring error (that has since been fixed).
At present, this feature is exclusively accessible through the iOS app, with forthcoming plans, as reported by TechCrunch, to extend its availability to Android users in the near future.
With the aim of recreating the virtual fitting room experience, Google has rolled out an advanced virtual try-on feature powered by generative AI. Initially, this tool is exclusively accessible for women’s tops.
Customers have the option to choose items bearing a “Try On” badge, which enables them to pick a virtual model for trying on tops offered by brands such as H&M, Loft, Everlane, and Anthropologie.
This cutting-edge technology aids people in envisioning how these clothing items might appear on a diverse range of models, spanning from size XXS to 4XL, with varying skin tones, body shapes, and hair types.
Google has additionally enhanced its shopping filters, allowing users to fine-tune their search based on factors such as color, style, and pattern. These filters, initially accessible for tops, harness the power of machine learning and visual matching algorithms to streamline the shopping experience.
Impact of AI-Generated Content on Sellers
AI-generated content like product descriptions and review summaries carry substantial consequences for sellers, given their prominent placement on search results or product detail pages.
For example, Amazon’s review summaries are accompanied by a disclaimer, clearly stating that it originates from AI analysis of customer reviews.
While sellers with predominantly positive product reviews stand to gain advantages, the introduction of AI-generated content introduces certain challenges for other sellers as well.
Amazon’s AI has the capability to assess a collection of reviews and oversimplify apparent product issues. The summarization process might fail to capture subtle nuances, such as instances of user error, which may turn potential customers away thinking that the product was defective.
Even when a product boasts an excellent overall rating, AI-generated summaries might highlight a single negative aspect frequently cited in customer feedback. This singular negative emphasis, driven by AI, could be enough to steer customers toward alternative product offers.
In addition, the potential for inaccuracies or even “hallucinations” (content that deviates from reality) in AI-generated customer review summaries should be a point of concern for sellers. Ill-executed AI content holds the potential to create misconceptions and unjustly tarnish reputations.
In fact, Amazon itself has recently required some authors to disclose the use of AI in their books following increasing complaints from the Authors Guild.
Per Amazon’s guideline, content, be it text, images, or translations within a book, is regarded as AI-generated when an AI tool played a significant role in its creation, even if the writer later made extensive edits. Conversely, if the writer initially generated the content and subsequently utilized AI for editing, refinement, or creative brainstorming, the content is categorized as AI-assisted.
Overall, AI-generated content has a direct influence on sellers, as it helps shape the way prospective customers view their products.
Although the advantages of a more effective and simplified shopping experience are evident, sellers may be right to be concerned about the potential of these AI tools to generate inaccurate descriptions or impartial reviews.
As eComm companies continually explores generative AI solutions, sellers must stay vigilant, recognizing the advantages and possible pitfalls associated with AI technology.