Expectations for easy, seamless experiences — driven in part by the high bar set by Netflix, Spotify, and TikTok — have spilled over into other facets of consumers’ lives, including shopping.
These global brands, and many B2C and B2B commerce brands, employ artificial intelligence to drive their customer experiences. However, AI isn’t a single science. It’s a landscape of technologies and tools working together to help machines think, understand and behave more like humans, and achieve intelligent behavior.
With the number of AI experiences and applications currently being employed in retail environments, it’s clear we’re well into the era of AI-powered commerce. Generative AI solutions, including ChatGPT, have led to increased expectations from shoppers and created a higher standard for relevant, conversational experiences.
While there’s an array of AI offerings available to brands, many of these solutions don’t offer genuine AI capabilities. And when AI is implemented, it’s sporadically used to re-rank products on search pages or applied in isolated parts of the search and discovery process. This can lead to a disjointed customer journey and experience.
The challenge for brands is to harness the power of true AI to make the biggest impact on customer engagement, conversions, loyalty and, ultimately, revenues. Two applications that can make an almost immediate impact on margins and the bottom line are product recommendations and search.
Product Recommendations Have Evolved
Amazon.com kicked off the e-commerce product recommendations revolution more than 20 years ago. Today, McKinsey reports that recommendations drive an impressive 35 percent of Amazon’s sales. However, what works for a global marketplace like Amazon may not work so well for brands and retailers with limited historical data, which continues to be a major challenge. We’ve found that 70 percent to 95 percent of customers visit some brands’ websites only once or twice a year, usually without logging in or setting up an account. This puts retailers at a disadvantage and can lead to a less-than-on-point shopping experience.
AI can be used to deliver recommendations to even the most sporadic visitor by leveraging in-session signals to provide tailored recommendations without shopping history data — e.g., if a shopper views a kayak product detail page (PDP), recommendations suggest related accessories such as a paddle or lifejacket. These session-aware recommendations represent a radical paradigm shift by enabling unique, personalized shopping experiences without relying on extensive data or logged-in users.
AI is Critical to Improving Search
Modern search engines are expected to guide and inspire queries, offer curated filters, retrieve accurate results, provide rich answers, and ensure fast, reliable and conversational interactions. Thanks to Google’s introduction of query suggestions in 2004, shoppers have become accustomed to the relevant, helpful and personalized suggestions provided by Google or Amazon. This approach can formulate effective queries, correct typos, save time, and standardize language. However, retailers sometimes find it complex and challenging to implement query suggestions on their site. Without AI to ensure relevance, consumers could encounter problems finding the products they’re looking for, creating frustration, lower conversions, and fewer repeat visits.
Today’s modern search must be able to:
- Handle vague and long-tail queries, interpreting intent to return relevant results.
- Vector search — specifically, word embeddings — can address the mismatch problem in which consumer queries don’t align with product information.
- Incorporate vector search along with other methods to calibrate for the most relevant results.
Machine learning can help increase the precision of shopper searches by pre-filtering products on the results page based on an attribute in the query. This also helps customers find what they want, faster.
Gen AI and Next-Gen Product Discovery
Gen AI won’t take the place of search anytime soon, but it can increase the accuracy and relevance of e-commerce product discovery journeys. Gen AI is the ideal tool for answering shoppers’ more complex and specific product questions. For example, “What is the best type of BBQ for a small patio?” or “What is the ideal black dress for a night wedding in wintertime?” It enables consumers to ask questions at any point in the product discovery process, with answers extracted from support documentation and delivered to them immediately. Gen AI has the potential to radically transform product discovery for B2C and B2B brands, showcasing their product expertise using information from blogs, FAQs, documentation, etc.
In 2024, shoppers have a world of choice at their fingertips. They expect individualized, consistent experiences regardless of how they interact with a brand throughout their purchasing journey. These expectations have driven demand for a new era in product recommendations, search and discovery. These demands have set the stage for advanced AI and chatbots, exemplified by ChatGPT, to provide seamless, relevant and conversational experiences. The brands that employ an “AI everywhere at once” strategy will not only see a fast return on their investment, but their customers will also “feel seen” and more engaged as a result.
Andrea Polonioli is a senior AI product marketing manager at Coveo, the AI platform built to make digital experiences delightful, relevant and profitable.
View Original Article
