Conversational Shopping: Here to Stay

Have you had the chance to try Rufus, Amazon's conversational shopping assistant powered by generative AI? Personally, I used it during my last holiday shopping, and it went very well.

Amazon chose the perfect time to introduce its shopping assistant: after an initial launch in the United States during the summer of 2024, Rufus was offered in Canada, Europe, and India just in time for the holiday season. It's safe to say that the arrival of this AI assistant has broken down the retention barriers of an unstoppable river: that of conversational shopping!

The conversational shopping train has left the station

If Rufus's launch is so significant, it is because it is being undertaken by a multinational like Amazon, not a lesser-known small player. This is, without a doubt, a fundamental structural change in our online shopping habits. Gone are the days of navigating through irrelevant, useless, or nonexistent filters. Gone are the endless loops on result pages clogged with sponsored ads.

The arrival of Rufus and conversational shopping in the digital landscape reminds me of that sweet moment when I discovered Netflix ("wow, television without ads?") and when I cancelled my cable services never to go back (I almost cried tears of joy!). It was time to jump on the train and follow the revolution. And today's new revolution in digital shopping is undoubtedly the arrival of conversational agents powered by artificial intelligence!

Shopping without ads and without being steered

Do you remember the days when we had to drag ourselves to the shopping mall and waste time browsing the aisles to find what we were looking for? Well, those days are over! Now, you open your mobile, click on Rufus, tell it what you want in the words you want, land on the product page, add it to the cart, confirm, and then you're off. It's like walking into the mall, talking to a clerk for 30 seconds, checking out without a queue, and then going back to playing basketball or watching Netflix. That's my kind of shopping experience!

Want a more concrete conversational shopping example? Here’s a request I recently gave to Rufus: "Find me a quality electric shaver for men with good reviews. I don't care about the price. I want one that shaves really close. And don't give me one with 3 pivoting heads; they’re not comfortable. I want a straight one, you know? Oh, and I want a durable shaver, not some junk."

There are two important takeaways from this search request:

1. I made a single query with all my purchasing criteria. So, goodbye to search queries with multiple little keywords and the use of more or less relevant filters.
2. I used the everyday language that I speak in everyday life, without having to adapt to keywords suggested by SEO specialists. Instead, I let the large language model (also called LLM or Large Language Model) of the AI system do all the work.

After that, all I had to do was look at the 3 or 4 products suggested and click on the option that inspired me the most. And right away, I found the perfect product for me!

The key: understanding the need behind the keywords

Beyond artificial intelligence, conversational shopping assistants rely on advanced infrastructures based on LLMs powered by ChatGPT, LLaMA, Grok, Claude, among others. These technologies interpret complex queries formulated in natural language and then provide precise results.

Creating AI solutions capable of processing requests made in natural language obviously involves developing and maintaining rich and detailed product descriptions. But it is also important to optimize these descriptions for semantic and voice search, because these shopping assistants go far beyond simple keyword matching: they incorporate natural language processing and understanding to interpret the intent behind a search query. So, if you search for "comfortable winter hiking shoes," these systems will not only focus on the individual words you used but will also consider your underlying need for warm and durable shoes suitable for outdoor activities.

Optimal product data management is crucial

To support real-time updates and simultaneous queries by the millions, AI platforms need scalable programming interfaces (APIs) built on resilient cloud infrastructures. These APIs ensure that the most up-to-date information about stock levels, prices, and product features is displayed. To achieve this, an architecture capable of managing massive volumes of structured data is necessary. Companies must also communicate all relevant product data effectively.

In short, effective management of an online product catalog starts with the creation of comprehensive data, including detailed metadata such as SKU codes and barcodes (UPC), as well as accurate descriptions. Product attributes, such as size, color, materials, or technical specifications, must also be standardized to facilitate comparison by AI algorithms.

It is also important to properly implement JSON-LD (JavaScript Object Notation for Linked Data) tags in a conversational shopping system, as these are essential for structuring product data. These tags enable algorithms to quickly detect and analyze information to associate products with standard categories. For example, a product like a coffee table will be better recognized if it is correctly labeled with attributes such as dimensions, materials, or style.

Furthermore, data management must accommodate all possible product variations when multiple sizes, colors, or models are offered. Each product variant must be presented clearly and distinctly in the results. Rich descriptions, including semantic keywords and multimedia elements, add an additional layer of value by facilitating complex, semantic, and voice searches.

Interoperability at the heart of the ecosystem

Interoperability is the hidden engine of this online shopping revolution. Conversational assistants like Perplexity do more than just search for products: they directly interact with data traditionally reserved for CRM/ERP systems and marketplaces to ensure the accuracy of real-time information and maintain perfect consistency across different platforms.

In addition to reducing errors, these dynamic connections between systems also help to decrease processing times, thereby enhancing the user experience. Augmented reality and 3D visualization, combined with these real-time information flows, also allow customers to visualize their product before placing an order.

Ultimately, your product catalog must provide access to API endpoints that allow real-time retrieval of your product information. Gone are the days of periodic information transfers to Google! AI shopping assistants must now always have up-to-date information on your products (prices, sales, inventories, etc.). Moreover, Shop like a Pro, Perplexity's shopping assistant, transfers the entire online shopping aspect of businesses into their interfaces. The commercial sites of these businesses thus act as providers of raw data, which greatly facilitates real-time sales tracking. Such functionality is currently not compatible with traditional tracking of your metrics on Google Analytics.

In summary, what actions should be taken?

Here's a brief recap of the right actions to take to prepare your e-commerce store for the arrival of conversational shopping.

Product Data Management

• Create rich and detailed product descriptions;
• Optimize product descriptions for semantic and voice search;
• Include comprehensive metadata (SKU, UPC, and precise attributes such as size, color, materials, etc.);
• Standardize product attributes;
• Properly implement JSON-LD tags;
• Manage product variations (sizes, colors, models, etc.).

Visual and Technical Product Optimization

• Add high-resolution images of your products from multiple angles;
• Create optimized alternative texts for each image;
• Develop API endpoints that allow for real-time information updates;
• Ensure interoperability among your different systems (CRM and ERP);
• Prepare dynamic and instantaneous data flows.

Search Strategy and Accessibility

• Adapt content to natural and conversational language;
• Consider complex and intuitive search queries;
• Facilitate understanding of the intent behind each search;
• Prepare your catalog for AI assistants.

Major transformations ahead for your online business

The advent of conversational shopping marks a major structural transformation in the world of e-commerce, where artificial intelligence, complex queries in natural language, and precise data management become central. Businesses must adapt to this new technological reality by rethinking their approach to cataloging, searching, and customer interaction to offer a more intuitive, efficient, and personalized shopping experience to their customers.

Need to prepare your online store for these new trends? Our team of e-commerce specialists is here to help you through this process!

Our e-commerce services