One Tech Tip: Up your Christmas shopping game with AI tools

Sales assistant chatbots were a new thing a year ago. Now they are everywhere.

After deploying AI-based assistants, online retailers and technology companies have added more artificial intelligence features to make online shopping easier and more convenient.

The latest generation of AI-powered shopping services and tools debuted in recent weeks, just in time to kick off the holiday shopping season that begins with Black Friday.

Here’s a look at existing and recently launched AI services that can help you in your search for the perfect gift in the run-up to Christmas:

Amazon led the way by rolling out its Rufus chatbot in 2024. Other e-commerce sites followed with their own AI assistants to improve the online shopping experience.

Walmart’s Sparky chatbot is available on the retail giant’s app and can summarize reviews or offer product recommendations based on occasions, like Christmas. Target recently unveiled a gift-finding chatbot on its app, but it’s only available for the holiday season. Ralph Lauren has partnered with Microsoft on the “Ask Ralph” chatbot to provide style recommendations.

The goal of deploying chatbots is to make it easier for Internet users to find what they are looking for. Instead of typing search terms and keywords, you can type or use voice dictation for a conversational exchange.

The results, in my experience, can be mixed.

I recently tried Rufus to find a replacement stainless steel pot for my rice cooker, as well as a protective trivet for my kitchen sink faucet. In both cases, the results weren’t quite right and didn’t seem to reflect the wide range of products available. Sometimes the results were completely unsuitable.

I ended up doing a more careful search of product listing photos and specifications to find the right items. Part of the problem, I suppose, is that I was looking for generic products. Searches for branded products may produce better results.

Perhaps you don’t want to limit yourself to just one retailer’s website in your Christmas shopping search. Or you don’t know exactly where to find the perfect gift.

Tech platforms have deployed AI-powered shopping tools that can cast a wider net by searching across multiple sites.

OpenAI added a new “shopping search” feature to ChatGPT last week, which can provide personalized shopping advice for products that require a lot of detailed specifications, like electronics or home appliances. The feature will activate if you ask ChatGPT a purchasing-related question or manually enable it in the chat window. OpenAI claims it can go beyond simple questions, such as verifying a price or feature, that classic ChatGPT can easily answer.

Google users can have a similar experience when using its AI-enabled search engine, which recently received a major update for shopping searches. The company says users can describe what they’re looking for as if they were talking to a friend and get an “intelligently curated response” based on 50 billion product listings, with images alongside prices, reviews and inventory information.

Google added similar shopping features to its Gemini AI chatbot app for US users last month.

Meanwhile, Perplexity last week unveiled its own shopping assistant feature, capable of tailoring recommendations based on previous searches.

I asked all three to find a soft cotton flannel shirt. ChatGPT and Perplexity asked me for specific requirements, such as budget and must-have features. ChatGPT’s response was the most detailed, with options from six brands, including its top pick, and included photos, pricing, and point-form summaries for each shirt. The results were also compiled into a comparative table.

The results from Google, which didn’t ask follow-up questions after my initial request, seem the most general. Perplexity’s results fall somewhere in between.

So you think you’ve found a stylish cardigan for your partner. But you’re not sure about the silhouette or the mood.

Generative AI try-on tools allow users to see how a garment might look on the wearer.

Existing virtual locker room tools rely on complex 3D rendering, real photo shoots and augmented reality. Often, buyers were limited to choosing the model that best suited their body type to see how the clothes fit them.

Google is now leveraging AI to allow shoppers to virtually try on clothes and shoes using photos of themselves in simple poses. Among the exceptions: accessories such as hats or jewelry, swimsuits and lingerie.

To use the feature, available through Google’s desktop search and mobile app in Australia, Japan, Canada and the United States, simply tap the “Try” button on a product photo, then add a full-length photo of yourself. You can then save the image of you with the item tested or share it. The original photo is also saved to your account, so you don’t need to keep uploading new images.

If you’re buying a gift for someone, Google says you can upload their photo, but only if you have their permission.

Now that you’ve figured out what exactly to get for the special people on your Christmas gift list, it’s time to buy. But if you want to outsource some of the necessary steps, there are “agentic AI” tools that can help you.

Amazon users can use an “AI agent” to purchase a product on their behalf if the price falls to the desired level. Google has launched its own “agent payment” feature, which allows you to automatically purchase a product you’re monitoring through its price tracking feature. Google’s feature has been rolled out to a small group of retailers, including Wayfair, Chewy and Quince, as well as some Shopify merchants.

Both companies say they will always confirm with you before the AI ​​agent makes the purchase.

Amazon goes a step further by allowing shoppers to purchase items that aren’t in stock directly from other brands’ websites. If you see a product on the Amazon Shopping app with a “Buy for Me” button, you can purchase it through Amazon’s regular checkout page, but the AI ​​agent will then complete the transaction on the other brand’s website with your encrypted payment information. The feature was in test mode but is being rolled out more widely.

Prefer to buy in person? It’s a good idea to make sure a physical store carries the product you want before you go there. Google has launched an AI service that will call local stores to ask them.

It is only available in the United States for toys, electronics, and health and beauty products. When you do a Google search for the product you want, add “near me” to the end of your search query. Then, if you see “Let Google Call” when you scroll through the results, you can tap the “Get Started” button. Answer a few questions about what you’re looking for, whether you want email or text updates.

Google will then contact stores near you to ask if the item is in stock.

The bot works quickly but results may be limited. When an AP reporter in New Jersey asked Google to call about a specific Acer monitor, the agent returned quickly with a response from a local computer repair shop that sold refurbished monitors. He seemed unaware of the big box stores selling electronics nearby.

According to Google’s text update, the local repair shop didn’t have the monitor, but had one of similar size – without the other bells and whistles – for a lower price.

___

Is there a technical topic that you think is worth explaining? Email us at onetechtip@ap.org with your suggestions for future editions of One Tech Tip.