Skip to Main Content
Skip to Footer
At The Lighthouse in Venice Beach, Shopify’s Adena Merabi moderated a live conversation with Simplistic’s Zac Cherin and TikTok’s Aaron Kay on the changing path to purchase. They got into how LLMs are changing product discovery, where TikTok fits in, and what merchants should be doing now. Watch the highlights below, or keep reading for some of our favorite takeaways.
Just a few years ago, someone looking for the best hiking boots for an upcoming trip might have typed “best hiking boots” into Google, opened a few listicles, and compared the options.
Today, that same shopper can ask ChatGPT for recommendations based on their destination, budget, and specific needs. From there, they might search TikTok to see people wearing and reviewing the boots, then head to the PDP to compare the details and make a purchase.
When Does AI Go From Recommending to Buying?
Right now, most AI shopping is still conversational. A shopper asks questions, gets recommendations, and uses those answers to decide, but they’re still the one who clicks “buy.”
Agentic commerce takes that a step further. Instead of helping with the decision, the agent can take on the task itself: researching the options, choosing a product, and completing the purchase on the shopper’s behalf.
Zac expects the shift to agentic commerce to happen faster for some purchases than others. Handing an agent the reins on a commodity purchase is one thing. Asking it to buy you a car or a specific piece of furniture is another. For those everyday purchases you don’t want to spend your time thinking about, Zac doesn’t think it will be long before you can just tell an agent what you need and when you need it, and let it handle the rest.
Your Product Data Matters More Now
Product data used to be a nice-to-have. Now, it can determine whether your product shows up when someone asks ChatGPT what they should buy.
Zac used footwear to explain why the details matter. A catalog that stops at “high top” or “low top” can only tell an LLM so much. Add details like “men’s” and “waterproof,” and it has more to go on when deciding whether that particular shoe matches what someone is asking for.
That doesn’t mean writing your product descriptions for a machine; they still have to work for people, too. You need to give an LLM enough detail to understand the product, but the copy still needs to sound natural and true to your brand.
Your catalog isn’t the only source for an LLM, either. Zac compared LLM citations to backlinks in SEO. Reviews, editorial coverage, Reddit threads, and other credible mentions can all become sources that ChatGPT cites when recommending products.
Where Does TikTok Fit Into the New Path to Purchase?
An LLM’s recommendation doesn’t necessarily end the search. Shoppers still want to see products in action, and TikTok gives them a place to do that. Creator content, reviews, and real-world examples can all help turn a recommendation into a purchase.
For merchants wondering how to get their products onto TikTok, the TikTok integration with Shopify makes it easy. The sales channel lets you sync your product catalog, set up tracking, and launch ads directly through Shopify. You can be up and running in just a few clicks.
Aaron shared the example of one merchant whose product was picked up by a creator. The merchant reached out to the creator and used the content they’d already made to promote the product. That became their ad strategy. It’s a good example of how TikTok can get your products in front of creators as well as customers.
Your Site Still Has to Make the Sale
Even though direct checkout through AI is an option, most purchases still happen on the brand’s website. Zac noted that over 90% of transactions attributed to ChatGPT still come from shoppers being referred back to the merchant’s site. So you want your products to appear in ChatGPT, but your site still has to convince the shopper to buy once they get there.
There’s another advantage to having the customer buy on your own site: you own the data. With Shopify, you can use that customer data to retarget them, find similar customers, and bring them back again. On a marketplace like Amazon, the marketplace owns that data.
3 Ways to Improve Your LLM Visibility
-
Check your 10 best-selling product pages.
Read each one as a customer first, then from an LLM’s perspective. If someone asked a specific question about the product, would the page be able to answer it? Make sure details like fit, materials, sizing, use cases, and where the product is made are all included.
-
Add the questions your customers are already asking.
Look at the questions customers ask most often through your support channels and add them to the relevant PDPs. The FAQs can be expandable. What’s important is that the answers are on the page for an LLM to read.
-
Get more specific with your product data.
Check your tags, descriptions, and collections. Are they specific enough to tell an LLM what makes one product different from another? “High top” says a lot less than “men’s high top waterproof.” If product variants are listed separately, group them under the same product.
What’s Next for AI Commerce?
We’re still in the early days of AI commerce, but what merchants do now could make a big difference as it becomes a more common way for people to shop.
We’re already working with Shopify merchants on product data, SEO, GEO, and AI search optimization. Talking directly with Shopify and TikTok means we can take what we’re learning and apply it to the work we do with brands.
Missed the event? Catch the highlights from The Lighthouse.
Want to know where your brand stands? Request a free SEO/GEO audit to find the gaps in your AI discoverability and how we can help you close them.
Leave a Comment