Why Shopping GPT Might Actually Become a Thing

Normally I like to kick these posts off with a funnier fun thing, but this time I’ll have to start with a humble confession. You see. The other day, when OpenAI and Shopify announced that from now on online merchants would be able to sell directly through ChatGPT, my first instinct was to file it under “interesting, given Google’s similar, recent, announcement, but probably overhyped” at least temporarily. But. The more I've thought about it, the more it feels like the wrong impulse.

Yes, “revolutionary” is almost always a bit when it comes to the tech industry characterizing the latest thing but ChatGPT shopping resembles something that's actually working right now, so it’s not too much to say that it’s likely to work too and that if it does work more or less as advertised a variety of things we mostly take for granted about e-commerce will change. Period. So what does it “resemble” in my opinion? Well. It reminds me of thriving social commerce platforms like Twitch Commerce, Snapchat Shopping or TikTok Shop aka a $821billion to $1.2trillion business.

How ChatGPT Shopping Will Actually Work

Ok, let's start with the mechanics, because understanding the technical side tends to help clarify the strategic implications in cases like this. When store owners enable the ChatGPT integration with Shopify, Shopping GPT will be able to access their product catalog. So. If a consumer asks “I need an inflatable green frog costume for my older sister that’s easy to both breathe and run in, preferably for under $100,” ChatGPT can surface relevant products from connected Shopify stores, complete with images, descriptions, prices and even reviews all within the chat interface.

That shopper would be able to find a favorite frog then complete the purchase without leaving the conversation. Payment processing happens through Shopify's infrastructure, inventory will sync in real time, and order fulfillment will work exactly like any other Shopify order. From any merchant's perspective, it's essentially just another sales channel (like Instagram or Pinterest Shopping) but embedded in a conversational AI interface that 100million consumers use weekly.

The technical barrier to entry is pretty low. Retailers will just need to install the ChatGPT sales channel from the Shopify app store, select the products of theirs that they’d like to make available, and the integration will purportedly handle the rest. There won’t be any custom app development or complicated API wrangling or new store infrastructure to maintain (at least not beyond the product metafield optimization merchants should already be doing for conversational search.)

How ShoppingGPT Parallels Social Commerce

That said, the low-technical barrier for store owners wouldn’t count for all that much if the barrier were too high on the consumer adoption side; and I compared this to TikTok Shop and YouTube Shopping up top because I don’t think it is. You see. Social commerce is increasingly popular because shopping is integrated with lifestyle and so collapses the distance between discovery and purchase. Twitch users aren't tuning in to shop. But when a creator demonstrates a product in a way that resonates, the friction between “that's interesting” and “I just bought it” shrinks to 0.

Users are buying in moments of inspiration, before rational any considerations or the impulse to comparison shop even has a chance to dilute their intent. ShoppingGPT seems to follow similar behavioral logic. You’ll hit it up to solve some problem, get a recommendation or research a decision. Not to shop. But your conversations will often have latent commercial intent. “What is ‘frog's together strong’ in reference to? That will naturally lead to products. “How much do those Shepard Fairey HOP signs weigh?" can create context where a frog costume becomes relevant.

In other words, purchase intent doesn't need to exist at the start of a conversation in ChatGPT.

Just like Instagram users don't open the app planning to buy Rihanna’s butta drop deluxe or that red leather Greta Constantine JAMAICA top Jodie Turner-Smith popularized from Tron: Ares red carpets but they consumed the lifestyles and then purchased the products associated with them

“Fighting The Friction That Foils e-Commerce”

Roughly 41% of consumers would be willing to pay more for simple, fast and efficient shopping experiences and fully 50% would switch retailers for less friction in the shopping experience. So. It’s safe to say that they find traditional e-commerce purchase paths frustrating. Formulating a search query, evaluating results, comparing options across multiple tabs and across successive queries, reading reviews from multiple sources, double checking specifications. Each of those steps is another opportunity for consumers to become distracted, to doubt, or abandon outright.

But like social commerce, conversational commerce compresses this journey. LLMs handle product discovery, answer questions in real time, make personalized recommendations based on stated preferences, and facilitate a purchase. All within one conversational chat. No new browser tabs. No switching between tabs. No checking or re-checking info across multiple sites.

Which just seems like something that’s going to matter to shoppers shopping, particularly for certain kinds of products. Complex products that require explanation benefit from conversational context. Personalized recommendations gain credibility when they emerge from dialogue rather than algorithmic suggestions. Impulse buys dependent on maintaining buyer momentum are more likely if the path from “I am interested in that” to “I am buying that” can be measured in seconds instead of in minutes or more.

Just consider the difference in friction encountered in these two scenarios.

Traditional Path … The consumer googles "inflatable frog costume," opens eight tabs comparing cooling fan or inflation pump options, gets distracted by an email notification, forgets which tab had the funniest one with best fan, pump and customer reviews, then eventually abandons your cart, because checkout requires creating yet another account.

Shopping GPT … The consumer asks ChatGPT "I need an inflatable frog costume with an internal cooling fan and the pump to inflate it included," receives three tailored recommendations with explanations of why each fits their size, feature and price preferences, asks follow-up questions about customer reviews and warranty variations, then completes purchase for them … without their ever having to leave that original chat.

The reduction in friction is substantial. Whether that translates to improved conversion rates will likely depend on product category and customer psychology but the structural advantage’s clear

That Customer Behavior Implications of ShoppingGPT

The interesting question probably isn't whether or not your customers will be willing to adopt and buy using SHoppingGPT (or using its competitor in Google AI Mode) but rather what types of products are likely to align best with the consumer behaviors most likely to emerge from its use.

These three categories seem like particularly likely candidates.

Discovery-driven purchases … Of products consumers didn't know they needed until they surfaced by the LLM in a chat. Gift shopping falls here naturally. So do solutions to problems people are actively researching. Like the cooling fans and inflation pumps I’ve been joking about above, were suggested in GPT’s answers to my initial joke questions. The conversational format just excels at connecting latent needs with relevant products.

Purchases dependent on advice … Which is to say of products consumers tend to want guidance with, but don't trust algorithmic recommendations of. Things that fit into my larger point herein like say oil body skin that’d be bought on Instagram Shopping after it was seen on an influencer’s skin or hobbyist tech like a razer klishi ultra controller bought via Twitch Commerce after the shopper watched another gamer finish off Ghost of Yōtei using it. So products suited to contexts where the illusion (or the reality) of the shopper having gotten personalized advice from the seller adds value to the transaction.

Purchases of convenience … i.e. boring replenishment orders, straightforward replacements, items where the consumer knows what they want but value is the reduced friction of not having to comparison shop. “I need more of the blue, pink and white mixed color glitter I bought last month’s glitter bombs again this month” becomes a quick single conversational exchange with your store, rather than a multi-step navigation to your site.

So right now you’re probably wondering what kinds of products probably won’t work? Well. I’m tempted to say the kind of purchases heavy deliberation typically proceeds. Like we don't really impulse buy home furnishings. I’d add similar visual-heavy categories where browsing images is central to decision-making, or products where touching/trying before buying matters significantly.

Which Merchants Should Experiment with ChatGPT?

Even though I think that maybe this is one of those e-commerce “revolutions” that might really change many things if not everything, I don’t think that every online retailer with a successful Shopify store needs to rush toward their ShoppingGPT integration … YET. That said, there are specific kinds of merchants that should seriously consider jumping on this AI bandwagon early.

You're a good candidate if your products solve specific problems consumers actively research and your customer acquisition costs are high enough that any new channel that promises to extend your organic reach warrants testing. Does your product catalog lend itself to recommendation and personalization? Does your average order value make even modest bump conversion rates worthwhile? Do you already have strong product data, clear, use case centric, descriptions, accurate specifications and product imagery?

You should probably wait if your products require extensive visually intensive browsing to build purchase intent. Does your business model depend on customers comparing multiple options simultaneously? Are your margins too thin to absorb experimentation that might not deliver immediate ROI? Is your product catalog less than well-organized?.

Realistic Assessment of ShoppingGPT

ChatGPT Shopping or rather ShoppingGPT probably won't replace your existing multi-channel sales strategy or single-handedly transform your business anytime in the near future, and it probably won't ever eliminate all of the fundamental challenges involved in being a successful e-commerce retailer … competitive differentiation, customer acquisition, operational excellence.

But at some point, it just might transform the behavioral context within which e-commerce occurs. Turn it or evolve it into one that reduces friction in ways that benefit certain product categories and customer journeys. Its similarity to social commerce suggests it could work when the integration between conversation (the chat) and transaction feels natural rather than forced.

For Shopify merchants in the right verticals, with strong product data and room in their multi and or omni-channel strategy for a little experimentation, the technical barrier is low enough and the potential upside interesting enough to warrant testing the waters. Just approach it as what it is. Just an experiment backed by a reasonable hypothesis, rather than the latest "game-changing" platform.

Sometimes the most revolutionary thing about a new channel is that it actually works as it was advertised. We'll know in about six months whether ChatGPT Shopping joins that rare category.