ChatGPT Giving up on In-Chat eCommerce?

[music] Hello and welcome to the dojo, a search marketing podcast by Exposure Ninja. My name is Dale Davies. I'm the head of marketing at Exposure Ninja and I'm joined by our fantastic CEO, Charlie Merchant. Hey Charlie, how's it going? Going great, thanks Dale. How are you? Pretty good. I'm uh I've got a lot to say, a lot to share about our news this week because it excites me terribly. So, I'm going to get us straight into it. Um what's the news, Charlie? The news. So, the news that we're talking about this week is that ChatGpt, which is owned by OpenAI, are pulling back from having instant checkout directly inside ChatGpt. So instead of purchases being able to be made on an instant checkout inside Chat GBT's conversational setup, they're instead going to have apps that connect to Chat GBT from retailers. So that actually they're not trying to get people to check out immediately in the Chat GBT conversation, but they are funneling them through. There's so much I have to say about this. It's like it really it kind of excites me and also frustrates me and kind of makes me question like is it are things just not going well over there that they would change the way that this works that the functionality works. Like why do you think that they've decided to make this change? Well, I expect they've been testing it and it probably just hasn't had the uptake that they hoped it would have, but perhaps they overestimated how much uptake it was going to have. I was always a little bit skeptical only because building consumer trust takes a long time and if we're used to checking out when we're buying if we're thinking about e-commerce and we're thinking about buying products we trust certain sites right a lot of the time someone will go to Amazon rather than buy from a small retailer even because they've just got prime they're used to buying on Amazon it's a habit that they've got that sticks or you want to go to the site that actually owns the thing you want to buy your speed cats from the Puma website or you want to buy it from a retailer that you know and trust like shoe or office shoes for example rather than a random site that you're not sure about that you haven't heard about. So I think there was always going to be a trust issue with bringing something like instant checkout into chat GPT for consumers. My hope had been when they actually brought this in though that they were just going to keep pushing to eventually change that search behavior and they weren't going to expect like an immediate uptick on people doing instant checkout inside chat GBT results because I don't think they were ever going to have a major off the bat success story with that. It just takes much much longer to build the trust with people who are searching in chat GBT. the same for other AI results in my opinion as well. I think Google's AI mode and Gemini are gonna have more clout because Google is of course dominant. We're used to searching on do on on on Google that we're used to that being our dominant search engine. So I think they're going to have an easier time of introducing something like this whereas I think Open AI were just going to have to have patience and it feels like they've pulled back really early to me. But on the other hand, I can see why they've done it if they were thinking perhaps they were moving too fast for the average consumer. Yeah, you nailed exactly though. The thing that really interests me is around like that trust of this the user that uses Chat GBT is so unfamiliar with the platform. Even the ones who are more power users, they're not really kind of used to a certain type of shopping or a certain type of behavior when it comes to navigating through the web. Especially when, and this is the bit that truly excites me the most of all is the different types of audiences that are using chat GBT versus a Google or a Claude or anywhere else like as my as far as I understand it, it's a certain type of person that uses chat GBT. It's a lot of people like looking for educational support and a lot of people in academia and so on versus Gemini sorry versus claw Google which is more of a general audience is like the everyday person who as you said has grown familiar with Google over years and years and years whereas the people using chat GBT are like help me to get through my homework faster or easier. Yeah. And we also know that the majority of chatbt's user base overlaps with Google. It's not that the users who are searching on chatbt have just given up Google and that's the end of that. They're using both. So what we believe is happening is people are actually having top offunnel conversations in chatbt. They're doing all of their research phase, their discovery looking for what is the right product for them. They're looking for the right running shoes. They're looking for the right paint color for their bedroom. They're looking for a kitchen tap replacement that's not going to spew water all over their wooden worktops inside the chat GBT conversation. But when they've actually then found the product that they want, they're just going and typing it into Google. And that might be a branded search. So they might decide, you know, I'm definitely buying my new tap from B andQ. They just type B andQ. Or it might be that they found a specific type of shoe. They're like, "Okay, I want the navy blue Puma Speed Cat shoes." And they type that product name into Google. So, they're finishing their purchase journey outside of ChatBT. And sometimes completely unconnected, right? So, in in that case, they're not following through a link from ChachiBT results. Sometimes they would be able to, especially I think if you're looking at things like running shoes as a perfect example, that is going to bring up something more like shopping results. you're going to start to get visual products that you can click through from chatbt a lot of the time in those types of conversations. Some people will do that, but others will just still go to Google and do it the old-fashioned way. It's really interesting that they're pulling back. And I think you may be right and very likely to be right in that maybe they just introduce this stuff later on. Mhm. At the same time, Google are also updating their main like instant checkout uh version of instant checkout within Google and within AI mode, adding new things like cart. So, not only can you add a product and buy it within AI mode or maybe an AI view or however that is working, you can now add additional items from the same shop within the same like purchase journey as well. And on top of that, they've added like a new functionality for catalog. So you can see, you know, a full range of other things that available from like Lowe's or whatever e-commerce, whichever retailer is you're going through. And as well as that, they've added in or it's coming soon, the functionality to still get the loyalty program points and all that kind of stuff. So you would typically complete your purchase through the ecom store and you get your points for shopping, whatever it is. They're finding a way to bring that in it, too. So, you don't even have to leave AI mode or whatever, you know, interface in order to get all the benefits of shopping through that website without having to visit the website and you just, in fact, in the example I saw uh this morning, [clears throat] if you're on if you're searching for a product and you don't currently have an account, it says, "Would you like to create an account with your Google account?" So one click it creates you an account for that for that retailer and then I presume gives you all that loyalty kind of stuff and then you just use your Google wallet to complete the purchase. Literally everything is everything is within that interface and at the same time chat pushing pulling back. It's wild. It's wild, but Google has the trust here, right? Like Agentic Commerce, I don't think anyone's denying that Agentic Commerce is the future. That is the direction that e-commerce is moving. It is what's going to be happening. The only question is the infrastructure to do it, which Google is building now, and it's complicated. It's not as easy as it sounds, like, oh, we'll just put in an instant checkout and it will work. There's a there's a lot going on with the UX, with the consumer trust, with how we shop and what shopping habits online actually look like. So, I guess Google's operating from a a kind of forward focused position of they're the major player. They are going to be the ones that lead in aentic commerce. And I guess OpenAI was hoping that Chat GBT could take some lead and has perhaps just realized that they're not in a position to do it, that they don't have the trust level needed to do it. And so backtracking or backpedaling to actually then be thinking about okay let's just think about top of the funnel. Let's think about using it as a discovery. Let's get people used to navigating through to websites and buying products from chat GBT because they are having success there as well. We already know that there's a lot of businesses experiencing higher conversion rates from traffic that comes through from ChachiBT. Just last week, I was looking at uh a window installer, a national one here in the UK, and their average conversion rate is around 4.5% for organic traffic. But then you look at the chatbt traffic, which is a much smaller percentage, but it's converting around 7.3%. It's significantly higher. And it kind of makes sense in something as complex as buying new windows. It's really difficult to convert on a website straight off the bat because it's a high ticket item. You're comparing lots of different installers. A lot of the time people not even sure necessarily what they're looking for because it's an infrequent purchase. It can be 10, 20, 30 years if not more apart or it can be a one-off thing that actually it might be the only time someone purchases windows for their home. So you don't know what you don't know in that kind of research journey, but you can actually have that conversation in chatbt. you can learn quite rapidly and if you've been having that discovery conversation you are going to move down the funnel faster into consideration stage into actually making an inquiry and getting a quote from a couple of different providers. So perhaps it is just the case that open AI thinking where are we successful where are we actually going to have influence and to get more people using the platform and ultimately more businesses investing in search strategies that include chat GPT. They want to just dominate the area that they can do really well and that may well be driving higher conversions through the funnel rather than actually closing the sale. I suppose Google has added in this additional functionality of the universal commerce protocol which is with you know a whole bunch of different merchandising like you know companies which I don't believe includes Shopify and perhaps it does um they're moving towards this as a retailer as an e-commerce store owner would you be putting your focus on trying to figure out how to get set up with UCP uh with this you universal e-commerce pro whatever called or would you be focusing on just that initial how do I just make ourselves more visible within the conversations that are happening or do you does it depending on the size of the company I definitely think it depends on scale if you are a very large marketleading e-commerce you need to be doing both large market leading e-commerce are not going to be sleeping on the agentic commerce side they know that it's the future and they are going to be wanting to hook it up now as as you should as they should be. They should definitely be thinking about that. Whereas I think a smaller e-commerce store, if you're turning over sort of around 10 million less, 20 million less, you'll probably need to be thinking more about how you actually take market share in your sector. And AI search is a fantastic way to do that. Actually showing up in chatbt results. If you can increase the conversion rates once traffic comes through to your site or you're getting your products featured in shopping, particularly if it is, you know, something like running shoes or uh paint colors, they're perfect examples of of the types of products that people would click through from TACBT results, then that's going to be a fantastic result that's more impactful short-term to the bottom line, knowing that things like Aentic Commerce are going to be made easier in the future for smaller retailers once they're actually rolled out at scale. that will just become normal in setups with Shopify, Magento, whatever type of e-commerce CMS you have operating. You and I were sitting on a call the other day discussing like e-commerce stores and looking through like some of their visibility within traditional kind of SEO within Google rankings and you'd see really large stores doing like really well organically but with very little to no visibility within the most basic of AI features like not appearing in AI overviews and I imagine they're unlikely to be appearing in chat GPT everywhere else too. Yeah. And I think it's a huge missed opportunity. One of the troubles with e-commerce is that there tends to be so much focus on paid ads on paid Google ads. That becomes the main strategy. The business then starts relying on it. Then you you are getting sales through it, right? It becomes the main lifeblood of an e-commerce business. But then you have the trouble of the rising costs which we all know costs are going up year on year when it comes to Google ads. We've got Google playing with the ads, moving them into different places in the search results page. So something that might have been a shorefire bet for you previously if you were doing well on your ad bidding then doesn't become and it can change super quickly. If a competitor gets investment and suddenly starts being more aggressive on certain keyword bids, well then they're suddenly holding the prime real estate. You're not. It super affects the funnel because you've just got all of your eggs in that basket. Organic on the other hand, there are so many different channels that you can be looking at and people are thinking about still the 10 blue links in organic and holding those top three. But actually so many search results, I think it's around 25 30% actually show AI overviews instead. And if you've got keywords doing well in organic page one, page two, there's a huge opportunity to push those into the AI overviews and then you would really start dominating the search engine results page. I mean, what you ideally want is to saturate it, right? You want to be in prime spots in the ads. You want to be in the AI overview and then you want to hold a high organic position as well because then it's your brand three times that's actually seen in those results. If someone has been searching top offunnel discovery conversations in chatbt and they've seen your brand there, then they go on to Google, they type it in and you're dominating that specific search result for what they're looking for. You know, some of your big products, some of your hero products, it's as good as closed. You've owned that whole funnel. And I think one of the reasons that e-commerce haven't been focusing as much on chat GBT is this preoccupation with Google ads and chat GBT becoming such a new organic channel that it's almost fallen by the wayside except for some very smart future focused forwardinking first movers in e-commerce who are like we've got to be dominating with our products in chat GPT. Somebody's we're recording this live and somebody's asking about do we really effectively need to be doing super well in three different domains with organic keywords with AI and ads and I think you answered that just then but it makes me wonder if I'm a marketing manager or marketing director or I'm responsible for a large ecom and perhaps it's a retailer whether it's an omni channel thing I only have so much resource I can probably dedicate a bit of time to some of these. My problem is scale. Like should I be focused on thoseformational kind of chats, thoseformational searches that you just highlighted, or should I be focused on the product pages? Because I could be dealing with a 10,000 page website. That's a lot of different pages to then optimize for chat if I have to go about this in a very particular way that may be different to what I was doing with SEO. If there is a different way of doing it to SEO like where would I focus my time on thatformational search? Do I just focus double down on blogs? Do I or do I just only address the product pages? I think this is the hard truth for marketing managers who are managing at scale is there is just a lot to do and it can feel really overwhelming and you need to have great product pages because ultimately if someone hits your product page and they don't convert there's no point driving traffic there. So your product pages have to be great. The way I would think more towards organic search results and chat GBT and other AI search platforms is what are your highest profit products? What are your hero products? What do you definitely need to be dominating? If you're working with 10,000 different products, you're thinking how am I getting 10,000 different products into Chachi PT? Don't start there. Think with how am I getting this specific group of products that have really good profit margins for the business into chat GBT. That's my test bed. I want to own the organic results. We've already got great paid ads running for them. Let's own that top and middle of funnel that happens in chat GBT conversations as well. And to do that, you are going to have to do some informational content still, but you're also going to have to do a lot of PR, a lot of digital PR work, a lot of reviews. you hopefully got UGC going already if you're in e-commerce and running ads. So all of that is going to add to actually doing well in AI search. So just coming back to the the the news that we started with with uh CHBT um kind of pulling back as I phrased it on CHBT. Do you think that they're pulling back long term or do you think this is just a temporary kind of thing? I think this is going to be temporary and I think some of it has perhaps just been a reaction to uh Shopify because the Shopify president said that you know only a dozen Shopify merchants are actually using the AI tool integrations even though they support them. And I think this is probably just a short-term kickback where OpenAI are focusing on Chat GBT on what they think is going to be the highest success rate, but over time I think this will be back. And uh before we finish off for this week, um last week you were speaking at E-commerce Scotland and uh by the account everybody I spoke to while I was there absolutely loved everything you had to share. What are the main kind of takeaways you took from that event that would be useful for people to know about and be aware of and maybe consider to put into action for the rest of 2026? Firstly, that was such an amazing event. It was an incredible event with a really, really nice vibe. If you work in e-commerce marketing, I highly recommend looking up E-commerce Scotland because it's such a great day event. Really, really lovely team. Um, I had the pleasure of speaking about how to dominate in AI search in 2026, which is of course one of my favorite topics. Uh, and for e-commerce, a lot of my takeaways were around making sure that your website is technically, and I'm talking about technical SEO, but I'm also talking about things like your Cloudflare setup actually set up to be showing in AI search results. There are some big brands, very very big brands we have seen who were accidentally blocking crawlers from chat GPT and other AI tools, meaning that their content was never going to be surfaced. They were not going to be showing up in AI search results at all. And you think that that's perhaps a mistake that only small businesses would make, but it is not. So, making sure that you've got your technical set up, making sure that you have a very clear positioning content strategy that is focused on AI search and what you want to show up for and making sure that your digital PR is supporting that as well. Those were probably the three main takeaways from my talk. Superb. And if you'd like Charlie to come and speak at your conference or your event and she does uh you know all kinds of you know small events if you need somebody for hen party to do a little speech as well. I'm sure she'll be able to come along and talk about instant checkout too while she's there. Um you can go to exposure.ninja/charlie hyphen marching speaker. We should probably change that to just slash Charlie. [laughter] Thinking about a long one. Yeah. I'll make sure I drop in the description as well. You can reach out to Charlie and get her booked in for your um conference or event later this year. Any final thoughts before I uh finish this one this week? Just to say that my next event I'm going to be in London on a search panel for an event called Sip and Search London, which is a fantastic event that I absolutely recommend. So, if you want to come along, I would love it. I would love to say hi. Amazing. It looks like a fantastic event. I saw some photos and videos the other day. Looks like a lot of fun. So, I encourage everyone to give it a try. Super. Well, thanks everyone for joining us this week and we'll speak to you next week as well. Take care. Bye bye.