Latest / The Search Signal / Google Universal Cart What Does It Mean for Ecommerce Brands?

Transcript

Google Universal Cart What Does It Mean for Ecommerce Brands? · The Search Signal

• michael: think about what Google sees with this cart. It sees every product a shopper saves from every single store, and it sees every price that they're watching. The store, on the other hand, only sees what gets transferred to its site or what ultimately gets bought. So the more of the shopping journey that happens inside of Google, from comparing all the way to checking out, the more Google knows about what people are weighing. And the more of that journey it can put ads in front of. Hey, welcome back to The Search Signal. I'm Michael Transon. I am the founder and CEO of a search marketing agency called Victorious. And here at The Search Signal, our goal is very simple. We want to help marketers understand and then take action on what is happening in the quickly moving world of search marketing. We ourselves have been running campaigns for nearly 13 years. We also work with some of the biggest and also some of the fastest growing brands that are out there today. So we bring what we learn from those experiences with our clients, plus a lot of first party research that we perform ourselves to The Search Signal to help you with your own work. It is a very, very busy week in the world of search marketing. So let's get into what happened this week in the news. Okay, we are starting with something that affects pretty much everybody who uses a rank tracking tool. So since the middle of September, Google, we've learned, has been serving some of these rank trackers complete and normal looking result pages with the listings completely swapped out. Nothing on these pages look wrong. The rankings are just simply not the ones that a real person would see if they did the same search. A company called Nozzle, which makes a rank tracker, had studied about 18 million of these and had found that about one in 12 had been altered. And also in late September, they saw that it was close to a quarter. So Google has not commented on this yet, but the likely target from my point of view is AI companies and data resellers, and the ones that are scraping Google's results to feed their own answers in their own software. And some of these SEO tools and rank trackers are getting caught right in the middle of it. So what does this mean? When your ranking reports from late September and October show drops and they don't match your traffic, you need to go to Search Console before you make any changes or before you report any losses to anybody. Now, Search Console is Google's own data on real searches, and because of this, right now I would trust it over any of the third-party trackers you might be using. It also, in my opinion, makes sense to ask whatever rank tracking vendor you use how they are detecting swapped pages because I would hope that the good ones are going to have an answer for this soon. Also, I would say the altered data from our POV has a bit of a pattern too. A keyword will drop pretty hard for a couple of days and then it will pop back to where it was, and your traffic probably didn't ever move. Now with that said, the worst of this seems to have landed between late September and early October. So if I was you, those are the weeks that I would recheck before you start building any sort of story around things changing that is not actually true. For the next thing that happened this week in the news, let's talk about a bit of a measurement problem on the AI side of things. So after OpenAI released GPT-6 and that was earlier this month, we found that links in the ChatGPT answers stopped carrying the tag that tells your analytics that a visit came from ChatGPT and they stopped passing a refer. So what that means is those visits are landing in your reports likely as direct traffic. There was an analyst that I saw from Adobe that looked at about 500 pages where ChatGPT was sending a meaningful share of visits. And on those pages, the ChatGPT referral traffic fell by about 31,000 visits a day on those 500 pages, while direct traffic rose by about 26,000 visits a day. Now, this is fixed. OpenAI fixed it on the 15th of October. So this was a bug, but we got a lesson here that's gonna outlast that bug, When your AI traffic suddenly drops, we need to look at direct traffic on those same pages before we assume anything has changed with our visibility. The fact is, measurement for AI search still breaks into ways nobody knows or announces every single day. So we have to confirm the tracking is working before we start to react to any numbers. So to check your own data, I would just take the pages that usually get ChatGPT traffic and then compare their direct traffic, you know, maybe with the week before, If direct went up by about what ChatGPT has lost, not to worry. Those visits did not go anywhere. I would say only the label has changed for that traffic. Okay, let's talk about testing ads in AI Mode. So Google has released a feature called personal intelligence that lets people connect apps like Gmail to their AI Mode. And this is so that the answers can be more personal, right? And in March, Google had said there were no ads for people who connected their apps, and that that was not changing right now. Now, last week, Google suddenly confirmed it's now testing ads there. Its help page now says that ads won't be based directly on your Gmail inbox or your photos library if you connect that, but summaries and inferences from your connected apps may be used to show you ads. Now, my take has been the same this whole time. I've always said this: ads always arrive. Google said no ads in this experience in March. And then seven months later, it's now testing them, The point is when Google opens a new surface without ads, we gotta treat that as a window and building our organic presence there while that window is still open. Now for advertisers, this also means ads in AI Mode can reach people whose answers are personalized from their own connected apps, which could make that targeting a lot more precise. Now Google says it won't use sensitive categories like health or religion to personalize these ads. And on top of that, you can disconnect your apps at any time. Staying on the AI Mode narrative, let's give another update on what happened this week. Google is also testing the left sidebar from AI Mode on the regular search results page. Now that is the small Google logo and your chat history and a button to start AI Mode chat and settings all sitting next to the normal search results. Now, tests, as you know, that doesn't mean that it's gonna ship, but every one of these little tests moves the regular results page a little bit closer to AI Mode. And I've said this before. I think that's the direction that regular search is heading. It's just doing it one piece at a time. And the last story from this very busy week is Google's Universal Cart, which went live to the US on October 15th. It is one shopping cart that works across Google search, including AI Mode and the Gemini app, where shoppers can save products from different stores and then can check out without leaving Google. And that is what today's episode is about. So let's dive in. So I spent a lot of time with the Universal Cart before recording this. I was adding products from, you know, different stores, getting to check out and just to see where it worked and then also where it didn't. So I will walk you through today what launched, whether this is quote unquote agentic shopping, who it matters for, and then also why Google wants this to be as seamless as possible. We'll talk about what Google reads when it is comparing products. And then also how to set it up before Black Friday, because that's where I think this needs to go next. So let's start with what actually got launched. So Universal Cart went live on October 15th and it was in the US. It works in Google search. It also works in AI Mode and in the Gemini app. And like I said before, this is the same cart in all of these places. So you can, for example, save a pair of shoes while you are searching on Google and then also add a sweater later while you are chatting with Gemini and they end up together. Now this cart does a few different jobs at once. It watches the products that you saved and it also tells you when a price drops or when something comes back in stock. It also shows promo codes. So it connects with Google Wallet. So it can suggest which of your cards will give you the best perks on that purchase. And then it also saves your progress. So you can come back actually days later and you can pick up right from where you left off. Now, checkout. That works in two different ways. So for some stores, you can actually buy right inside the cart with Google Pay. And then for others, Google is going to send your cart over to the store's own website so you can finish there without having to actually add everything again to the cart. Either way, the store is going to stay the seller of record. So Google has named Nike, Sephora, Target, Ulta, Walmart, Wayfair, and Shopify, I think. Those are the stores that were part of the launch. Now, when I tried it, I used Ulta because that was a really clear example. So I found a fragrance kit in Google Shopping, and there was a buy now button just right there. And I could have paid with my Google Wallet without ever having to leave Google itself. And now Google says that people shop around Google and use Google more than a billion times a day, now I will say this. Google has used roughly that same number since like 2021. So it hasn't told us whether shopping on Google is growing. But what has grown is the catalog behind it, which has more than doubled since 2021. So the question becomes: like, what's different from before, Google Shopping has shown products from lots of stores for many, many years. But every click sent you off to that store. And then you built a separate cart on all of those websites. But now there is just one cart across all of them. And for some stores, the entire purchase can happen just straight on Google. Now, a couple things from Google's preview back in May when they talked about this that didn't make it into the actual launch we're seeing right now. So far, I can't tell whether or not YouTube and Gmail, I don't think those are connected yet. And the launch announcement doesn't mention price history or loyalty program info. And those were both in the preview from May. So I would say expect those to happen even if they're not here yet. But that brings up the question a lot of people are asking about this launch, which is Is this agentic shopping? There's a popular take right now that a launch like the one we're talking about means brands have to start selling to AI instead of to people. I personally don't think that is where we're at today. So let's start by talking about how Google actually describes this. In its launch for the Universal Cart, Google splits its shopping work into two different tracks. The first one is agents that act on your behalf. And then the other is tools that give shoppers confidence and gives them control. Now Google put Universal Cart on the second track. And that is a bit of a change from May in the announcement when Google had introduced the cart as part of the foundation for agentic commerce and Nick Fox, who runs the search and commerce at Google, has said publicly that he doesn't subscribe to the idea that everything is going to be, you know, agents just talking to other agents. Now, in practice, a person taps buy in this every single time. Even Google's separate buy for me feature, which you can purchase something when it hits a price that you set. Asks you actually to confirm the purchase and all of your shipping details before it actually makes a transaction. So when I had asked for Nike shoes I had used the AI Mode chat for this, I got a loading screen that said it was connecting to the Nike store. And then it said it was checking Nike's best prices, and then it got me to the checkout. Now I did not need a Nike account to do this, but I still had to, confirm this actual purchase myself before it was completed. If you recall too, ChatGPT actually tried something similar to this last year with checkout inside of its chat and shoppers there also had to confirm every purchase. Now here's the kicker: it didn't work. Walmart said that purchases inside of ChatGPT converted at about a third. Of the rate of purchases where the shoppers clicked through to Walmart's actual own website. And in March, OpenAI actually pulled back on that to focus on instead just helping people find and discover products. Now Google is starting from a much different place, People have shopped with Google and Google Shopping for years, and they trust it with things like their payment information and When a store prefers it, the cart can actually hand the purchase, like we talked about, to the store's own website. And here's the reality of the market too. Shoppers are saying the same thing, right? There's a company, Worldpay. If you're familiar with them, they're payments company. They had surveyed people in a couple of different countries this year. And among the people that were actually open to shopping with AI, more than half. still wanted to approve every purchase themselves. And now Worldpay, their model benefits when more shopping runs through AI. So a result that goes through the interest of that company, in my opinion, carries more weight, and it's probably true. So my view on this is that, you know, right now, people are still the ones that are gonna be deciding what they want. Nobody's telling Google to go find them a generic thing and just, you know, go buy it. They are actually still deciding what they want. And the AI is helping them compare options and get to the checkout faster. Now that might change. The technology for fully agentic shopping already is there. It already exists. What's missing is whether people trust it enough to actually use it. And my opinion, if shoppers adopt Universal Cart this holiday season, I think that that is a strong sign Google might be moving toward buying actually on your behalf. And the quote unquote sell the AI idea, you know, that came from a vendor interview, right? They made a really fair point inside the interview, which is that when products look the same, an AI is gonna like break the tie on things like price and delivery speed and you know, reviews or availability. And brands with a clear reason to be chosen still have room. That interview, the way that he put it was that this is really about helping a product get chosen. That's what this does, helps a product get chosen. And I agree with that part. But the shopper doing the choosing is still a human and not an AI. So let's talk about who this whole change actually matters for. So Universal Cart matters most for purchases people make quickly. And in my opinion, at prices that they just don't really agonize over. This would be like beauty products or apparel, household basics, I don't think anyone's buying a $10,000 item with a couple clicks from a cart that they found on Google. But it also matters a lot at this time of the year, right? One of the worst, let's be real here. One of the worst parts about Black Friday is just figuring out where the deals are at. Now, a cart that watches prices across stores and shows all of the promo codes in one place really, really helps with that. Now, the launch partners around Universal Cart right now are also big retailers. So they are going to be the ones that are benefiting first. With that said, if you're on Shopify, this is important. Shopify enrolled its eligible US stores by default, which is going to give smaller brands on Shopify a way straight into Universal Cart from day one. And there is a, in my opinion, a much bigger shift that is underneath all of this. Google Shopping organizes everything by product first and by store second, I was talking about fragrance kits earlier. So you pick a fragrance kit, and then after you pick the product, then you see every store that sells it and at what price, and oftentimes you'll see how far they are away. And it'll show you whether they're in stock and if you can buy it right there, right now. So, in a lot of ways, Google is building what another company built, Amazon. One place to compare everything with just one checkout. Now, the difference is that with Google, each store can keep its own website. And when the cart transfers, the store keeps the customer too. So for a brand that has always had to choose between selling on Amazon and actually owning the customer relationship, this is a very, very meaningful difference. Now, separately, if you sell services or if you generate leads, I will say the cart doesn't matter or change much for you yet. Reason why I say that. Google builds these shopping tools because buying is where the ad money is. When Google does something similar for services, like the feature where it's that exists now where AI can call local businesses to check on prices and availability for us. I think the reason is the same. More of the search stays on Google, and the more it stays on Google, the more it carries Google ads, which will raise the next question around Google's motives, which is why Google wants this whole Universal Cart experience to be seamless as possible. If you don't know this, Google doesn't charge merchants or shoppers any fees for Google Pay. So what we can learn from that is the money for Google isn't in the actual transaction itself. It is in ads and also in keeping that shopping experience and journey on Google. So when I searched in Google Shopping, sponsored products and sponsored stores, those sat at the very, very top, AI Mode already runs offers that are labeled as sponsored deals. And like we talked about earlier in the news, Google is now testing ads in AI Mode for people who are connecting all of their apps to it. Now, Google hasn't decided whether the ads yet will show up inside of the cart itself. I was talking about Nick Fox from Google earlier. When he was asked this, he had said merchandising at the cart stage is quote unquote interesting for retailers. And that how it evolves is from what he said still to be decided. But think about what Google sees with this cart. It sees every product a shopper saves from every single store, and it sees every price that they're watching. The store, on the other hand, only sees what gets transferred to its site or what ultimately gets bought. So the more of the shopping journey that happens inside of Google, from comparing all the way to checking out, the more Google knows about what people are weighing. And the more of that journey it can put ads in front of. That is why Google wants to be as smooth as possible with this rollout. And now, you know, stores used to see a lot of that themselves. When a shopper got to your site, you could see. in analytics, what they looked at, how long they stayed, what they put in the cart, and what they left behind. Now, when comparing happens inside of Google, you don't see, you know, which other products they had considered or what other stores that they checked. And you cannot tell whether a price drop at a competitor, for example, is why you didn't get the sale. You just see the end of the journey if you do get to see much of it at all. So For your brand, that means the comparison that consumers are making is happening somewhere you cannot fully see. And what Google knows about your products is going to decide how well you hold up in that space that you do not control or see. So let's talk about what Google does read when it compares products. So the cart is built on Google's shopping graph, which Google describes as more than 60 billion product listings and they are refreshing it all of the time. A big part of what feeds it for your products is going to be your Merchant Center feed. And that is your prices, that's your stock levels, your shipping, if you've got a return policy and all of the reviews for your business. Google also added six optional fields to Merchant Center earlier this year. And I think they call them conversational attributes. And they include, from what I've seen, question and answer pairs about your products and related products like accessories or substitutes, and then also how popular an item is within your own catalog. So Google says they are mainly meant for conversations in AI Mode, and Google hasn't said that they can affect the cart directly. And it hasn't necessarily promised a ranking boost inside of carts or shopping, but AI Mode and Gemini are where a lot of products do get found and saved in the first place. So the way that I would think about your product data is the way, you know, way I would think about it if I was like a good salesperson. The shoppers got a lot of questions. Does your product run small? Does it work with what they've already got at their house? How quickly can you ship it? What's the return window on it? A good salesperson is going to be able to answer those questions before the shopper even has to ask them, And a good website does that too. Your product data should do the same thing because when you know Google lines your product up against three different other competing products, that data they have is gonna be your only salesperson. And your product pages on your site honestly do matter a lot here too. Google reads product information from your website, including the structured data that marks up things like your price, availability, your shipping, returns. And they check that information against your product feed. And when a cart transfers your product page is where that shopper is going to land. So the point being is that page needs to answer the same questions and importantly, show the same price, show the same price as your feed. So a couple of places that I would start. First, write questions and answers that pair for your best selling products is the first thing you do. You probably already have the questions. Go check your customer service inbox, go check your reviews. Then you can mark your related products. Like if you sell a case that goes with a phone you sell, or if you've got a product that has a refill that goes with the starter kit. Those are good examples. I would also rank your catalog by popularity. Just use your own sales, because as you might be aware, Google can't always tell your best sellers from everything else. And then make sure the price and the stock in your feed match your site exactly because. This is important. A mismatch can actually get your product pulled from Google Shopping altogether. So let's talk about the practical part. With all that being said, specifically, let's talk about how to set this all up before Black Friday. Black Friday is November 27th, which is about according to my calendar, about five weeks from now. So, at least from this episode's recording. So, my recommendation here is you gotta move fast. And I would do all of this before then if you can. We just talked about it. Clean up your product data, turn on checkout. But this all starts in Merchant Center. Okay. If your products are not in Merchant Center with accurate prices and stock and shipping and return info. Nothing else in this is going to work. If you are on Shopify, like we talked about earlier, this is very easy. You could just look under the sales channels and then look in Agentic. Now Shopify has already turned on direct checkout for eligible US stores. So you might even have this turned on without realizing it. But from there, you can choose your path. You can choose checkout on Google, or you can choose the cart transfer for your website. Now, if you're not on Shopify, the setup lives in a checkout hub inside of your Merchant Center, which Google is rolling out gradually in the US right now. So your team, or if you've got a platform, your platform provider connects your store to Google's Commerce protocol. And you have to meet Google's onboarding requirements before either of these things turn on. So checking out on Google and transferring the cart to your site, those are set up separately. So you can start with one and then later on you can have the other. Now the two paths here, let's talk about them because they are a real trade-off. Okay. Checking out on Google using that option means that there are going to be fewer taps for the shopper. Cart transfer keeps the purchase and the customer relationship, on your own website. I'm not going to tell you what to pick, but whichever you do pick, we need to set it up all the way. For me, when I had tested this, I had two stores that let me save items to the cart in Google Shopping. But when I clicked through, I just landed on a product page and I had an empty cart and there was no tracking tag on the link. And that is what happens when cart transfer is not set up. And that is a shopper who has already, like my example, decided to buy from you. And they show up to your website and they have to start completely over. Then after that is done, I would set up your measurement because there are some gaps on this right now. So Merchant Center does have reports for checkouts and for cart transfers. So I would definitely take a peek at those and see them. But I would also, if possible, add your own tracking to the link that Google uses to send cards to your site. Or those sites might not be labeled correctly inside of your analytics. Now on Shopify, purchases that are made through direct checkout do not show up in Google Analytics. So you will need to look at Shopify's own reporting for now. And last thing I'll say about that too is Google doesn't give you a report for how many shoppers save your product to their cart and never bought it. So that's something that you need to take into consideration in your decision. So before Black Friday, I would definitely test it. And I would test it the way that a shopper would test it. So just go, you can even do this right now. Pop up in a new tab, search for one of your best sellers in Google Shopping, or you can do AI Mode or Gemini, save it to the cart, and then go all the way to checkout. You can even just do one from regular search and then you can also do one from AI Mode chat if you want to try both. Check that the cart arrives full and that the price matches your website and the link also carries your tracking, and that nothing is asking your shopper to start the whole process over. This only takes about 10 minutes, and it's how I found every problem that I just described in today's episode. Which all of this leads me to where I think now that we know what's happening right now. Where I think this is all going to head next. All right, here're my predictions. And you know, for each of them, I'll give you a reason why it might not happen also, because as we all know, it's really easy to overpredict, especially with something this new. So, first, I think you'll be able to buy inside of AI Mode chat. You can just say, buy it, and it'll ask which card from your wallet, it'll ask which address you want to ship it to, and it'll probably, I don't know, confirm with a passkey or use your fingerprint or something. Now, the reason why that might not happen, buying from any browser where someone is signed into Google can create some real fraud and security questions. So that could potentially slow it down. The second thing I think will happen is I think the cart will show up in regular search results alongside the normal links. The reason why it might not work, Google does still change its main results. generally carefully, even while it is testing actively AI Mode's layout in that same surface. Third, I think ChatGPT is going to come back into checkout. Now the reason why that might not happen, that first attempt it did was bad. It converted very poorly. And you know, when you take a peek at that and you see what happened, you know, AI moved towards discovery after it didn't work. And they ended up letting retailers run their own checkout. So that might not happen, but I think it will. The fourth thing I think might happen, this is the big question, in my opinion, whether Amazon and like Costco join. If they did, Google would start to look a lot like Amazon, which is the one place to shop across everyone. Now, the reason why it might not, obviously, Amazon has very little reason to send its shoppers through Google's cart. It's sending them to a competitor. Then fifth, which I think is happening soon, I think ads will end up inside of the actual cart itself. The reason why it might not, Google has said that it hasn't decided. And I don't know. Ads at the moment, someone is about to buy, you know, could cost Google some trust. But I think there's a good chance. And then sixth, and this is the last thing, I think that Google eventually nudges shoppers towards checking out on Google. over transferring the cart. The reason why it might not, I don't know. stores, you know, want the customer relationship. And, you know, you can look at Walmart's experience with ChatGPT and that suggests that, you know, shoppers still like finishing on the store's own website. But the thing that I would watch the most through all of this is adoption. If shoppers use this cart over this upcoming holiday season. Expect, in my opinion, the rest of these predictions to come a lot faster. But taking a step back on this whole thing, Universal Cart is just simply a layer on top of how people already shop on Google. People are still the ones deciding what they want. And the AI that now exists is helping them, like we talked about earlier, compare and get to checkouts faster. And the brands that do well in it are going to be the ones whose product data answers a shopper's question and whose checkout works when someone actually shows up ready to buy. The technology to buy on your behalf already exists, but the trust with people that is the missing piece. And if people start trusting it this season, I think fully agentic shopping is going to get a lot closer. So this week, my recommendation for you is fix your product data and walk through your own checkout the same way that a shopper would in Universal Cart. Those two things are going to decide how you show up when the shopping season does start. That's today's episode on Google's Universal Cart. Thank you for listening to this week's episode of The Search Signal. If this was useful, please go ahead and subscribe wherever this episode does find you. And that is it for me today. I will see you for the next episode next week.