EP 240 - Inside Google Local Rankings and the Review Purge
Claudia Tomina joins us to break down what the Google API leak reveals about rankings and the role of business name and categories, why reviews are really disappearing, and how AI Overviews, Gemini, and AI Mode are reshaping local discovery and conversion.
Why does adding one word suddenly put a business in the top 3—and removing it makes them disappear? Claudia Tomina explains how Google’s local ranking system prioritizes one-to-one matches between search queries, business names, and primary categories—and why business name is still the cheat code. And why are reviews still disappearing? Tip: It's not a bug.
Sign up for our new Legal Marketing Newsletter. Once a month, no fluff, just good ideas.
The Podcast Deets
- Local Rankings & Business Names (00:00–10:45)
- Reviews, Enforcement & Risk (10:45–21:40)
- AI Overviews, Gemini & the Future Funnel (21:40–42:30)
Key Takeaways
- Business names and primary categories outweigh most other ranking signals
- Missing reviews are due to a Google algorithm update NOT a bug
- Review volume and patterns matter more than intent
- AI interfaces are becoming the final decision layer
- Google Business Profiles are now the last mile, not the first
👇 Watch by topic:
00:00 – Introduction & Google API Leak
08:40 – Business Names, Categories & Brand Searches Role in Ranking
10:45 – Review Removals, Volume & Enforcement
21:40 – AI Overviews & Local Search
32:45 – Gemini, AI Mode & Agentic Shopping
40:30 – What Marketers Should Do Now
Related Links
Claudia Tomina Linked-In Profile
Watch for Claudia's article on business name in ranking

Interested in sponsoring this podcast or our newsletters please reach out to mblumenthal@nearmedia.co
Full Transcript
Greg Sterling: Hey everybody, welcome back to the Near Media podcast, the Near Memo with Greg and Mike and today Claudia Tomina. Did I pronounce your surname correctly? Tomina? Tomina? Tomina, okay, good. I wanna always make sure that I pronounce people's names correctly. I hear it in my mind as Tomina, but Tomina is the correct pronunciation. Okay, good. She's the founder of Reputation Arm and she's a local SEO, pretty prominent local SEO who's...
Claudia Tomina: Tomina, Tomina
Greg Sterling: posting a lot, writing a lot, kind of making the rounds on podcasts and here to talk to, and she's been on this podcast before a couple of times, I think, at least once. I mean, you've been, yeah, and you've been in our office hours. So this is your first solo outing with us. So we're really gonna put your feet to the fire here, I think. So thank you for being with us and welcome Claudia and...
Claudia Tomina: Yeah, on a panel. Yes.
Greg Sterling: I want to say you've got a piece coming out in Search Engine Land where you're talking about local ranking on Google and some of the things that you have discerned, learned from your research into the Google API link, leak rather. Why don't you tell us about sort of the high level takeaways from that, not to preempt them or anything. And then we can get into it. And we also want to talk about reviews in the future. the interface and local AI stuff. So we'll be getting to all of that.
Claudia Tomina: Yeah, OK, so thank you for having me. Appreciate it. Episode 240. So. No, no, I think you did. So yeah, I wrote this article because I do like to explore the API leak and I think you can find a wealth of information and kind of just validate some of your theories at the end of the day. We don't always know how these algorithms are working, but we have a good idea and case studies are always great.
Greg Sterling: 240, did I not say that? Thank you for saying that. Okay, all right.
Claudia Tomina: So I've been looking into specific search queries, like why can somebody rank for this and then add this one keyword in front of it and all of a sudden they're ranking in top three. You remove it, make a little bit of a broader search, and then you're eliminated. So I started looking at the API and from my understanding, Google obviously is looking for a one-to-one match. So when when the search query can match, the category and a token in the business name, then obviously you're going to become more relevant and you're going to rank. So everybody knows that business name is your cheat code and to entering into the SERP. So the idea was, what happens when there's a search query that doesn't really match a category? How is Google looking at that? And so that's where business name really becomes a priority, but business name is a priority regardless from. From what I understand, I think it's the top layer. And then when there needs to be a tiebreaker or Google needs to validate something, they're looking at your category. And primary category is going to be the key to whether you're going to show up in the top three. If you have a secondary category, you're less likely to be that one-to-one match for Google. So yeah, I mean, it's interesting to see. I think when you have a broader search term like restaurant in general, then there's a lot of other factors that come into play. Reviews, behavioral clicks, all that kind of stuff comes more into play. But when the search is in the name or the category, it has to be extremely relevant. So I think for the most part, people have to realize, don't chase every keyword. You're not going to rank for everything. sometimes you have to look at other channels to promote certain services, and then become the brand search later on. So don't waste your SEO dollars on trying to rank for 20 or 30 keywords in Google Maps. It's a waste of time.
Greg Sterling: Well, so what you just said about promote your business on other channels and become the brand search, I think that's the ultimate strategy. But does putting keywords in the business name detract from that at some point? Like if you're thinking about long term building a brand, building your, I'm not disputing anything you're saying about the way Google thinks about the business, but if you're trying to build.
Mike Blumenthal: You think Plumber San Francisco is not a good brand name? Is that what you're saying?
Greg Sterling: Precisely. Best San Francisco plumber. But people have been doing that. I mean, the keyword stuffing advice has been pretty prevalent and people have been doing a version of that. But if you're going to build a brand long term, maybe this is what I was trying to think of when we were in the green room. It doesn't make a lot of sense to have a generic business name like that. I mean, what would you say to that,
Mike Blumenthal: There you go.
Claudia Tomina: I mean, so, okay, in the article, there's a specific example, and it's for a, I searched smoothie near me, and the second listing in a local map pack is a smoke shop. So everyone thinks like, your category, your category, but there's not, it's not a juice bar, it's Smokies, smoke, and smoothie. And just the fact that the smoke shop, and, has pictures of smoothies and has smoothie in the business name, allowed it to over rank in a business that's completely more relevant. So you just have to look at every specific search query. It's like super unique. That's my point is every, every search query is going to give you different results and maybe weigh different rankings. But the idea is business name category. And I think domain URL is another huge one. I mean, it's not a mistake that every garage door has garage door in the name, garage door is the primary category, and garage door in their URLs. Like, they know what they're doing.
Mike Blumenthal: But to answer your question, Greg, I would say that Google has driven this behavior in the marketplace. There's this dialectic between ranking and degradation of unique business names. And I think a business to achieve both effects, which is the ability to have your brand stand out and stand out at Google, needs to thread a needle between a creative name and a keyword name.
Greg Sterling: Of course.
Mike Blumenthal: And I think that it's doable, but I think it's, I think people don't give it enough thought because they think that Google delivers so many customers. And the reality is we've seen in our legal search that outside branding probably delivers more customers. It's just that Google is so good at claiming last click attribution. So I would say to Claudia that, that the answer to your question or to you is some combination. something between a unique name and a keyword that is memorable.
Claudia Tomina: But what other way is there for Google to do this, right? If somebody wants to search a name, they have to show the knowledge panel. They have to match the business name.
Mike Blumenthal: Right. This is their way of discerning brand and it's a, it ends up distorting the market. I'm not saying it's wrong. I'm just saying it is. I mean, it certainly.
Greg Sterling: Well... So I have another example. I think the rationale that you've articulated is right, that Google is trying to do brand matches as much as it can. And when we did some research in Europe with travel, and what we saw is that booking.com is obviously a huge travel site. and especially in Europe. And when people would put booking in the query, it wasn't necessarily a brand search that they were doing. I mean, you could tell that because we watched the videos, but Google was interpreting it as a brand search. in almost every case and Booking.com would show up in that number one position consistently and they also were buying ads. So they had ads, then they had the number one organic position. So Google was biased towards sort of showing a brand when they could infer that it was a brand query, but they were over inferring that because it wasn't always the case that people were looking to navigate to Booking.com.
Claudia Tomina: And the same thing happens on Amazon with product searches. It's the same idea.
Greg Sterling: So is there any tactical advice there?
Claudia Tomina: I mean, would I just like...
Greg Sterling: Beyond keyword stuffing
Claudia Tomina: I just think everybody needs to kind of evaluate the search query and then not get so worked up when the rankings aren't going in their favor. And then look to the other channels, you know, especially like, like I give the example of the Med Spa, you know, they provide weight loss drugs or IV therapy, but they're not going to be able to compete for IV therapy because it's a secondary category and there might be too much competition in the area with that business name and that as a primary category. So, so move on and And if you want to build out that service, do it on social channels and become the Brand Search.
Greg Sterling: Yeah, somebody was remarking yesterday, I don't know who it was that, or maybe it wasn't yesterday, but I saw yesterday on LinkedIn that Instagram searches were influencing Google rankings. know, that information on Instagram was showing up in Google rankings.
Mike Blumenthal: Well, it was actually, I think you're referring to the one where the Instagram search was interpreted by Google as an event and was included as a justification in the display. So that though is a known feature. just, was, what was different about this post was that Google used the event information or the daily special information, which is what Google interpreted it as, as a justification, which justifications typically don't
Claudia Tomina: Yes.
Greg Sterling: Yeah, yeah, yeah. Right.
Mike Blumenthal: as you'll at least don't influence rank, but do influence conversions. And it's Google's way of saying, here's what we know that might be relevant to your queries, this about this place. So, we're, I mean, Claudia can speak to this more, but tying your social in with your profile at Google will increase the amount of secondary information that Google has about things like your daily specials.
Greg Sterling: Don't impede Drake King's Re...
Mike Blumenthal: and can certainly help with conversions. I mean, if you're looking for liver and onions special on Wednesday, which, it is still on the menu around here just as a special, just as a note, I use it because people respond to it. you know.
Greg Sterling: I'll never be looking for liver and onions. That'll never be a search query that I enter into Google. Yeah. My deceased relatives would be very excited by that, but not I. All right, so let's talk about reviews. You've quite a bit over the, as we all have, quite a bit over the last year, there was just a lot of kind of chaos and confusion around reviews, Google taking down reviews, people trying to get reviews reinstated, a lot of discussion of review fraud. We had the GMB API. folks on and they were talking about the patterns in review removals. What are your takeaways from the last year, in this review management world? And you wrote about the punishment of people who were soliciting reviews at the point of sale.
Claudia Tomina: Yeah, that's the main issue is these businesses are asking for reviews at checkout. Google's tracking that the reviews were left inside the building. And then when they do their audits, they're removing, I believe they're removing all the reviews that were left inside the building. This is why I say that.
Mike Blumenthal: You don't think it's a function of just sheer volume?
Claudia Tomina: So I think the volume is what triggers the, you know, the flags. And then from there, they're looking at the patterns of where the review was left, because I had a client that all the negative reviews stayed, but all the positive reviews were removed. And my thought is, well, somebody left the building and went home and then left the negative review later. So that's why those stuck, but all the positive ones got removed because they were being left after the point of sale. So you're less likely to be inside somebody's business and leave a negative review. So why did Google leave those, right? That's my theory is that.
Mike Blumenthal: Well, there
Greg Sterling: Well-
Mike Blumenthal: could be, as we've seen, are now the ability for users to report incentives. And if there were incentives involved with this, even subtle ones, that could be another confounding factor. It's a little hard to know for sure. But volume is certainly a trigger.
Greg Sterling: It seems like Google is inferring some bad behavior or unethical behavior on the part of the business if the review is left in the facility or in the restaurant or whatever. But it's not necessarily the case.
Claudia Tomina: Well, I think they're looking for patterns. And so when they see these huge spikes, that might trigger an audit. That's one theory. then when GMB API showed us that there is that local map question of, did this business incentivize reviews? I mean, I'm sure that's triggering something as well. So I started to think, well, was that the trigger or was the trigger
Greg Sterling: Yep.
Claudia Tomina: the actual like huge spike in reviews. But both both are present, right? It's just it's interesting. At the end of the day, you know, this was not the most businesses think like, well, I'm I'm not giving away anything for free. Everyone I talked to, the handful of customers that this happened to, they all claim that they did not incentivize or give away anything for free, that they just were asking for the review at the counter with a QR code. But how would Google know?
Greg Sterling: And well, that they're inferring from presence inside the building, right?
Claudia Tomina: Correct. Exactly.
Mike Blumenthal: Yeah. Although Google's ability to discern location isn't as, I it's only, it's about a quarter of a mile. And so it might or might, I mean, to me, volume is a much Occam's razor would tell me that a simpler solution for Google, a less costly solution would be to flag them for volume. And then perhaps as Claudia pointed out, review them for location and see, but it seems more likely that volume is a prime.
Claudia Tomina: you-
Mike Blumenthal: me, more likely that they would invest in a very simple, too many and too short a time in this category.
Greg Sterling: It's. Right. So velocity is another factor here. It's like the concentration. Yeah.
Mike Blumenthal: I would think, I mean, it's just a simpler factor to track.
Claudia Tomina: And I think the way that the review was left, like did they search the business, then leave the review or was a QR code scanned? do you know what I'm saying? Like Google's looking at that too.
Greg Sterling: But see, but Google facilitates that themselves. Google makes it easy. They do that themselves.
Mike Blumenthal: They do facilitate it, but it could be part of Google's plan to discern between brand searches and review requests, right? A true brand search wouldn't use that. And so it might even be tactically advisable for people to not use Google's review code and just have people do brand searches to leave a review. I mean, it may, I would hypothesize that that would serve them better over the five-year period than that review link, which Google can easily track.
Claudia Tomina: Right, because it's... That share this link to leave a review, I never recommend that for clients because I would say that you're like 80 % more likely to get filtered when you use that feature.
Greg Sterling: that would infuriate. Yeah.
Mike Blumenthal: And we've talked about in the past about how brand searches reinforce and rank. And so there may be multiple reasons to not use Google's link, not the least of which is they, as Claudia points out, they're likely to use it as a signal.
Greg Sterling: Yes. So the follow-up email is, hey, if you liked us, leave us a review. Search for us on Google, and then leave us a review. And here's the template where you can mention all these keywords in your review.
Claudia Tomina: Well, so the business, one of the businesses that lost a bunch of reviews. So see, there was like 400 reviews in one month that was left. And I believe like upwards of 300 plus were taken down. Why were 80 or so left? My theory was that they were actual brand search reviews. And that they weren't left. Yes. Or they do text campaigns as well. So they may have sent out an email or a text.
Greg Sterling: Brand Search preceded the review.
Mike Blumenthal: and
Claudia Tomina: which is sending a link that's not the, leave us a review link that Google provides. It's more of a branded search link. And that was not filtered out.
Greg Sterling: Do you, I mean.
Mike Blumenthal: And you could provide a link in an email to a brand search, Greg. And I don't think Google would distinguish between that and an actual brand search. I think it would fall within this is a brand search.
Claudia Tomina: That's how our app works. Like it's a branded search.
Greg Sterling: So you pre-populate the brand search and people just click on it.
Claudia Tomina: Yeah, and then it pops up for the business when they click.
Greg Sterling: and then Google.
Mike Blumenthal: And then you say this keyword and that keyword and the other keyword and go from there.
Greg Sterling: Yeah. Yeah.
Claudia Tomina: Well, you can, in the text or in the email, can, hey, thanks for catering with us. So hopefully they will mention catering.
Greg Sterling: Yeah. One of my sort of beefs has been that Google has not done a lot of education of small businesses, of local businesses about what is permitted and what's not permitted. I mean, they rely to a high degree on this network of agencies that are managing these accounts to understand the rules and to advise them. But there are many, many more businesses out there that aren't working with agencies that are just trying to do this on their own. And, you know, some of them are operating in bad faith and may be cheating. Others are operating in good faith and simply don't understand what's going on. or don't have clarity around the rules. To what extent do you think that that's true or not?
Claudia Tomina: Yeah, it's totally true, but everything's fluid with them. They cannot make public statements and hold themselves liable to something if they want to change it down the line. So, you know, they're trained not to speak because they can't. That's just like at the end of the day, that algorithms change or they want to prioritize something or they want to adapt to the Yelp model and say, hey, you know what? We decided you shouldn't ask for reviews. If they've already made other public statements educating people on how to ask for reviews, then it kind of makes it difficult for them to change.
Greg Sterling: For sure, but I mean, don't you think they have an obligation to be clear about what the rules are so people can follow them?
Mike Blumenthal: They publish the rules and they intentionally leave a lot of vagueness so that they have the latitude to make a decision within a very broad framework to take it down. In other words, in compliance with 230, they're given the ability to editorially remove reviews and they want that to be as broad as possible. So they write it down in a very somewhat vague way to give them the most freedom possible and the least cost possible. don't think, you know,
Claudia Tomina: Yes.
Greg Sterling: Thank
Claudia Tomina: If there's one thing I've learned.
Mike Blumenthal: And they've never spent money on either education or not in local on promotion.
Greg Sterling: That's not 100 % true. Years ago, they did a whole bunch of workshops. They did a lot of ad workshops early on, and then they did. Right. But they did.
Mike Blumenthal: Not 100 % true. And years ago they sponsored Local U but that was more, I mean, it wasn't with an effort to clarify what they wanted. It was an effort to get people to spend money on ads.
Greg Sterling: They did these local boot camps for a while. They did local boot camps for a while. Interestingly, OpenAI is doing a sort of a version of that now. Okay, so.
Claudia Tomina: just think everything Google does is very intentional. You know, we may think, it's a bug, it's a bug. Yeah, but it's not, they're never really bugs. And if it is a bug, typically, like you would see it kind of get worked out within a month or two, if it's a major bug. I mean, there's bugs that are never going to get cleared out. But you know, like the review purge, I think in the beginning, even I felt like, like what's going on? there, I mean, obviously I knew there was some kind of algorithmic update and I thought,
Greg Sterling: of course it is. you
Claudia Tomina: will they fix this? But then it became clear that it was extremely intentional and that there was no fix and that it was our job to kind of let these businesses know like you lost your reviews. That's it. It's over. Move on. But it's
Greg Sterling: Well, I mean, they for years and years, they denied that there was any behavioral element to to ranking and that proved to be completely false, you know, so that that loan, you know, it's it's a version of deception, right? You know, you can give them the benefit of the doubt and say they don't want to reveal too much because then spammers would take over, blah, blah, blah. But but but they they they do deceptive things with some regularity. would I would use that term. All right, so what is the practical advice to business owners now? So where we are with the state of reviews? What would you say? What are the things that they need to really understand?
Claudia Tomina: Everyone should back up their reviews. Everyone should be like scraping their reviews or having copies of their reviews because they never know if they're going to need to go back. there were threads where I escalated in the forums and I got reviews back. I'm not saying like it's entirely impossible. There are times where like the hundreds or thousands that you lost, may be somebody said that you incentivize and so they see that flag and then you're not getting your reviews back. But there are times where when they do make updates to the algorithm and then genuine reviews get filtered out, you can possibly get them back. But you need the proof. If you don't have proof of the review, if you're like, I I had 300 reviews and now today I have 270. Well, if you cannot provide the 30 reviews to Google support, I don't believe you'll be able to get them back. What do you think, Mike?
Mike Blumenthal: I absolutely agree that if it's that volume and there are several ways to do this, some software like GMB API actually tracks removed reviews and gives them to you in a downloadable format. ⁓ the other way is, is archiving the emails that Google might send to you. Sometimes they send, re review announcements that there's a new review to the.
Claudia Tomina: I do the same thing.
Mike Blumenthal: owner, manager it's not a terribly reliable system. But if you archive those, I agree that absolutely you should be tracking every review, the text, the username, and ideally the link to it if you can and keep it in case you ever need it.
Claudia Tomina: Especially with suspensions, there's times where you get reinstated and then all of a sudden your reviews are missing. I anything can happen. So, right.
Mike Blumenthal: And they can't find them unless you give them, you spoon feed it back to them. They're amazingly incompetent in that.
Greg Sterling: All right, let's move on to the inevitable topic of AI. So when Google SGE showed up, there was a lot of duplication. If you're doing a local search, they would essentially duplicate most of the local pack in those SGE results at the top of the page. You had the local pack, and then very often two out of the three that would show up. Sometimes they had five, sometimes they had more than five in the SGE result. And it was totally redundant, or almost totally redundant. know what to do with the SGE result. And then local disappeared when AI overviews launched. Local wasn't present. Over time, that's changed. And now, probably more than anybody, Joy Hawkins is kind of tracking the degree to which the local pack or some version of the local pack is starting to show up in AI overviews, which portends the disappearance of the traditional local pack. thoughts.
Claudia Tomina: So I think it's affecting a lot like specific verticals more than others. I'm not seeing it in, you know, my clients where I'm primarily focused on the restaurants. She's primarily focused on service based companies and lawyers. And so yes, like it's, it's definitely something and probably because those type of businesses are the ones that do the most ads. So they're just focused on that category for that reason. Some type of monetization is going to be coming out of that.
Greg Sterling: Well, yeah.
Mike Blumenthal: And it could also be a function of query length in the category, right? We saw very clearly in the legal research that when queries got more than four and a half words or five words long, we saw an increase. Users are making longer queries. And I think there's this, again, learning that users can make longer queries. as they do, I think we may see a shift towards more AI. So we don't know exactly what's going trigger it, but I think cost
Greg Sterling: Right. you get the AIOverview
Claudia Tomina: Yeah.
Mike Blumenthal: is a big issue. the local pack is very cheap for them, fast and cheap for them to deliver as our LSAs and their LSAs and local pack, presume are profitable for them. They charge enough for them and,
Greg Sterling: Well, LSA is for sure.
Claudia Tomina: Well, and then it's triggering like get quotes and those type of things. So that's probably why they're testing it more than other categories.
Greg Sterling: Yeah, well, I a restaurant query is more, I mean, you know, occasionally you'll see these longer queries, but for the most part, a restaurant query is going to be best cuisine type near me, brand name, you know, or whatever. It's going to be, these are going to be fairly compact queries. And so there's no, this phenomenon of ambiguity is that triggers the AI overview or is it an informational surge? it's not, you know, that's not going to show up as much, but, but the, the You're suggesting that by putting the local pack in the AI overview that they can monetize that? Is that what you guys were saying? Or Claudia, you were saying?
Claudia Tomina: Well, then your trigger to go into AI mode maybe, they're just, I think they're just constantly just testing behavioral signals and then what they can do with it after. I don't think they know what they want, what they're gonna do. I think it's all tests.
Greg Sterling: Yep. Right. There's a lot more tests going on. I mean, you you've pointed this out and we've done our own version of this. There's just a ton of testing happening all the time now. And yeah, and it's, it's, it's very disorienting obviously, because, you know, is this going to be something that takes hold? Is this just a test? I mean, the thing that you saw the other day was Gemini, right? Where there was all this AI generated information before you got to the result. Describe that a little bit.
Claudia Tomina: all the time. It's quite obnoxious. Yes. Yeah, I was was like shocked to see it because I've never seen that in Gemini. And then all of sudden I did a like I had a restaurant search query in Gemini. I wanted to see what pop up and it was just this like extra layer. And then at the bottom there was a call to action that said open in maps. But everything that was there was like it was kind of similar to like know before you go, but not exactly. It was more like tips and everything review related. So like everything was pulled from reviews. It wasn't the most attractive UI, but it was just like a one pager, lots of data from reviews, and then open in Maps.
Greg Sterling: This was a brand search for a particular restaurant.
Claudia Tomina: Yes. Or no, and it was was just best sushi in Michigan. That's what I was searching.
Greg Sterling: Okay. And so what and did they show you, did they hint at multiple businesses or what did they actually show you?
Claudia Tomina: So yes, they displayed, I wanna say four businesses, and when I would click on the business name, it was linked, it would take me to that page, and then I had the option to open in Maps. So it was like a layer before the layer.
Greg Sterling: to the profile. Yeah. Yeah, that's like too many clicks going on.
Claudia Tomina: Yeah, it wasn't the best and it's only in mobile, it's not in desktop. So that's another thing. It's like what you test in mobile is not always going to be the same as desktop and so, and then AI mode, it's just going to display a little bit differently. So I like to test in all places because I like to see how things are being presented and what's their focus. And it's always typically reviews, photos, videos, content.
Greg Sterling: Well, so we've got AI overviews, AI mode, Gemini, Web Guide, which fewer people are talking about and using. I've opted into Web Guide So all of my desktop results, at least in one of my. browser screens is Web Guide, which I initially liked a lot and thought was kind of a good UI as sort of a hybrid. But now I don't like it because there's less information on the page. It's just much more text, fewer results. I don't like it as much. And it's especially bad in mobile, think, because you have to scroll and scroll and scroll and there's too much text. where many people, including you recently, or not so recently perhaps, speculated that AI mode is the future of the Google SERP. I don't think that that's true. I think there's elements of AI mode that make their way into the traditional SERP. But where do you think all of this is going? I mean, it's kind of a mess.
Claudia Tomina: I honestly think Google doesn't know where they're going with it. I think that they're still brainstorming through it. And I think they're just constantly testing. I don't know, like is there really a purpose to have AI mode and Gemini? Like why have them separated? Sometimes I think about that.
Greg Sterling: Right. Well, Gemini is a more general tool. It's their true ChatGPT competitor, whereas AI Mode is not really. It sort of is, but not really.
Mike Blumenthal: And as you saw in the recent announcement, Gemini is going to become the personal assistant technology. Well, no, it's going to become Google's personal assistant for you. They want you to tie in your email and your calendar personalization of Gemini. And they've also said that Gemini, they don't feel like Gemini is even closely ready for ads. So AI mode and
Greg Sterling: for Siri. you're talking about the personalization thing, Yeah.
Mike Blumenthal: and AI overviews and search is where they generate their money. So when you look at the big picture, from my point of view, firstly, it's competitive. They want to win against ChatGPT To do that, they've got to keep Gemini relatively pure. The second is income and the third is usability and how they mix and match those three things, the competitive nature of it. And then the fourth is a cost, right? AI mode is very expensive for them to generate some of these responses. And so they have to balance those four things, how that's going to all balance out, which of those four things Google is going to pick is most important. You know, think initially to be competitive and then as the competition fails or ChatGPT is absorbed by Microsoft when they run out of money, you know, who knows what we'll see. But for now it's those things are balancing each other. So we're not seeing the degradation of Gemini yet or AI mode yet.
Greg Sterling: So I saw something interesting the other day, which I posted on LinkedIn about, and may have talked about on one of these podcasts, where my wife and I went and saw the song sung blue, which is the Neil Diamond movie. You saw that, I know, Well, so one of the things, it was totally different than what I expected. So we came out of the movie, and it's based on a true story. And so I wanted to figure out what elements of the story were true and what happened to the people after the... Right, okay, so I did...
Mike Blumenthal: Very sweet movie, wasn't it? Same search I did afterwards. She did get hit by a car twice. Whoops, spoiler alert.
Claudia Tomina: Hahaha
Greg Sterling: Yeah, yeah, that's just an incredible detail that we don't need to talk about the movie. So so I did this. I did this search. I don't remember what the verbatim query was. It was something like, you know, song sung blue, true story or something, you know, question mark or whatever.
Mike Blumenthal: Yes, we do. reg Sterling (30:05) And it showed me an AI overview. I was on my iPhone. It showed me an AI overview, and it gave me a little bit of information. And then it said something like, it was a prompt to open the AI overview. It didn't say AI mode at the bottom. It said see more or something comparable to that, like learn more or whatever. It was some prompt to get the full, expand the AI overview. And what it did is it took me into AI mode. It didn't open the, open the AI overview in the SERP, which normally happens. It took me into AI mode, which I didn't recognize at the time. And so I'm just looking at the information. And then I thought, I'm in AI mode. That's weird. And then I compared the two, like the top of the AI overview with the AI mode result. And they were verbatim the same. So it was essentially the same information. I couldn't open the AI overview to see what was below it, because it just sent me the AI mode. And I thought, that's a path forward for Google with how to. reconcile the two of these things. So you go into AI mode and then you ask your follow-up question. I know they've been doing it, but I had not experienced that personally. They talked about that at some point in the past, but I had not seen it. Typically what you see, also what was striking to me is that typically you see something referencing AI mode to learn more or whatever, try this search on AI mode or whatever they're saying. This did not do that. It was just a generic prompt to learn more. And I thought,
Claudia Tomina: Well, they've been doing that. Yeah, they've been doing that.
Greg Sterling: Go ahead.
Mike Blumenthal: So let me, let me just interject slightly. did the exact same query or same intent of query. I don't remember some version of that. And perhaps, you know, they see all these new queries coming all the time, right? So a new movie comes out, a new query gets associated with it. Perhaps this is caching of AI mode information. In other words, that would reduce their costs. This is a query that's happening over and over again. So number of people went into AI mode.
Greg Sterling: some version of that, right?
Mike Blumenthal: They see that people go into AI mode. then automatically make AI mode with a cached response, lower cost, as a path. What do think of that theory? Good theory.
Greg Sterling: It's a good theory. It's a good theory. But why do you have AI overviews and AI mode? You don't need both together.
Claudia Tomina: Well, it's a way to get you into AI mode because people, everyone's used to going to Google, right? So it's like, there's Google and then how do they get them to start utilizing AI mode? They start with the AI overview, move them into AI mode. But I think they're starting to see a lot of people using Gemini and doing the same type of query searches that they would normally do on AI mode they're doing in Gemini. And so maybe they're trying to figure out how to balance that.
Greg Sterling: Well, for sure. For sure. But-
Claudia Tomina: It's the same thing.
Greg Sterling: Well, let's talk sort of for this final section, let's talk about sort of the future of AI and how that impacts all of the things that we talk about and write about and think about. Google made a big announcement about agentic shopping, including this thing called Business Agent, which is really interesting, which I think will come to GBP. There's Ask Maps is a version of that already, right? Where you have a chat bot. that you can ask any question about the business and it'll get you an answer. But what the agentic shopping blog post, which had a bunch of announcements in it, suggests is that more and more of the activity will be in an AI interface. It'll be you ask the question, you ask a follow-up question, and then you zero in on the product or the business that you want to, you know, do business with or want to buy. And then maybe you go to the website, or maybe you go someplace else to buy the thing. Google wants to sell you stuff within Google. I mean, this suggests a bunch of things to me. One is more zero-click activity, which is less of a problem in local because there's an offline transaction. More zero-click activity, less visitation of websites, less looking directly at reviews and more that this is all about data for the AI engine to assimilate and present to you. Thoughts on any of that stuff?
Claudia Tomina: So I used to sell on Amazon before I did marketing. So in the like 2000s, I was selling on Amazon before Prime even existed. And I've seen the evolution and I've seen how Amazon, you know, can dictate basically pricing just by not granting you buy box and all that kind of stuff. I kind of welcome this in a sense where perhaps maybe Amazon might have to change some of their practices because more people are already going to be in Google AI and searching for a product. And if they give a merchant the ability to run a transaction without having to run it through Amazon and worrying about the buy box and 15 to 20 % in fees, I don't know what Google takes and fees or what they're going to come out of this. But I kind of, I like the idea of where this is going to go. I think it's going to really catch on. I think Amazon is going to have to come up with some, I mean, Amazon's always going to do well. They have prime two day shipping, but most merchants use their fulfillment centers regardless. So you could, you could still use the fulfillment center to ship your stuff, but sell on other places. But at the end of the day, it's so hard to sell on your own website. You can't show up in the SERPs, right? Like Amazon dominates and all these other big, you know, marketplaces are dominating Walmart everywhere else. So if you're not selling on your, your brand on Walmart, it's very difficult to sell your product on your own website.
Greg Sterling: Let's talk about local businesses though. The businesses that we're chiefly interested in here, not necessarily product sellers, although local businesses can be product sellers. Do you see a further migration to some sort of AI, kind of the full funnel happens in AI, there's no click. And then there's a visitation or a call to a local business. People aren't reading reviews. aren't looking at, everything they look at is within the AI, in the AI environment. Right.
Claudia Tomina: Well, yeah, they're getting Well, they have merchant center. There's merchant center. And so that you have your product feeds in there. Eventually there's gonna be the ability to check out. Like they've been already trying to do that or pick up.
Greg Sterling: they're, yeah, they're, they, that's what they, they explicitly announced it. Yeah.
Claudia Tomina: Yeah, so that's all going to just start getting developed and there's going to be third parties that are going to be able to facilitate that for you. Maybe Shopify, all that is just going to be integrated right into AI mode. And so all of a sudden, you know, your Shopify site can run a transaction in AI mode. So you're less likely to maybe shop it on Amazon.
Mike Blumenthal: And we're seeing a version of that with restaurants now in AI mode, where it'll actually allow you to make a reservation without leaving AI mode using Google, Google reserve as the tool.
Greg Sterling: Right.
Claudia Tomina: Anytime you can have some sort of integration, any API integration, I've always said this, whether it's your menu where it's constantly updated because you have your toast integration, whether it's your open table or resi reservation list, like you have to be fully integrated into Google Business Profiles. So constantly be paying attention to that because at the end of the day, if somebody can't finish the transaction in AI mode, somebody else might win.
Greg Sterling: Well, that's, that's, go ahead, Mike.
Mike Blumenthal: I was going to say that more dystopian version of what you're talking about though, is what OpenAI suggested in terms of an OS for the internet, where all apps come within three or four or five agentic AI LLMs, Amazon, Facebook, Google, OpenAI, Apple, where this is the world, right? In other words, it could go well beyond reservations. What's that?
Greg Sterling: Yep.
Claudia Tomina: TikTok shop. TikTok shop.
Greg Sterling: TikTok Shop
Mike Blumenthal: Yes, all where these things become totally integrated. Everything from graphic editing to reservations could be totally integrated into two or three or four OS's that then battle it out so that in the end we have three OS's on the internet, right? I mean, and that's entirely possible. So it could go.
Greg Sterling: Well, that's where everybody wants to go. That's where everybody wants to go with this. totally. Because, well, Google would like nothing better than to own the entire funnel. That's what's at stake is that Google is the place where you discover everything. Google is the place where you learn and compare everything. And Google is the place where you buy things. I mean, they were unsuccessful in launching their own shopping cart.
Mike Blumenthal: Right. And one has to presume that Google wants to go there. If ChatGPT wants to go there, Google wants to go there, right?
Greg Sterling: a number of years ago as a kind of competitive tool against Amazon. But now with this, they can. But they want to own the entire funnel, show you ads at potentially maybe the top and the bottom or everywhere. And I think that we interviewed Jes Schultz a number of weeks ago. And she was making an argument about users and sort of the, I mean, users fundamentally like the AI experience. They like it better than search, than traditional Google search. They trust traditional Google, they trust the data and they're familiar with traditional Google, but they like.
Mike Blumenthal: If I can get Google to do a fan out query for me, I'm in because it gives me 25 sources and it summarizes them for me.
Greg Sterling: Right. That's right. people like the AI experience better than they like the traditional search experience. And Google understands this very clearly. And so that's why it doesn't matter if OpenAI disappears or all these other sites disappear. They are still migrating, maybe not at the same speed, but they are still migrating to some version of this going forward. And that's the reality that everybody has to sort of prepare for, I think.
Mike Blumenthal: or moving to the woods.
Greg Sterling: Yeah, that's right. That's the other alternative.
Claudia Tomina: It's very intrusive, but at the same time, everybody likes it and it's going to catch on no matter what.
Greg Sterling: Well, it's just, yeah, it's just more fluid. It conforms to your own. You don't have to do these kind of truncated clipped phrases. You can just speak to it. You can zero in on something. It's just much more flexible than what search has been.
Claudia Tomina: love when Amazon reminds me to buy something at the end of my checkout. you know, when you're checking out for Amazon and then it reminds me, do you also purchase this in the past? Do you need it now again? So Google's gonna do the same type of stuff.
Greg Sterling: Totally, and that's what the personalization thing where you are linking all these apps and letting them have all your data, they're just gonna be making recommendation after recommendation in an effort to get you to buy. So in this last sort of couple of minutes that we have, as people look forward to 2026, which we're sort of in the third week of, I guess, what should they be doing this year? I mean, there's gonna be a lot of change, there's gonna be a lot of volatility. sort of probably more than 2025. What is your advice to marketers?
Claudia Tomina: My advice is look at your Google business profile, not so much as like obviously the first step, but more or less the last step. Can somebody book or run a transaction from AI mode from your Google business profile? Look at it as like a portfolio. At the end of the day, if somebody's doing a brand search, how do you appear? Do you have reviews like? There's many times where financial advisors will say, I don't get leads from Amazon or Google, so I don't really care. But everyone should care about their Google business profile because it could potentially be the last touch point. And if you don't look good, somebody will bounce off and you're not going to win that search. So all your efforts that you do outside of Google on social and everywhere else, if somebody ends up searching your brand and you're not presented well, you couldn't lose. So make sure at the end of the day, your Google business profile looks the right way and check everywhere. Ask AI mode about your brand and ask Gemini about your brand and make sure that the data that you want presented is there. And if there's something about your brand that isn't there, start producing better content on your website.
Greg Sterling: That's good advice. Thank you, Claudia. And we've seen exactly what you described in our user testing, where people will disqualify potential vendors on the basis of what is or isn't present in their Google Business Profile. All right. Well, thanks for joining us. Great conversation, great advice. And we will obviously have you back in the future. And where can people find you if they want to contact you?
Claudia Tomina: LinkedIn is always a good spot to find me or you can go on reputationarm.com and fill out a contact form and schedule a call.
Greg Sterling: Okay, all right, well thanks and thanks for listening everybody. We'll see you next week.
Claudia Tomina: Thanks.
