Thank you to Cindy Krum for joining us for a Campfire Chat on Youtube on May 14, 2025.

Re-watch the Youtube interview with Cindy as she helps us dissect some recent Google Trial Docs, the future of SEO and a lot more. Cindy also joined us the next day for an SEO interview, live in Denver. We're so grateful for her time and energy. 

Read the full transcript: 

Show Transcript

Noah Learner: Hey there, I'm Noah Lerner and welcome to our 10th Campfire Chat today with Cindy Krum.

Just to let you know all about her, she's been working in search ever since 2003, she's the founder and CEO of Mobile Moxie and has worked with some of the largest and most iconic brands as an embedded consultant. She was named by USA Today as one of the 10 most influential SEOs in our industry back in 2022.

She has spoken at nearly every major SEO event in the world multiple times. She was the co-winner of SEO Oktoberfest in 2024. She's a foodie, global traveler, animal lover, author, and someone I'm lucky to have as a friend.

Cindy, it's amazing to have you with us on Campfire Chat. Before we jump in, I want to thank our community partners that make all of this possible.

Thank you!

Ahrefs, AirOps, Bento, Citation Labs, Domain Sphere, Drivefly Digital, ModX, and SEO jobs.

Cindy, we got to start at the beginning. How'd you get started in SEO?

Cindy Krum: was almost an accident. I was building websites and they weren't ranking well. And so I was just tasked with figuring it out. But then it's more interesting how I got into mobile. I was going to leave this website building job and just do straight marketing. And I saw a job for mobile marketing. And I was like, that sounds interesting. And I thought it was cell phone stuff. But it turned out to be

wrapping like horse trailers with graphic design stuff. That was their idea of mobile marketing. And I didn't figure that out until the interview.

Noah Learner: That's amazing. That is wild. So, and you wrote a book about mobile marketing back in 2010.

Cindy Krum: Yeah, cell phone stuff, not horse trailers.

Noah Learner: Yeah, I was wondering if the two things, did you talk about that in the book? man.

Cindy Krum: No, it was more of a technical book. The one criticism that I got that really resonated with me on that book was that it was not a lot of fun. It was just technical.

Noah Learner: bummer. But at the same time, saw the reviews. The reviews were strong. You had a whole bunch of them. I have been slightly obsessed with the Google trial recently. As per you and I have talked about this stuff privately. So I did a little bit of reading to get ready for stuff. Have you been following the trial in the past two weeks or so?

Cindy Krum: Well, loosely, I don't read everything. I usually read more of the pundits and stuff like that. But I think that a lot of things that we as an industry have been thinking for a while seem to have been proven out this week. That's what I've seen. Is that what you've seen?

Noah Learner: Mm-hmm.

Yeah, and I can't tell if I'm actually showing something on the screen. Okay, so for everybody at home, this link will take you to something on the government, our government's website, which will allow you to see all of the exhibits in the trial. So you get to see all of these like internal documents, you get to see emails going back and forth, you get to see strategy documents inside Google.

some open AI stuff that's internal and this one thing completely grabbed me and I want to share my screen to show everybody this because this is so wild. Okay, I'm not super great at sharing things so let's see if I can figure this out. Okay, so this one document really, really, really caught my eye.

yeah, so this is Google talking internally to their search board of directors and I need to get rid of, this is like the worst share ever, so excuse me on this, but I wanted to get your take on this.

Noah Learner: So you and I've talked a ton about how now search is all about the journey. It's about bio journeys and just life journeys. And what I found to be super interesting in this particular slide was how they talked about how many important questions are actually projects from life decisions. How do I start a career in data science? I've actually looked into this to everyday routines, like what's a dinner plan?

This gets me thinking a lot about the future of our careers, of the type of work that we're going to do, and the types of content that we need to produce. I wanted to get your feedback on this, because I feel like this one slide that Google is talking to their own internal board of directors, like they're saying, guess what? Search is going to be entirely new.

It's going to be entirely different than how it's been before.

AI mode or whatever it's going to be named in its final iteration is probably what the primary interface is going to look like. What are your thoughts on that?

Cindy Krum: Yeah, I mean, this is super telling. Wait, can you put it back? Super telling with two different parts of this I think are especially interesting. The first part is the last sentence in the first paragraph that talks about answers that are reimagined as a, well, that are translated and potentially even reimagined as a podcast. So that,

That speaks to what we're seeing right now, and I think what I've been predicting is about to happen, which is that when Mobile First Indexing launched, they tried to get rid of all of the Google CCTLDs and consolidate towards one CCTLD and no language-specific, country-specific content. They messed it up, and they had to roll back, and they left the CCTLDs.

back in 2018 when they first tried. Now they're trying again, and I think that it speaks to what we're getting ready for, which is, you know, it's suspicious that they do that in the same week that they also bring Google Discover to desktop, which is two weeks before Google I.O.

, where they're slotted to announce that Gemini is coming to desktop. And I think all of this is kind of...

coming towards a kind of consolidation, a centralization of what Google is in terms of Google Discover, AI Overviews, and international. But the translation piece and the reimagining piece are both very active. Google's not in this paragraph shying away from creating the content that you would consume that would answer your question.

So if it didn't exist in your language, they will translate it for you.

That's not new, but it does speak to one of the major problems Google tries to solve, which is Google's great at big languages and less good at answering questions in small languages. But if they understand things at a topical entity journey level, then they can just translate the journey and the entity information and the response from a language that they're good at, like English, and put it in whatever smaller language where they have less training data available.

Cindy Krum: So that makes them better everywhere. So this is great for users, users who speak and search in smaller languages. But then the reimagined into a podcast, I've mixed feeling that sounds great for me because I'm a listener. I like to listen. But they're again, creating content that didn't exist in the first place. And then in the next paragraph, you keyed into projects. But I think what's interesting

is more the last sentence again, search will also serve as a thought partner that's AI organized. And to me, that speaks to the idea of conversational search more than anything. In 2018, when Mobile First Indexing rolled out, they said they wanted it to be predictive, they wanted it to be conversational, and they wanted it to be kind of, I think the third one was AI organized or like,

intelligent. That's not the right word. But there were three things that they said, conversational, predictive, and somehow intelligent. And this is what we're seeing now come to life, The time between 2018 and now, they've been getting to this goal that I think that they're going to try and launch at Google I.

O. Sorry, long rant. That's that. Yeah.

Noah Learner: I love it. I love it. So the last sentence is like jumping off the page at me. Search will also serve as a thought partner that is AI organized. The first thing that they say that it is, is AI organized, interactive and enabled by an ecosystem that will grow alongside search. So I'm hearing like Gemini, I'm hearing Vertex, I'm here. Yeah. Wild. So

Cindy Krum: with or without creator content and with or without potentially Google created content, Google synthesized content.

Noah Learner: Yeah. Wild. So you've been teaching us how to think about SEO forever. Like, I feel like you got me to think about mobile before anyone else had really drilled down into my brain hole. You got me to think about entities. You got me to think about Fraggles. You got me to think about mum. I mean,

It's like your entire career of changing the way that we all think about and talk about the practice of SEO. When we had lunch recently, I asked you like, how the hell do you do this? Like how, where does this come from?

Where does, how do you see things that other people don't? And I was hoping you could share a little bit of that conversation with us and your thought process around it.

Cindy Krum: Yeah, yeah, for sure. I thought of the third word. It was personalized. They wanted to personalize, predictive, and conversational. And that kind of gets to the answer, too, of your next question, which is how do I manage to kind of understand where Google is going? And usually it's because I kind of try and break it down and figure out or look for patterns in what are they doing, what are they saying, and

Noah Learner: Yeah.

Cindy Krum: like why would they do those things from a business perspective? So understanding Google as a business that has costs that they want to keep down and profits that they want to drive up. so looking at everything that they do and questioning it, how does this make money? How does this lose money? That kind of helps you understand.

what they're trying to do, like where they're investing, where they might be spending more money than they're making in the early days, but then they anticipate a bigger payout later. So I think that's kind of where we've been, where Google's been using everything that's happening post-Mobile First Indexing to add to the understanding of the topic layer and entities and then the journeys that are associated with those entities. Once they've learned that, then they can hit the next level of these AI-organized predictive search results.

And when you think about predictive, personalized, and conversational, that's going to help Google take what they know from the open web and take what they know from, marry it with what they know from ads and help connect their advertisers where they make money with better, more users who are more likely to convert and be happy with the product.

in a faster, more predictive way. And they've had to kind of add this layer of obfuscation in both organic and paid with the AI stuff where they can kind of take all the things that they know, regardless of where they learned it, and then remix it to sell this user to the highest bidder, right? A la PMAX.

If you're just an organic...

like SEO person, PMAX is the new AI layer in ads. And it can get you to all different kinds of ads when you set up a PMAX campaign, but it's AI. it does it with like more of, they sell it as a hands-off approach to your advertising.

But what that means is you have less control and you have less learnings that you walk away from. So like it used to be when you were bidding on a keyword, you could use a PPC to test.

Cindy Krum: things like a meta description or a title, you could like A-B test it with your titles and ads and see which gets the most clicks and be like, all right, this gets the most clicks. So if we're in number one, maybe we switch out whatever title tag we had to something that's now just gonna get a bunch of clicks and keep us there. You used to be able to do that. Now with PMAX, you have less reporting and so there's less learnings and less takeaway. So in many ways, we're the...

marketers are losing on both sides of the equation. We created all this content so that Google can learn and build out the topic layer. And now we're not getting credit there.

And we set up all these ads and Google just kind of uses this obfuscation layer to send users to whichever ad they're going to potentially make the most money on or what have you.

bit in my mind disingenuous, but they get away with it because there aren't competitors. That's so much in there. There's a lot to unpack. I'm sorry.

Noah Learner: I love it. So start with the money.

Cindy Krum: Yeah, start with the money. Always, they're a business. They've got to make money.

Noah Learner: Yeah. So I feel like we're doing our own court presentation here today. So I don't know if you saw this one. So, so Pandu was, he testified last year or the year before, and there's a bunch of stuff that he was talking about. Um, I want to share another tab first, which is this one with a guy named H.J. Kim. I believe they call him H.J.

Where they talk about hand crafting of signals. So I think people thought that a lot of this stuff was out of the engineers control and this document that was just released kind of changes that perspective a lot. Have you seen this before?

Cindy Krum: Yeah, I took a quick look at this and I thought the hand crafting of signals was interesting because you're right, I did assume that it was more AI controlled and I wonder how much things are still hand crafted because it does seem like in some cases Google, you know, like it's a mystery even to Google why things rank.

Noah Learner: Mm-hmm.

Cindy Krum: or why something would rank number one when it doesn't have as good a signals as other things. So I wonder how true this hand crafting of signals is now. I think the hand crafting of signals was important to train the AI in how to build an algorithm. But so this is okay. So February 2025. Yeah, it's wild.

Noah Learner: Yeah, isn't that wild? This blew my mind. So what I thought was super crazy about this was that they came to the conclusion that the only way that they would understand how to debug anything was if they made it simpler and made it so that they could control the inputs and tweak knobs as it were. And that HJ's first project, I guess, or second project was NavBoost.

super wild and then ABC signals. I hadn't heard it described this way before. So it's like links. You know what's in the body of the of the text of the page and then click data.

Those are the three things that make up something that they describe as topicality. And so link builders are going to love a they're going to love the fact that links still matter and then.

When I think about making my content matter, I think about vector embeddings, right? Like because matching the query to the score of the page and then click data is Chrome, right?

Cindy Krum: Exactly. And this is what that one presentation that won Oktoberfest was about Google's switch to mobile-first indexing being a way that Google switched to leaning much harder on the click and engagement data from Chrome as an evaluation of the success of a page or a site.

Noah Learner: Okay, there's some more stuff that I saw here and I wanted to know if you had any understanding of this at all. Ranking signals curves. So they basically have over a hundred different raw signals and that those become top level signals that bleed into their understanding. And that each of these different ranking signals has a curve.

and that they can plot the curve to sort of tweak the knobs to determine how results happen. And what was super interesting in a lot of these different documents is that the engineers were saying that their algorithm could be reverse engineered pretty easily. They said, they actually used the term pretty easily as long as the people had the underlying data.

Their competitors don't have the underlying data, which I thought was super interesting.

Cindy Krum: Yeah, and the underlying data, I think, is coming from Chrome. And I think these ranking signal curves, what they're talking about is they have kind of a

data about how normal successful pages look when they're engaged with and how they don't look, right? And part of these curves are to find low quality pages, but also bot manipulation. Because the bot manipulation isn't trained to fit a curve, it's trained to do one thing, to go and do a thing and click. And that makes the pattern recognition

for manipulation much easier when you do it on a normal curve versus an abnormal.

Noah Learner: So this sentence that's at the top of the page just completely jumped out at me when I read it. If Google's forced to give information on clicks, URLs and queries, that's like the inputs, it would be easy for competitors to figure out the high level buckets that compose the final information retrieval score. High level buckets are EBC topicality, nav boost and quality. And then...

Cindy Krum: Thank

Noah Learner: What I thought was pretty interesting is equality is generally static. That was really interesting to me.

Let's see, page quality is cute. That's a metric I hadn't heard before as Q star. And they talk about how important it is and then quality score is still super important.

And then I didn't know if you knew about all of this stuff. E-deep rank is an LLM system that uses BERT transformers. I didn't know if you had any thoughts on that.

Cindy Krum: So it looks like E-deep rank is related to what I would call fraggles, what Google would call passages, right? It's trying to break one page with multiple topicalities out potentially into component pages or component parts. And maybe it's one page with one topicality, but it's trying to find then subtopics on the page. That's what I see there.

Noah Learner: Yeah.

Cindy Krum: And it's straight up saying LLM based signals to decompose. yeah.

Noah Learner: Are you enjoying this by the way? feel like I've never done like court document breakdowns with anyone in a podcast and I want to make sure you're into it.

Cindy Krum: Yeah, no, you're ahead of me on this. I haven't read the actual court documents, so I'm enjoying this a lot and I'm learning stuff. I wonder what is that word that's before popularity signal that uses Chrome? Is it click popularity signal? Or Q star popularity signal? I wonder what is blocked out in that last dash, because it's a short word.

Noah Learner: Yeah, me too.

Noah Learner: I think it's CTR.

Cindy Krum: yeah.

Noah Learner: I had read this earlier and I was thinking about like what acronym would be short enough to fit there. User side data.

Cindy Krum: What fits there? Yeah. Yeah.

Noah Learner: That to me is Chrome.

Cindy Krum: Yeah, interaction data, not the content data. So this, think, might be, have you ever looked at Chrome histograms?

Noah Learner: Ooh, make us smarter.

Cindy Krum: Okay, so open up a Chrome window, and ideally one that you have other tabs opened in, and then type in chrome colon slash slash histograms, H-I-S-T-O-G-R-A-M-S, and hit enter, and share it.

Noah Learner: histograms.

Noah Learner: Whoa.

Cindy Krum: Yeah, don't don't don't hug it. Share it first. Then you can look.

Noah Learner: sorry. I don't know what the heck I'm looking at.

Cindy Krum: So, okay, so if you want, can hit switch to monitor mode. That's more fun.

Cindy Krum: So what this is going to do if you scroll is it's moving, right? It's changing as we go. And it's capturing all of the things that are happening on all of the different tabs. And this is how Google reports back your user behavior in Chrome. And so my idea in that one talk that was so popular was that Google is using Chrome histograms and Chrome rendering.

as part of mobile-first indexing to understand not just the contents of the page, but how people interact with the page and how the page loads, things that you might get out of Core Web Vitals, which, of course, launched right after mobile-first indexing. So I think this goes on forever. But hit a plus box.

Hit a plus box. Hit a plus on the left. Yeah.

Noah Learner: There's so much information here. I can't believe how much information there is. This is it on any of them? Okay.

Cindy Krum: So keep hitting them, that one was boring. What a histogram is, is just a sideways bar chart. So if you keep clicking around, you'll get some that have lots of information because you clicked or did some behavior on a page. These ones are boring histograms. But if we find one that we know you did, for instance, it has some dashed lines that are really long.

Noah Learner: Food navigation.

Cindy Krum: And so it's giving us the information about how you behaved on a particular page. And if you search in here, you'll find things like filled in a form, filled in a credit card information, clicked an anchor link. You can find stuff on the XY axis. And this is all kind of a little bar chart of your behavior. There's one.

of your behavior on a particular page. So API storage access header recorded. Okay. So I don't know what this means, but you did it.

Noah Learner: my god, this is the riot. Okay, I'm gonna share this tab. Okay.

Cindy Krum: Do a Control-F for anchor, let's see. But it also, if you look in here, it comes, you can see a lot of the Core Web Vitals stuff in here with Largest Contentful Paint. There's one.

Noah Learner: Ooh, that's like me clicking links, right?

Cindy Krum: You clicked on an anchor somewhere a couple times, right?

Noah Learner: Whoa. So what are these things represent? Is that like a, an anchor link on the page with an ID of 175?

Cindy Krum: No, think that would probably be the pixel height.

Noah Learner: wild.

Yeah, wild. OK, back to the court documents, because this stuff is. my God.

Cindy Krum: Yeah, so this is what that's talking about, I think.

Noah Learner: Let's see Tetris. What do know about Tetris?

Other than I'm not very good at the game.

Cindy Krum: We called 10 grams. Did you ever play 10 grams the game for kids?

Noah Learner: No, I played Tetris and I sucked at it.

Cindy Krum: I played Tetris as well. So tangrams is you have to replicate certain shapes, but I don't know what it means. That's what the game is. And Tetris also you're replicating in some ways. You're using shapes to create blocks. Tangrams to apply basic principles of search to all of the features. So I imagine, like I said, I haven't read all this.

What I imagine is by observing data logs, blah, blah, I imagine this is a fitting thing. So, and it's kind of like the curves. So it's to say to rank here, you must fit this shape, this overall shape on multiple categories, right?

So the top rank, it's kind of like what we say when we're talking about EEAT, the top ranking pages all have signals that look like this.

Top ranking pages have these shapes of curves in their signals, and you should have that shape too, if you want to rank. That's my guess.

Noah Learner: Wow. This jumped off the page once again. The documents alone do not give you enough details to figure it out, but the data likely does.

Cindy Krum: Well, and that's why Google is saying that they cannot sever Chrome as a business entity, right? And that speaks to the idea of what I was saying is like Chrome is now critical. Chrome rendering and Chrome data collection is now critical for the algorithm and the ads, right? It makes both of those things viable businesses without Chrome. It would not potentially be viable, which is interesting from an international perspective because the EU says that

I'm can me.

Noah Learner: I'm gonna switch to another document because this jumped off the page at me as well. Rank embed and how disclosure of Google's information could be used by competitors to reverse engineer Google search innovations. Like I'm not trying to beat a dead horse with that statement, but the first bullet point there, rank embed is a dual encoder model that embeds both query and document into an embedding space.

Embedding space consider semantic properties of query and document in addition to other signals. Retrieval and ranking are then a dot product distance measure in the embedding space. So I read that and I was like, I need to learn more about vector embeddings and I need to start incorporating that into tooling somehow.

Cindy Krum: Yeah, but I know if rank embed and vector embed are the same thing, like rank embed to me sounds like if it's connecting the query in the document, what it's saying is that like vector embedding is kind of about content and more like topical relationships within the vector space on a page. This to me is more about

Noah Learner: Yeah.

Cindy Krum: kind of the way that you tell if a ranking organization is correct, so position one, two, and three, is by connecting the document and the query. And so when users consistently click on something in a particular ranking, that it entrenches that ranking and that topic with

the content. So to say that if we do a trial balloon, we're Google, and we see new content on your site, we throw it up high in the rankings, and it gets no clicks, it falls out. But we have new content, we think we know what it's about, we throw it up high in the rankings, and it sticks, it gets enough clicks, then I think that would be like an

a rank embed kind of situation where it's entrenching it, this page is entrenching itself on this topic and then potentially the vectors related to it to say this is ranking for this topic. It might also be related to the other things in that vector.

Noah Learner: Yeah.

Noah Learner: NavBoost, that was one thing that came out was that people thought NavBoost was an LLM type of situation and this document made it clear that it's a query to document lookup table where they're tracking how many times a document got clicked over the last 13 months for a specific query. And I thought that was really interesting. I didn't quite have that level of understanding previously.

Cindy Krum: Well, the lookup table is kind of wild because that speaks to the idea of maybe the relationship between nav boost and quality, right? Where quality is somewhat static, maybe potentially coming from this nav boost table. But yeah, that's not how I envisioned it either. But maybe when they say like nav boost table, it's just basically the topic layer. And it's basically like

entity relationships and vectors and stuff like that. I don't know. You're right. It's some of this stuff is more surprising than I would have expected.

Cause you expected it to be a moving target kind of, right? Like with, with ongoing data.

Noah Learner: I thought we were in complete black box land where anytime there's a new update, know, Google's like, ooh, you know, don't blame us. It's the algorithm, you know, versus they're still super manually involved. we went super deep, super technical, super granular. Hopefully we didn't lose our audience. this does tie me into one of your recent trips though.

Do you want to keep talking about this? I kind of feel like I want to hear about SAO week. I want to hear about your experience there. I want to hear what you learned, what you found interesting, what it was like to speak at the venue.

Everybody involved seemed to come away pretty blown away.

Cindy Krum: Yeah, see how we quiz fantastic. Of course, you know, no one would expect anything less from Mike King. He puts on a a good event, a great event, really. The the speakers were all top notch. The venue was great as a speaker. I really appreciated the way that they had the slides set up. So they had, you know, they had a confidence monitor, which is great. Speakers love that.

But they also had the slides kind of behind you and on the walls all around you. So everywhere you looked, you could see your slides without having to use the confidence monitor, which is fantastic. Yeah, never have to do one of those.

Because sometimes the stuff on the confidence monitor is too little. So you have to turn around. But no, was just everywhere. So that was great. Of course, the evening events were fantastic.

Noah Learner: Yeah, you didn't have to do one of these. You didn't have to. Yeah.

Cindy Krum: Busta Rhymes. So I spoke, I was the first speaker after Busta Rhymes, so I made a joke that Busta Rhymes was my opener, which got stuck the first thing in the morning. But also Mike did his blindfolded rapping. I don't know if you've ever seen videos of that or anything like that, but it's wild. So he blindfolds himself and

Noah Learner: No.

Cindy Krum: gets up on stage and people just send stuff to the stage from their purses or backpacks or just things they find in the venue. And he had to rap about it while he's figuring it out, like just in his hands blindfolded. It's great.

Noah Learner: That's hilarious.

Did you learn anything?

Cindy Krum: I met a lot of great people. Yes, actually I did learn, especially some of it from a stage. Some of the speakers had some really great tips about optimizing for AI. And then also, you know, at the conversations at the bar, some good tips and tricks about visibility in various.

AI search functions and chat functions in terms of, let's see, let me generalize. Being found in AI does seem to be.

So like you can get visibility in some platforms just by having a surge of tweets and a surge of retweets and stuff like that, or having a conversation with a lot of likes on Reddit, interesting stuff like that.

Noah Learner: someone asked how many tabs I had open. I had one, two, three, four, five, six, eight tabs, Brian, eight tabs open. that was one thing that I was, quite taken. Well, let me frame that after looking at all this people talking about SEO week, it seemed like everybody was talking about brand and that that was the biggest takeaway from that event.

That's entirely absent from what we just looked at. There's no brand signals.

Cindy Krum: Well, brand, I think, isn't brand part of Navboost.

Noah Learner: That's well, navboost is a query to document lookup table. So how many times it got clicked is so popularity, right? Like it's pop, you know, that popularity of the brand or links as a measure of popularity.

Cindy Krum: But before I eat.

Cindy Krum: Well, but brand names can also be queries, right? So brand understanding could happen in NavBoost that

Noah Learner: Mm-hmm.

Noah Learner: I guess what I'm saying is I had a hard time connecting the dots in that. So make me smarter. Like what was everybody talking about at SEO week where people's takeaway was all about brand and how do I keep those two things together or tie them together?

Cindy Krum: So truthfully, I have to say, the takeaway from most of the conferences I've spoken at in 2025 has been brand. And it's because brand is always the answer when Google is removing traffic that used to be there. Then there aren't tricks that will work that get rid of AI overviews, for instance. And so if AI overviews are taking your spot, there's no way to

Noah Learner: Mm-hmm.

Cindy Krum: outrank an AI overview, you can get into the AI overview, but even when you're in an AI overview, it's decreasing clicks. so, you know, safe marketers always, when they hunker down, it's towards brand and towards building what we can, controlling what we can control, and that's usually brand. But in a separate kind of vein, still associated with brand, what you do

when you're building out your brand is creating an association of your brand with a particular product. And that comes out really well in LLMs. That's something that LLMs can synthesize, which is the top hiking boot for people with wide feet or whatever.

Like this is something that LLMs...

are meant for because you don't want to read all the pages and know whether or not to trust all the hiking boot companies. But LLMs are great at that. as a brand, what companies can do is figure out what is your unique selling proposition because if you don't have one, you will be lost.

Noah Learner: The answer is Solomon, by the way. The answer is Solomon. Yeah.

Cindy Krum: Google's not just rewarding people for existing. They have to have something that makes them cooler than the competitors in some niche. And it's not just Google. That's the other thing is if you are creating brand awareness in branded queries, then you have a much better chance of getting the ranking in the click-through without

as much effort. And so you're starting higher up in the funnel to say, make people search for Solomon, did you say Solomon? Solomon hiking boots, so that they're not just starting with hiking boots, or they're not starting with hiking boots for wide feet.

They've already been exposed to the TikToks and the Reddit threads and whatever, you already know you need the hiking boots.

from Solomon. So that's kind of where branding comes in is to hunker down and find what makes you different from the competitors instead of just trying to rank on all the keywords in your space, narrowing the fields and picking your battles is what can still have success in SEO potentially and in AI overviews, but also in the other channels where if you're not being found in Google,

you need other strategies. And that's what we need to realize as marketers and SEOs is if our only marketing strategy was SEO, that's not a strategy, right? We need to have a diverse approach to getting not just traffic and not just organic traffic, but all traffic and all awareness and all sales, whether it's social or, you know, other stuff or whether it's offline or paid or other things like

Putting all your eggs in the SEO basket was never a great idea, and this is kind of proving that case.

Noah Learner: This takes me to a couple different topics, one of which is some areas where you've been doing a lot of work. You've been looking at different types of platforms, whether they're marketplaces or whether they're social platforms. Can you tell us a little bit about what you've been working on recently?

Cindy Krum: Yeah, so I'm really excited. Either we have launched or we're about to launch. I'm not sure what the update is today. Tracking for Amazon in my tool set. So what that means is launching the ability to synthesize a search result, or actually it's not even synthesized. It's a real search result in Amazon.

From any country because there will be different results in Amazon in different countries and then since I see the landing pages to track particular products. What the landing page looks like what the reviews are stuff like that and doing that not just in real time but over time so setting up automatic capture of what the search result looks like if there's a new competitor.

And then also parsing in things like star rankings and reviews and stuff like that in a table over time. So that's really exciting because I think that more and more, Google is fighting really hard for the e-commerce and product searches, but so many of them just start in Amazon now. And Google is reporting about increase in query volume.

Apple is questioning that and saying, maybe not, or maybe you're mucking with the way you're measuring it.

It is reasonable to assume that query volume is going up. It's not reasonable to assume that all queries start in Google anymore. Because people's search behaviors have changed.

know for me, there are places where I'll start a search that are not Google. I already know, I'm not going to buy something I know I can get for Amazon at a reasonable price with no shipping charge.

tried three different websites that might charge me shipping and might or might not have a better price. Do know what I mean? So I think more searches are starting in Amazon and Google's trying to fight that and get that back, but with Merchant Center and stuff like that.

But until they have a competitor for Amazon Prime where shipping questions aren't a blocker, I think Google's going to struggle there.

Noah Learner: Yeah, it's the default. Yeah.

Cindy Krum: And then the other thing, the next thing we'll be adding is TikTok because lots of the younger generation are starting their search on TikTok. I mean, no, TikTok is great. Like it's almost too good because I...

Noah Learner: the kids.

Noah Learner: I don't have it installed because I'm like so afraid of my data going overseas. And my wife just became an addict like in the last month.

Cindy Krum: Yeah!

Cindy Krum: It is addictive, I see the problem. But it's also influential. like, you know, I was worried about my data too, and then I just gave up and I was like, so many people are here. I need to understand how this works. And the younger generation is searching on TikTok. And the thing that I like about TikTok is that it is more obviously authentic and not gamed yet.

So like if I see that you're recommending a tool or something like that or you're recommending Salomon boots and you're showing me what they look like and you're showing me what they look like after two months and 200 miles of hiking or whatever then like that's a real testimonial that I know you know if it's affiliate sponsored you'll probably say that maybe you won't but it's not

AI generated reviews on a review collection site that are all potentially crap and fake, right?

Noah Learner: Hey, so funny. I don't know if you saw that. I just threw something up on the stream. Your SEO Oktoberfest co winner Eric Wu shared to my understanding there are several time windows. It's not just 13 months. There are shorter windows that are also evaluated, so the documents talk about that Eric. They talk about like using the shortest amount of time window possible to with a high enough confidence. So if it's.

Cindy Krum: I want to, can we get Eric on the freaking call? Like I love Eric. He's talking about in the chat, I just saw he's talking about TikTok search. Yeah, it's nuts. And it's got more features than Google SERP. Yeah, and the shopping is nuts. And like if TikTok.

Noah Learner: he's not allowed. I've tried to get him. I think he is.

Cindy Krum: I don't think TikTok has an Amazon Prime thing, but if they did, holy moly, so much e-commerce would be lost. Because I want to buy everything I see on TikTok, and it's usually $3. Buy it, buy it, buy it.

Noah Learner: Have you bought glasses on TikToks? I know we both share a love of glasses.

Cindy Krum: they weren't they weren't I didn't buy them from TikTok but there are a lot of influencers on TikTok who like my glasses it's not these it's the pair I wear ones I'll just tell everyone like it's cool like this was actually innovative where they let you have different face plates that are magnetically attached right here and so you can change the color so if you see me with lots of different colored glasses it's because number one like they're all the same glasses with different toppers so they do but it's TikTok shop yeah Amy but

Is TikTok shop, it doesn't have prime where you prepay and then you never pay shipping and you have reliable returns and stuff like that. Right? Correct me if I'm wrong.

I try not to buy things on TikTok because I don't want to start. I don't want to, you know, open the faucet. Okay.

Noah Learner: So AI mode. One of the documents that I saw in the court trial, they showed like a user interface for what the bar was gonna look like and it say like all Gemini. So is AI mode gonna be renamed Gemini, think, or is it just gonna continue to be AI mode and is that our search surface of the future? And if so, like how far off?

Cindy Krum: Nice comment.

Cindy Krum: I think Google wants to make Gemini the search surface of the future. think that what they'll try and do, they'll do a lot of testing, right, after Google I.O. They'll make some big announcements at Google I.O., but they'll, you know, ultimately Google Discover is the predictive search that Google said they wanted, you know, 10 or 15 years ago, even before 2018, they wanted a predictive search that shows you the search results without you even having to search. And so that's all been AI driven.

by user behavior, which is why iOS users have never had as good an experience on Google Discover as Android users, because Google has way more information about Android users, because they own it all. But remember that last year at Google I.O.

, Google talked about how Gemini was coming in at the OS level. Gemini is not a Chrome feature or a search feature on...

Android anyway, it's at the OS level, which means it can see not just web interactions, but app interactions as well. And so it's going to have even more information and data, and it's going to be able to connect the dots. And even at Apple's event, they had the same thing, where their AI agent was coming in at the OS level.

And so that's going to be a much richer, more compelling experience for users. It's going to involve more data transfer.

if users don't like that, it's scarier, but I think AI mode and Gemini all use similar backend technology. So Google just has to test how they're going to position it. What's going to cross what they call the creepy line, which is actually in some of their documentation.

They say don't cross the creepy line. They did cross the creepy line.

When they talk about journeys in your history, they used to be much better, and it crossed the creepy line. Like if you go to your history in Chrome, there's the chronological history, and then there's another tab now that used to be called Journeys. Now it's called Groups.

But Journeys matched up and launched right after Google Mom was announced, where Google Mom talks about Journeys.

Cindy Krum: and it would collect every time you searched on something on a topic or related to the topic and what websites you clicked on. And I think it would even give details about how long you stayed on the site. That crossed the creepy line. So they rolled it back a little bit. They don't do as good a job at clustering the information, and they have multiple clusters on the same topic, which is dumb. And they call it groups instead of journeys. The other time they crossed the creepy line was right when the creepy line's like a theoretical thing that Google talks about as like,

how, like when users are gonna stop using something because it creeps them out. So like when that, when they started showing the journeys with all that detail and the information, they saw like in their testing, whatever testing they did, apparently it crossed the creepy line. So they backed it off because it was too much personal information on display.

They were kind of showing off how much they knew and it gave people the ick, right? So...

Noah Learner: I gave people the ick.

Cindy Krum: But the same thing happened years ago when Google started focusing on mobile search and they had the, when Google launched accounts, right? Because Google used to not have accounts. When Google launched accounts and it was across Chrome and your phone, even when you searched on desktop, it would say your last search was from your blah blah blah device. And it would tell what kind of phone you had. That crossed the creepy line too and they stopped that. But that was the clear signal that they were aggregating your information across devices and your search behavior and stuff like that.

So AI mode, to get back to your question, they'll call it whatever they want. It doesn't matter what they call it. If they call it Gemini, if they call it, like Gemini right now, the distinction is Gemini is more of a chat functionality and more like chat GBT, whereas AI mode is more of a search functionality.

Maybe they keep both, maybe they don't. It'll depend on how users like it.

Noah Learner: Yeah.

Cindy Krum: Before they called it Gemini, it was SGE, and it was more chat oriented too. And it had lots of continuing conversations, and it even had the voice icon and the speaker icon. It was meant to be conversational. So yeah, I don't have a hard answer. think Google's just going to try and see what users like, and then tune it towards that.

Noah Learner: Okay, I got invited, so I'm going to IELT next week, and I got invited to a meeting with Danny Sullivan and some of their engineers and stuff. If you had any one question that you wanted me to ask him, what would it be?

Cindy Krum: my gosh, I miss Danny. With Danny, just want to, you know, this isn't the right thing, but I would just want to catch up with him and ask him how it's going because I miss seeing Danny at events. But Google-oriented, SEO-oriented questions, what would I ask Danny?

I wasn't ready for this question. would probably, well, I would ask him how involved he is, his position is in all of the AI mode, Gemini, AI overviews kind of thing, or is it a separate function where he's not consulted?

Noah Learner: stumped you I didn't mean to

Noah Learner: Okay. And yes, I'm writing it down.

Cindy Krum: Because I think that a lot of people assume that he is deeply integrated. And I wonder if that's true or not. What do you think?

Noah Learner: Hmm, I don't know. I mean honestly, I don't know

I don't know. I get the sense from the documents that the woman who's the head of search is talking to folks in Vertex and that they're collaborating super closely in helping Vertex and Gemini stay grounded using search results. And it seems like they're super tightly wounded.

I don't know how his role fits into that though. I don't know.

Cindy Krum: When he talks about his role specifically, he has said, I think, before that SEOs misunderstand his role and that his role is to kind of represent the SEO community within Google meetings and not to represent Google meetings within the SEO community. That's kind of the flavor of what I've heard him say. And so the other question would be, what are the biggest concerns that you're representing to Google?

Noah Learner: Hm.

Cindy Krum: from the SEO community to different functions and groups within Google, right? And to me, my biggest concern if I were Danny Sullivan would be any kind of blurring of the lines between paid and organic, which it does seem like we're getting closer to with AI overuse.

Noah Learner: Isn't that horse already out of the barn though? I mean like hasn't that already happened?

Cindy Krum: I don't think Danny would say it has.

Noah Learner: Okay, well, my biggest question is, what does our job look like in five years, 10 years, two years, one year? I mean, like, that's the question that I, if he's representing our community, I think that's the question that I get asked the most when I talk to other SEOs. They want to know what the future looks like for them.

Cindy Krum: Yeah, so I think this is the other reason that brand keeps coming up is you control what you can control. And Google is going to be casting about and over-correcting and doing weird stuff for a while. That's what I think. there may be like even what you think is a reliable amount of SEO traffic might go away, might change. So I think diversifying.

Noah Learner: Mm-hmm.

Cindy Krum: your digital marketing and organic SEO kinds of investments away from Google towards this idea of branding or just kind of what's the right word, being where your users are, interacting with your users in positive ways, even if it's not on your website. So

Noah Learner: Mm-hmm.

Cindy Krum: participating in the communities on Reddit, participating on all the social platforms, having content in videos and short form videos, which are super addictive, just ask your wife. But they're super compelling as well. Like I do think that I've definitely been influenced by short form video where I was like, I want to buy that, where do I buy that?

And sometimes in some platforms, they make it hard to find the link because the link is somehow monetized and has to go through some kind of middle affiliate layer. So companies that do a good job on that, eliminating the middle layer, TikTok does a good job actually, will get more conversions if you have influencers or if you're doing the influencing yourself and you're just touting your product to your market or...

influencers in your market. I think that is the new SEO. I've said for a couple years now, pumping out the same content, even if it's easier now with AI helping you write it, everyone's going to be doing that.

It's not the answer. You have to come in with something that's more authentic and more meaningful and more resonant with your users. And I think that's why there's this bloom of short form video.

where it's long enough to give you the feeling of authenticity, meaning, and clarity, but short enough to not take up your whole day.

Noah Learner: I just had two or three thoughts going on firing at the same time. I mean, like the first thing is, yes, I'm so excited for everybody in the community. Cindy and I are going to be doing the SEO communities first in person meetup in Denver tomorrow. And I ordered pizza. really excited. It's not the best in Denver, but it's really good. And Cindy and I are going to be doing an interview on stage and I'm really excited because I still have questions left.

that there won't be overlap between the two. So I'm like crazy stoked, but I feel like we're getting to the end here. We got to tie this one up. Any parting thoughts?

Any soapbox stuff you want to get off here that you want to share?

Cindy Krum: The honest thought I just had is that I wish Eric was here so he could come hang out with us. Because I want to hear his thoughts. He's one of the smartest people in the industry, and he doesn't get on too many stages recently. And I want to hear what he thinks. But no parting thoughts. I'm looking forward to tomorrow. And I hope everyone's had a good time, and we didn't get too geeky here. I hope everyone has some ideas.

Noah Learner: Yeah.

Noah Learner: Yeah.

Noah Learner: I loved it. Cindy, this is so great. Okay, so for everybody, thanks for joining us for our 10th Campfire Chat. This was really, really fun. This is the first time we got to integrate the live questions that are coming in on YouTube and stuff. I thought that was fun. It made it really easy for me to just share their questions directly, not have to repeat them. And you got to read them in real time and react. I loved watching you being like, like when a question came up on screen, was awesome.

But everybody, thanks for joining us and I'm really excited to see you for our next one in two weeks with our guests to be determined. And Cindy, thanks so much for being here. Everybody else have a great day and we'll catch you soon.