How AI Search Is Reshaping Retail Media and Discovery

0:00Joe Doran: The Voices of Search podcast is a proud member of the I Hear Everything podcast network. Looking to launch or scale your podcast? I Hear Everything delivers podcast production, growth, and monetization solutions that transform your words into profit. Ready to give your brand a voice? Then visit IHearEverything.com. Welcome to the Voices of Search podcast and I Hear Everything production. In this podcast, we'll share the news, knowledge, and strategies you need to navigate the ever-changing world of SEO. Ready to expedite your company's organic growth efforts? Sit back, relax, and get ready for your daily dose of search engine optimization wisdom.

0:41Joe Doran: Here's today's host of the Voices of Search podcast, Tyson Stockton. My name's Tyson, and today I'm joined with Joe Dorian, Chief Product Officer at Botify. Welcome to the podcast, Joe. It's great to be here. Super excited to talk about the future of AgentecSearch and what's happening in the world and how retailers can get ready for it and actually help them sell more product. yeah and i i like that we're diving into like a specific kind of industry and vertical within it because a lot of times on the podcast we're talking about the industry as a whole we all know the industry is changing all these differences going on but today we want to kind of like dive more into like commerce and retail which is an area we both spent quite a bit of our career in And so to start things off, kind of like give us the lay of the land.

1:37Like how is commerce and e-commerce in particular changing? Yeah, it is changing dramatically. And I think it has a lot of leaders in the retail, e-commerce, DTC sections really scared about what will be the future of how consumers shop. right and how much control a retailer can actually have over that shopping experience when it's driven by an agent right or a bot uh that's actually there whether that is happening within the chat gpt and gemini or whether it's happening outside of that and in and around their site so this has a lot of retailers a lot of leadership at retailers and e-commerce providers thinking about their overall strategy and their overall investment so that they can control what's most precious to them is their brand and the shopping experience of their brand between them and their consumers.

2:36And I mean, we've talked about it before on the show a little bit as far as like how consumers are both going through like the buyer's journey, as well as just kind of like how people are actually like interacting with it like from what you're seeing today like is are conversions happening yet within kind of like these environments or are you seeing it more as like the discovery like research early stages of the cycle yeah i think it's mostly the early stage of the cycle or what we would commonly refer to as the upper funnel or the product discovery component. And the thing that's changing about that journey is really these chatbots, whether it's chat GPT, Gemini, or Perplexia, or any of them, they are surrounding the consumer with additional context that search engines could never provide, right?

3:32It's allowing them to go and look through their specific use case, their specific shopping use case. I need a certain outfit for a certain event. I need to have it within a certain budget. Those multiple constraints are really hard in a traditional search solution to go and solve for. The chatbots do a great job of that product discovery, narrowing it down and allowing you to go find what the full portfolio of products are out there. And what scares retailers is, are they showing up when their competitors' brands are showing up? How are they showing that their products actually outperform or are clearly differentiated and have a unique selling proposition vis-a-vis their competitors?

4:19And this is what really drives that fear among the retailer if they need to lean in or they're going to miss out on that. But one of the things that we are seeing is that the transactions, as you said, are not happening within the LLMs. It's not happening within the chat experience, even though they're trying to get there but we believe that that that that awareness that consideration that interest and intent is being formed and what happens when a referral comes to a retailer's site it's going right to the product description page and it's leading to a higher conversion so this channel looks like it can change the nature of the conversion rates right for the retailer but they but they're they're not in control of it so that that's what like oh my goodness so much promise so much risk yeah well it's like there's like a black box piece because it's like if you're looking just at like your internal metrics it's almost like the you have this impression that the buyer's journey is getting compressed down because from when someone enters the site to then maybe that transaction or that conversion happening it's a shorter time period but it's like i don't think that necessarily like our behavior as humans are changing as much as sure where we're getting the information is but it's like i'm still going to want to research and learn about different variances of products make my own opinions and do it but when you actually like land on in the site there's a very short time period and so i think it is something that's like both creating a lot of fear and uncertainty, but at the same time, it's like, hey, e-commerce isn't dead.

6:01Like people are still more than ever buying online. It's just like how we're competing and how we're being discovered is changing. Exactly. And one of the things that's really interesting, Tyson, is really when we talk to retailers, like about half of them will say what you said, which is what the industry talks about, is that the buyer cycle has been very compressed, right? You have a very compressed buyer cycle. The other half of the retailers will say, well, the old search experience was full of a lot of false positives. I would make a query. I would go to a product page and it wasn't what I wanted.

6:40I needed to go back, do another query. So it's actually more efficient and you got rid of the bad clicks. the clicks that were never leading to a conversion. So it's actually traditional search is wasteful. And, you know, the AI search is actually more efficient. Not really sure if there's data to prove which one is actually truly correct. But we do know that this zero click environment is happening. And it's like, what does it mean to me? And what does it mean to me as a retailer and my shopping experience? And how do I make sure that my products are getting the best chance to actually show up and be competitive?

7:24Well, like to that last point, like let's talk a little bit on like indexation. Because I feel like this is an area, I mean, one, if you're talking enterprise, these are substantial, really large websites. websites and I mean and if you just think about the amount of like filters and facets it's like you have this potential of just ballooning websites and then that becomes a challenge as far as like what are actual search engines now LLMs seen and discovering so like from your experience and your perspective, like, are retailers aware? And like, I guess, how familiar do you think like the industry is with like, the actual indexation and kind of crawl efficiency of their website?

8:18Yeah, I think there's a lot of retailers that believe that their website is fully crawlable, that they are indexed at a high rate, which, you know, because they've been doing this for a long time, 15, 20 years against this. But the challenge is a lot of these merchants and a lot of these retailers, just like what you said, are exploding their product portfolios, right? It is very rare that you see a tier one or tier two or even a tier three retailer with less than a million products. That means that their website is going to be well over a million pages that's there. When we look at retailers across like size and scale, we're only seeing them, and this is reported through Google Search Console, their indexation sits between 40 to 50%, which means only 40 to 50 % of their web pages are actually being indexed, let alone crawled, right?

9:18So their, and their crawlability is never at 100%. The crawlability is always really, really low, probably like maybe 10 to 20 points above their indexation rate, right? And that is really a clear indicator that you're not, you know, your products aren't even showing up, let alone, and Google's got the most sophisticated bot in the world, and they're not showing up with the most sophisticated bots and the most extensive crawl budget, right? All the chat bots and crawlers that are coming from ChatGPT and the rest of the LLMs are 20 years behind in being able to crawl with a significant amount of efficiency and scale that Google has.

10:08So they're just not as sophisticated. They have a hard time spending that budget efficiently and getting to what would be the most profitable or the most productive pages that a retailer has. So they're missing out. And this can be a major, if they fix that problem, that's a first step that they need to do in fixing the problem to drive visibility. Because they're not giving their products a chance, right? Well, I mean, I want to double down on that point because it's like 40 or 50 percent. Like that is crazy. As far as like, it doesn't matter how good your content is. it doesn't matter how fast your website loads if they're only seeing like that then that means almost half of your products don't even have a chance and let alone and i mean to me it's not necessarily like oh this is like an ai search problem it's like no this is a fundamental then of just online discovery and one of the things that like i've seen too is like you also have like varying like working with international e-commerce sites like we did a study with one of our clients who operates in a bunch of different countries each country had a significantly different percentage of indexation and some were a little stronger some were really bad but it's like to me that's like the lowest hanging fruit from an e-commerce side to actually have a significant gain if like you don't change your performance at all but you just improve that indexation like that has a potential to just exponentially increase like your bottom line yep no it it does and we see that significantly but you know a lot of our uh retail customers or a lot of the retail industry they've spent decades investing in their e-commerce site really to support human traffic, right?

12:10And the non-human traffic was always seen as waste with the exception of the Google bot. Now there's 15 to 20 other bots that matter or crawlers that matter that you need to go and support and drive against. And they're not serving the Google crawler efficiently today by itself. So this complexity is significantly increasing. So how do you go and make the investments in the infrastructure to make sure that all of your content and all your product information is fully ready and readable by these bots, right? And that is not only just like you said, performance time. So it has to be very, very fast in loading and delivering it.

12:54But it has to be able to fully read and fully render that page, right? And a lot of those pages that our retailers use are very heavy JavaScript or API laden that really cause problems for a bot or a crawler to truly fully read and render. So they give up because they're not going to, if I can't resolve this very, very fast or efficiently, I'm going to reduce the number of pages in your budget that I'm going to crawl. And that just, it just becomes a compounding effort. right and you'll flow out of the index and you will flow out of impressions you'll flow out of clicks and you'll lose organic revenue which is really really a shame right because they the retailers really need to get they really need that to power the rest of the e-commerce engine well let's let's dive in a little bit more because you made the comment a little earlier on how like the the bots and the crawlers of some of these emerging platforms companies are not they don't have the same sophistication or i mean essentially experience that google has in this space so like tell us a little bit more on like what are some of those differences how can we better understand these kind of i mean really like these bots that are growing in importance and significant to a business itself?

14:20Yeah, I think there's a couple of them. First off, just in full rendering, if you want to be able to render it, the biggest challenge is JavaScript on a page. Retailers use JavaScript very materially to have a performance side lift for consumers using a browser, right? So it is a very great tool for human traffic to increase performance and have have the right responsiveness on the page for what a consumer is actually looking for. Our friends at the crawlers and bots cannot render them. So what ends up happening is they see 30 to 40 % of the content that's on the page. They miss a tremendous amount of the content that's not really fully rendered.

15:07So what do they miss? They miss reviews. Reviews is almost always in JavaScript because it's really populated by a third party solution. A lot of times their specs and descriptions tend to get pulled through JavaScript and not kind of like directly pushed into the page like on HTML. Any offers or promotions tend to be rendered through JavaScript. So you get like 30 to 40 percent of what's on the page. And when we show this to a retailer and say this is what a crawler sees and this is what a consumer sees, it's it's they just get you know very irate because it's like oh i need to fix this because they they're missing like half the information um that's that's there so how do i go and and and do that and show that so that that's what is one of the biggest challenges that that really um is is part of the problem the other big challenge is the llms or the agentic search players like chat chat GPT and those, they operate differently.

16:12When you do a prompt into chat GPT, it will pull from its sources of index and what it actually has stored, but it's going to do a confirmation crawl against that page to go read it. If they can't read it, it's not going to be able to verify that data, and it probably will lead to, oh, it doesn't match what's in my index, so it's not verified, I can't trust it, I'm not going to show it. No mention, no citation. And so this goes differently. Google does that, but very, very specifically, Google does that really around Google Shopping, right? And when they're presenting a Google Carousel, because they're trying to confirm price and availability from what you've given to them and structured data.

16:58So common challenge, but because they operate a little bit differently, it's it's an additional complexity, right? that you as a retailer have to ensure that that you're serving these these this non-human traffic in the same way with the same care that you do for your human traffic because it it is part of the shopping experience that's it and it's super interesting and it's in some ways it's like it's it has like some deja vu elements of it where it's like i mean yeah javascript up until now Now, I mean, early days SEO, huge concern. Like that was a huge topic. We talked about it a ton, kind of started to fade away as Google became more and more sophisticated in how they were able to render pages.

17:49And like now it feels like we're kind of coming back. So it's like even though we're in this new space, in this new world, it's still like some familiarity to it. But I think your comment of like, hey, sometimes these systems or these bots are only seeing maybe 30 % of the page. I would assume in addition of like, hey, maybe there's important information that they're not seeing and it's not going to be included. There probably is also then like a perceived lower quality of content, which then could influence the amount of bandwidth that these crawlers are going to spend on those sites. So it's like, it feels like that's like a compounding issue, not just like, oh, hey, they didn't see that our product was on sale.

18:38It's like, oh, okay, maybe we would have had 60 % indexation instead of 40. Yes, that is exactly the challenge, right? So if I, as a crawler, can't efficiently crawl the breadth of your site, I'm going to continually pull back on it because there's other parts that I could probably spend on. The big challenge with anybody that has a crawler and they're trying to index the interweb, the Internet is massive. Even Google will have a very difficult time indexing the full Internet. So they have to make choices. right and they're making and they set a budget limit for what they can actually do and if you were uh have content that's important and you make it easy for that bot to go consume it with a high degree of efficiency right they're they're going to spend more money on you than on the things where it's harder to go do so you have to play this this this game because you have to work within the constraints that these big marketplaces, search engines, AI search engines are playing for us.

19:51And even though Googlebot can do a very good job, they're the best at rendering JavaScript, they would prefer not to because rendering JavaScript is super expensive. And it will eat in to part of the budget. So if you're relying on, oh, well, Google can crawl me because they're the most sophisticated, that is a false promise because they only have so many dollars to spend and a JavaScript page is going to cost three to five X to render than a fully rendered simple page for the client with all the information that they need. Yeah, and it's like that point is also like that, like the amount of cost and energy that's going into that area is so significant that it's like naturally there's gonna be trade-offs to it and i would say especially with like as the barriers to building have been decreasing you'd almost expect like the volume and the size and breadth of like i mean i don't know pages on the internet so it's like that's going to become more and more of a challenge i would assume for all these types of systems now i want to i want to double back a little bit on kind of like some of your previous experience so part of your career you spent with microsoft you worked with them from like the paid search side um i think that's probably pretty interesting with also in this new era because paid search you have a lot of like structured data elements with product feeds and And like, to me, like that is becoming more and more the competition ground, not just in paid search, but also in this new form of discovery.

21:45Absolutely. Absolutely. So within a paid search ecosystem, I think there is a lot of things that we thought about very, very clearly for our consumers, for the retailers, trying to get them the best associated, like sponsored ad to go associate with their query. and how do we go get it right around the ability to go bid efficiently, to boost against individuals that were most important, but also kind of like what was a structured element and the structured data that I would bring in that would be matching criteria to those queries as well as what we get displayed on the page and then what was the right balance between paid and unpaid.

22:32But I think the thing that becomes, and the other thing that we had is we would also, So for every one of our advertisers, we had our own crawler because we would crawl all the advertiser sites to make sure that there is a great connection between the landing page and all that fun stuff. So I think there is a big connection there. The other thing that we spent a lot of time is having the right balance between organic results and paid results. and what was the right balance between that, which I think is going to be something that we need to think about more in the world of AI and what you are thinking about and thinking about the contributions of organic traffic and search to the overall.

23:15But the product information is, I think, is critical, especially as a structured data associated with the product for making recommendations and making ranking that would be good predictors of a, in the paid search world, a click, right? And a lot of times we were just optimizing against those clicks based on what observed behaviors we would see. And we weren't taking into account the quality of the product, the fitment of the product to the prompt, other additional data that would be there that would be a better way to recommend a product that may lead to the ultimate goal instead of the ultimate goal being a click and ultimate goal being the consumer buying the product right because it was always the the that paid mechanism it's a shortcoming uh is because it's a step removed because google doesn't care if you convert or not they they kind of do because they want the return on ad spend but they get paid on a click right so that's what they do right yeah which is and that's like an interesting little nuance because it's like yeah if then that ultimately like the attribution slightly off it's like you're gonna have a different perspective on like your decision making will be different and i think like you see it sometimes within like organizations where it's like maybe an engineering team is focused more on kind of like um user behavior on it which is going to be very different so it's like they're looking more of like what happens when the person is on the site when if you're i don't know seo marketing you might be thinking a little bit more of like well how do we get more people into the site and it's like both of them are going to play into the ultimate business goal but i've seen that like conflict between perspectives that ultimately traces back to that attribution.

25:16Like what, what are you optimizing for essentially? Time for a one minute break to hear from our sponsor Previsible. So you're looking for SEO help and you got a couple of options. You could start replying to spam from agencies that claim they can get you to rank number one on Google. You can pay an hourly rate for a consultant who will inevitably nickel and dime you with hourly charges, or you can work with a cookie cutter agency to quickly launch a strategy-less project with low success rate. None of those sound very good, now do they? Well, that's where Previsible's integrated consulting model comes in.

25:50Previsible draws from a collective 40 years of SEO and digital marketing experience to unlock your organic growth opportunities. They build custom solutions that combine strategy, technical expertise, content, and reporting to effectively operationalize SEO for your business. Previsible's four-stage approach which ensures that your SEO programs thrive by starting off with a strategy-first approach. Then they support you in your efforts to create quality content, help you identify technical issues, and most importantly, they'll work with your cross-functional teams to integrate your SEO strategies to make sure that your SEO budget actually drives results, not just your agency's bottom line.

26:29So join brands like Yelp, eBay, Canva, Atlassian, Square, all who rely on the SEO consultants at Previsible. For more information, go to previsible.io. That's previsible, P-R-E-V-I-S-I-B-L-E dot I-O.

26:49Exactly. And I think that's what led us to, like, we really, we as an industry, the practitioners, as well as kind of like the industry itself, have kind of like laid in when you think about paid search, really on a session-based conversion, right? so that is a simple conversion there is no um no real look at how much does that organic traffic contribute to a paid search what other factors happen to a paid search transfer that came over to that led to a conversion how do you think about that from a multi-channel or kind of like universal measurement uh kind of like standpoint it and i think the with this transition with um AI discoverability and agent commerce and everything that's happening now, there's an opportunity to redefine what metrics matter, right, that drive the outcome, which is, especially in retail or commerce activities at all, is just getting the product sold, right, having that purchase done, and what are the touch points that kind of like really, really matter, and how do they contribute or attribute to kind of like that transaction, right?

28:00So they have a more holistic view because our world, because of everything looks at, you know, session based, you know, the paid search people don't really talk to the organic people, maybe a little bit. They don't talk to the open web paid media. They don't talk to the paid social. Right. The email is set up separate. it. So the, the, I think the, the opportunity of a lot of touch points being redefined by having AI search happening, I think is just very good and very healthy for our industry. A hundred percent. Now let's, let's go a little bit more into these kind of like agentic feeds and like, as far as like AI product catalog discoverability, Like how, how should we be looking at that?

28:50How should we be thinking about that? Whether it's from like preparation or even like emphasis and prioritization on kind of some of our roadmaps. Yeah, absolutely. I think that anyone that's a retailer should be leaning in very heavily to go deliver full, accurate, complete product feeds to the LLMs, right? any of them that'll actually take it and i think they should really double down and relook at what they're feeding into google merchant center right because google merchant center will support google shopping but it's also going to be the place where that feed is going to support ai overviews right um ai mode and gemini right from those components the thing that's changed in the world of AI search is that they want structured product data to come in that's full and complete, not just to inform a product carousel that shows up there or a product card, they're using it to inform the context in and around that response, right?

29:58Very, very important. And they're using it to validate and trust based on their crawl data that they're looking at crawling the data in association with what they have in the product feed right now the the product feed like when you look at the agentic commerce protocol from open ai when you look at uh universal commerce protocol from google their required fields as well as um um optional fields are a superset of what had been historically pushed for Google shopping or for other paid ads. Because of what I said, they need all that additional information to provide context for a proper answer. The product feeds before were really about filling up an ad.

30:50Product title, product description, price, reviews, image. They only cared about like five save fields within that feed. Now they need the rest of the information because these answer engines are actually looking for comparison. Is it the right fitment? Is it the right size? Is it going to fit my use case? Are there additional reviews there that I can see, not just the star rating, but actually the context of the review that fits the question that's actually there? So this means that me as a retailer, all that rich information that is on my product description page needs to be exposed, needs to be read, needs to be pushed, right?

31:36You know, needs to be crawlable and it needs to be pushed to these search engines or they're not going to have a good shopping experience. They're not going to show up, right, as they should. So this is really important. And the other thing I think, Tyson, that makes this really hard is that for the last 20 or 30 years, we've been adding a number of data elements to product description pages that don't sit in the retailer's product information system, right? or their information management system around the product because they've augmented with third-party data. They have additional first-party data that sits in and around it that's not in the core product information system.

32:19So it ends up being a problem where they've built this castle of a beautiful product description page. There's not one source of it. So it ends up being a huge problem for how do I go. And so when the retailers get asked for additional attributes and fields in these new product feeds, they don't know how to source it. And that and they can't source it at scale. That ends up being a challenge for a lot of these retailers. And that's why I think this is a very interesting time in and around it. And I think what those AI search engines want and what they're asking for today is going to grow exponentially.

33:04because they're going to want more and more information. When ACP launched from OpenAI, they had 75 fields. They're down to 42 because they knew that the additional 30 fields or 33 fields were too hard for the retailer to fill out, but they still want them. Can you give us a few examples of what are those attributes or what are those fields that we should be expecting and getting ahead of? Yeah, so first, the common one is review content. So they want the review content, not just a star rating. So that is one thing that actually needs to get pulled in at depth. And they don't just want, like, when they go and crawl a site, they usually see the two, three, maybe five that are displayed.

33:53They want 10 to 15, 20, right? And it's like, oh, how do I do that? I have to go, like, that ends up being a complicated problem. Q &A is one of the things that comes up time and time again that retailers just don't have on each product page. Right. And and I think a lot of commerce brands are missing questions and answers around their specific products. And this then causes a problem like shit. I don't have them. Pardon my French. We are a French company. But the how do I go produce them? For my 2 million, 5 million products that I have, I don't have staff or people to go produce that data or that Q &A that's there.

34:43The other thing that's really, really interesting since we've gotten into this business with our product called Agenic Catalog is that we're seeing a tremendous amount of missing fields. right so we'll first start with a standard product feed could be coming directly from their system it could be what they're shipping off to google merchant center today but we'll see dimensions are missing we'll see that the category is actually wrong we'll see that that that additional fields are just not populated because they don't exist in their product information system it could be the pattern that's associated with it it could be the material source it's associated with it So we end up seeing, you know, we'll go and do an analysis for a feed and we'll see nine out of 10 records or products are have at least three to seven errors in them.

35:38Right. Missing data, incorrect data, data that's inconsistent with what's on the product description page, you know, or fields that, you know, just were never popular. it yeah and a lot of this as you're describing it to me it's like you almost want to have the same attention or like thought that as seos we used to have more of just like on the page where it's like as you're describing that i'm thinking like yeah i've been in so many conversations where you're talking with retailers about just like product descriptions the reviews on the page, how you're displaying all this. And it's like, you're trying to build a more complete picture.

36:22And with how these new systems work, the feeds are then basically that same on-page experience. So it's like, yeah, if you are not populating all those things or you don't have the infrastructure in place to be populating them, then it's almost like it comes back to earlier in the conversation where you're only showing maybe 30 % of the page or the attributes of the product, which is then going to have all those other snowballing impacts that we were previously kind of discussing. So it's like conceptually as an SEO, it's almost like you have to treat the feeds as your on-page SEO efforts, the same as we've been doing on the actual page.

37:07Yeah. And one of the questions that that is absolutely right. And one of the questions that our clients ask us is kind of like, well, why is it that you're able to identify all these challenges and solve these problems? And it's like, because we crawl 100 % of your site at depth and we store that data. So we, as part of our crawling, because we're doing this all the time, we have to have a baseline of what is 100 % of your site across your full site maps. We compare and contrast to what Google says in the Google search console. We can compare and contrast what we see as being crawled by looking at the logs.

37:51So we have the full view of it. But when we look at their product feeds, I can compare it directly to the structured product data that I have from my crawl from your site. And I can tell you what's there and not there. Right. So I can show with you directly. Right. And that way, I think it's very, very telling to not only the, you know, the retailer itself, not only from the practitioner, but also from the technologist. I can show you here are the fields that you don't have that maybe you need to get your merchant team, your sourcing team to go and ask from this from whoever's the OEM is or if it's your product, what needs to be there that would be critical.

38:31But I can also tell you just the gaps that your product information system is missing or whatever is sourcing your product feeds is missing today because I see it all on your GDP. And the retailers have spent decades making those product description pages very robust. But we just have to extract that data. Interesting. And I mean, that approach totally makes sense because, again, earlier in the conversation, it's almost like the low-hanging fruit is like a lot of this information or data is existing. we're just looking then of how are we getting it and actually making sure it's being packaged up and made accessible essentially so it's like having that comparison of like we are going to ensure the crawl and then highlight the differences like and as I don't know as someone that would be on like the in-house or the e-commerce side to me it's like that has to be a little bit of a glass half full because that is a scalable opportunity to have a significant gains rather than it's like oh we need to rethink like and i don't know creating like product videos it's like that is a very slow like investment that is going to take a lot longer over time um now i could i could keep going down this rabbit hole with you but i do want to be mindful of your time here um and i want to hit on a few kind of like lightning round questions for the audience so i'm going to throw five questions at you a little more shorter form and the first question i'd like to ask is what's the most overrated e-commerce seo metric today yeah i i I think the most overrated SEO metric for e-commerce today is just tracking organic traffic as a standalone KPI and not comparing it with everything else that is actually happening in other channels.

40:48So if I'm an e-commerce provider, if I'm not thinking about my organic traffic from search and comparing, contrasting what my referral rates and my organic traffic is from the AI sources that are there, I'm missing out. If I'm not looking at my investment pressure that's coming from paid or from other channels, I'm missing out. If I'm not thinking about what's happening to my paid and organic Google shopping compared to my organic traffic, I'm not getting the full picture, right? And I think it's related to where we started, the the zero click searches uh am i getting cited and showing up kind of like what is the intent gap that's kind of like happening yeah and in a lot of ways too it's like both in how you describe that but also like in the industry as a whole i feel like we as sqos we do need to have like a wider lens of how we're looking at this because there's also similarities of like some of our activities are almost like swinging towards like digital PR type activities.

41:52So it's like, I do agree that you have to have that broad perspective to then be effective in kind of like this new era. Next question I'd like to ask is what's one AI search trend retailers are overreacting to? Yep. I think this really... When you look at AI search, one of the biggest things people will look at, the first thing they'll look at is they'll look at their share voice, which is kind of like how often I'm mentioned. But the other thing that they overreact to is prompt volume and how many prompts are actually being executed. Both those metrics are probabilistic metrics. Both of them are synthetic.

42:37No one is publishing accurate, transparent data around what is actually being answered by the search engines and what is the volume on there. There's panel providers that populate that data on behalf of a lot of the native AI kind of like visibility tools that are actually out there. But all these panels are small. They may not be directly representative of the retailers consumers because it's all internet like data you know picked up from chrome plugins or extensions or and i really question like are you really having an appropriate sample and then they extrapolate these small panels to a large population and you know when we've looked at it it's like that we have a really struggle we really struggle to go and represent to the marketer that this data is meaningful and is representative of their shoppers right and and it's like and then on top of that there's no way to connect the share of voice or your prompt volume to business outcomes where we go and show them directly it's like now your citation rate i can look at your citation rate and i can look at your crawl volume i can do a correlation if you increase citation rate i see a correlated increase in crawl volume which means that they're doing a retrieval crawl to go on and and get it i'd say i i can see that but your crawl volume is more representative of of your mentions and your your impressions than anything else right so that is one of the metrics that we and and that our industry is so hot and moving right now that that people are willing to go and they and and accept it without thinking about it i love that you brought up that point i've been struggling with that too and it's like i feel like so much time is probably wasted in that area where it's like unless you're trying to gain some like competitive understanding or maybe like managing some stakeholders in your organization it just tends to be that time vacuum and there's going to be so much more personalization and variance that it's like the integrity of that too is like i'm just not fully there on as much as i probably was yeah five ten years ago and probably more like 10 15 years ago next question um product contents or product feeds which will matter more in the next two years product feeds no no doubt no doubt i i believe like we discussed earlier uh a lot of retailers have very rich and robust product content today that sits on their pdps it's just fragmented across your ecosystem so consolidating it making sure it's full complete and accurate within a product feed is incredibly important right now one for the primary use case of like getting and pushing that structured data into the LLM so that the open AI and, and Google across Gemini and AI can better represent your products and services.

45:56But I think there's additional uses that you could do this. It's like, once you get that data like up and set and ready, maybe that data has to get fed back into your product information systems where the distribution actually serves to, to get there. It's like, how do you go and get that? I think everyone is trying to figure out how they bring a chat experience or an AI search experience to their site search. Right. One of the first things that they ask for is structured product data that is bigger and better than what they have. That is a clear application and saying I can go and help you do that and redeploy.

46:32This is where you you build once, make the investment once and deploy it in many locations to go and add to your thing. But product feed, I think, is what matters next over the next couple of years. Yeah, and it's the key to scalability in the same lens. And I think, like we mentioned earlier, it's almost like, you know, I'd encourage SEOs to think of the feeds as like the new on-page SEO and take the same critical eye to those feeds that we've been applying to when we're actually looking at the visible page. what's one infrastructure investment every retailer should make today yeah i i think every retailer should make a investment in their infrastructure about how they're specifically serving non-human traffic how do i serve the bots right now they they they'll they have a very robust cdn or content delivery network uh play they usually kind of like have multiple uh parties that are actually doing that to make sure that they deliver the best human experience possible, they're not making that investment to do it in the bot traffic, right?

47:44And those bots and non-humans have very different needs, right? And the shift has to be from thinking of non-human traffic is something that I kind of have to do, but it should be something where I have to invest here to have the best shopping experience, I have to service my non-human traffic, those crawlers very effectively, very efficiently with a high quality, right? If they think that way, that would be the one thing that they would need to do. And, you know, it can help them. It can help them drive revenue and bottom line and not just be about cost containment. Yeah. And it's the fundamental starting point that we hit on earlier of like, that is the initial like table stakes of getting the seat at the table to then compete in all the other areas.

48:38Yet we fixate on those more like visible competitive areas rather than like, well, are we even in the game? Last question I have for you. Five years from now, what will e-commerce teams wish they had started doing earlier? Yep. I think the number one thing that an e-commerce team should be thinking about and measuring today is attribution, right? How do I think about the outcome I'm trying to drive and the contribution of each of my channels to that outcome? So I'm trying to get somebody to buy a product and I should not leverage the past 10 to 20 years of thinking and think everything is based on a isolated session.

49:32Right. That if that starts with the referral traffic to that, I think are I think humans are are complex and messy people. I do not believe they just go to a single chat bot, make a request and then kind of like follow through and purchase. I think the shopping experience is messy. I think they do some things on their phone. I think they do some things on their their desktop. I think they do some things in traditional search, some things in in in agentic search. And I think they do some things. They're impacted by paid media, whether it's a TV ad, CTV ad, a video ad they saw on their phone or a retargeting ad.

50:11Right. Or something I see on site. So I think they need to lift their head and think broader, stay anchored on the purchase and and and look at the contribution or the attribution of all these efforts to that purchase. yeah and for me too like the way that i keep thinking about this is like in some ways i find myself keep coming back to like more traditional like consumer behavior elements and it's like that area within our practice to me is like always going to be the case because it's like you are still dealing with humans and yes we're competing in these other arenas and we have to think of bot traffic and all these other areas but then also it's like yeah humans are fickled like we're not going to make decisions immediately some people are going to stress over them some people are going to impulsively make them but there's that human element at the end of the day that it's like yeah we're operating in a new arena but we are still dealing with humans at the same time.

51:19So it's like those two pieces come together. And I think your point on attribution is in some ways, it's almost like the product feed where it's like you're creating those scalable kind of infrastructure that then you can build and make decisions off from that. Yep. Absolutely. 100%. Well, greatly appreciate the conversation. That's going to wrap up this episode of the Voice of Search podcast. Thanks again to Joe Dorian, Chief Product Officer at Botify for joining us. If you'd like to contact Joe, you can find a link to his LinkedIn profile in the show notes, or you can go on over and check out his company's website, botify.com.

52:04And if you haven't subscribed yet and you want a daily stream of SEO and content marketing knowledge in your podcast feed, hit that subscribe button in your podcast app or on YouTube, and we'll be back in your feed in the following day. That's all for today. Thanks for stopping by and we'll see you in the next episode.