⁠Dan Kabakov⁠ - Founder - ⁠Online Labs

0:00 Podcast Introduction and Dan Kabakov's Career Journey What's up everybody? Welcome back to another edition of the Ad Podcast. We have been exploring a series over the last few weeks on the evolution of the AI tools inside of the walled garden media buying platforms of Facebook and Google. 0:16 And we have another great expert joining to give his perspective today. Dan Kabakov is founder and a Google ad specialist at Online Labs. He's an e-commerce growth strategist and a certified Google partner and founder of Online Labs. 0:33 With over a decade of experience managing paid search and P Max campaigns for 50 plus e-commerce brands, Cavacov specializes in scaling direct consumer storefronts through data-driven optimization and systematic account frameworks, is the creator of the e-commerce ad system and the author of fixed and scale Google ads for e-commerce, focusing on high ticket retail accounts, attribution tracking and the leveraging of custom AI tooling alongside the human strategic judgement. 1:01 Dan, thanks for coming on. You know, I think you'll be a great person to give us the Google perspective and the new, you know, what is it AIAI Max, I think is the. 1:14 Speaker 2 Yeah. 1:15 Speaker 1 That they just rolled out. But yeah, before we dive into, you know, the prompt, maybe we could spend a little bit of time on your background. So where'd you grow up? You know, what did you study and how did you get involved in buying Google for all these amazing direct consumer brands? 1:34 Speaker 2 Oh wow. The story. The story starts somewhere around 12 years ago. I finished school and I thought to myself what should I choose as a main profession and I was always kind of curious about numbers and spreadsheets. 1:53 My friends were thinking I'm a bit autistic. So they, they suggested a digital marketing, specifically online marketing where you can actually play with numbers and analyze things. And I was, I was really, really thrilled by that. 2:10 So I started my journey there. I started working after I finished college, I started working at ad agency. And throughout the years, I kind of went on my own and started freelancing and which turned out to be an AD agency that became online labs. 2:32 And now I work in online labs. I provide consulting and management services for clients worldwide and specifically e-commerce. 2:44 Speaker 1 I love it. 2:45 Google's AI Tools: Advertiser Excitement and Concerns All right, so yeah, so the prompt that you answered unquoted and we did a little bit of a pre show as well was, you know, we we we received a sub stack article about, you know, aged media agencies that were complaining that they had, you know, basically been relegated to like log log police of like some of these AI modes inside of advantage plus and the new AI mode inside of Google and P Max and Google ads. 3:18 And, you know, for a long time, I think, you know, machine learning and AI and the ability to optimize using, you know, some of the best data in the world that these walled gardens, some of the best signals in the world that these walled gardens are able to collect was really, you know, the Holy Grail of the promise of machine learning and AI was, you know, these systems would learn how and when to serve their ads at the right moment to the right users to drive a conversion. 3:45 And, you know, maybe we had an expectation that they would start to optimize creative. So tag lines being optimized, you know, ad copy, even background images, you know, videos being changed or automatically generated resize. 4:05 And you know, brands spend a lot of time focused on their brand guidelines and how they show up in the world and how they differentiate from their competitor set and what are the hero features of their products, etcetera. So, you know, I've kind of had a large, you know, group of media agencies that have commented on these ideas from highly regulated spaces like pharmaceutical and finance, finance to other e-commerce brands and folks that have worked kind of in your space. 4:39 And I think there's kind of a a universal exception that these tools are, you know, probably should be in beta mode and shouldn't have been rolled out completely across the the open web. But instead, you know, we get to all be the beta testers in real time. 4:56 So I think that that was kind of a, a universal theme. I think most people are excited about it as long as the trade off in what brands are or, or their agency partners are giving up in control means they're getting something back. 5:13 So if there's a value exchange, that's great. But I had one client say that is it the minute they turned on AI mode, their CPMS went up by 10%. So they didn't really get anything back right. So, you know, sitting on the Google specialist side and somebody you know who works with clients in the e-commerce world who really care about low funnel metrics like cost per acquisition, ROAS, LTV, things of that nature. 5:42 Be curious, you know, how have you seen the evolution of the AI optimization, both creative and placement and otherwise evolve? And are you excited? Are you moving into it carefully? 5:59 You know, what is your big picture take? And if you want to give just the Google perspective, you know, since that's a area of expertise I know for you, we can, you know, we can frame it to that platform specifically if that works, if that's what you'd like to do. So sorry for my preamble there. 6:18 Speaker 2 It's all good. Yeah. So I, I feel kind of both excited and scared at the same time. I will be talking from Google ads perspective. So it all started in 2021 where Google first introduced Performance Max campaign type. 6:38 And this was the first campaign that actually started pushing AI and automatic settings on on us on the advertisers. Before that, it was pretty simple and straightforward. 6:54 We just had to create search campaign for what people are searching for, show the ad, target the relevant people and measure the results. And that's that. With Performance Max, Google came and said, hey guys, how about you give us the control and we'll take care of the targeting and we'll make sure to bring you the results. 7:15 So when it started, people were spending, first of all, they kind of slowly auto upgraded shopping campaigns to Performance Max without people have people not noticing and people were spending a lot of money and not necessarily getting the results. 7:35 And those that got results didn't know where it came from because it's a black box that it came from search campaigns did it came from YouTube? We don't know. There was like no proper reporting. And only now they starting after massive pushback from the advertiser, they started slowly adding features. 7:57 So performance Max is not a black box anymore. Last year we had a great feature added channel performance. I can see clearly where the traffic comes from and optimized towards it. So that's that's the story with with the P Max. 8:13 And then they came up with additional idea and they called it AI Max. And the way I see it is that Google, they kind of promise us to make the platform easy by giving the control over things like bidding, targeting to, to the platform and let the algorithm do its job, which which is a great idea, but it, it doesn't work as, as we can see at the moment. 8:48 And I bet a lot of other agencies that you interviewed kind of sharing the same opinion about it because we can clearly see it in in the results. The algorithm is simply is not there yet as where they wanted to be. 9:04 And if I'm being honest, some of these features, in my opinion, they're kind of designed to make more money for Google. 9:13 Speaker 1 By. 9:14 Speaker 2 Forcing advertisers to spend more, not necessarily getting the results. And something that you mentioned previously that we're kind of in, we are the better users of the tools. So each conversation with the Google representative starts with, have you tried this latest feature? 9:34 And this feature not necessarily is beneficial for that business. They just want to test it out and they want your, your budget to be spent on it, which is, which is interesting. 9:50 But on that, on the other hand, I do need to say that if if you feed the machine the proper data and you set it up correctly, it can bring good results. 10:06 But you need to know how to do it properly because it's not what Google recommends you to do because some of the features could be relevant for you. It's, it's somewhere in the middle ground. And this is basically our job as advertising managers to to be there for the client and be kind of a third party person who who gives an objective opinion about this. 10:35 Speaker 1 Yeah, I think that's great. 10:37 How AI Affects Media Buyers and Agency Relevance I think, you know, the, the, the identification that some of these features are, you know, they don't just optimize results, they optimize profits for Google. And you know, that's probably true of some of the other buying platforms. 10:54 You know, when you have a black box and Costco up, you know, and you can't see what happened and why And all these, you know, publicly traded companies, you know, whether it's better or Google, they're investing massive amounts of money in training AI models, building data centers. 11:13 You know, Meta is building the metaverse and building wearables and you know, Google's building self driving cars. So these guys have huge, huge moon shot investments out in the market every single quarter. And most of the time those, you know, new Frontier projects are being funded by the existing business, which is advertising. 11:36 So, you know, it doesn't shock me that sometimes, you know, it's not always in the best interest of the client necessarily. I guess there's a, a tug of there's a, there's a tug and pull, right? Because these platforms need to make more money and they need to, but they also need to deliver for clients or they'll go and try and find other channels that work. 11:59 You know, one of the things that I have noticed is that, you know, what you mentioned is, you know, let's say like over the last 15 years, there was a media buying speciality that was really focused on understanding how to set up multi variant tests within these ad buying platforms. 12:22 You know, for your side of the house, doing a lot of keywords stuff, really thinking about, like you said at the top of the episode, what search habits would be associated with what keywords and how do we find those audiences? My side of what I've been focused on over the last 15 years is more on the YouTube side. 12:43 So thinking about, OK, well, you know, I've got a piece of creative. I think it's relevant for a bunch of different audiences. So rather than do the new way of just like setting up one campaign, an objective, starting an end date and then like loading all my audiences into one campaign, I'm going to do multi variant testing. 13:05 I'm going to, I'm going to set up a bunch of campaigns with audiences as the key, the key differentiator between the campaigns. And then I'm going to track my performance and the audiences that are delivering good row as and good cost. 13:20 I'm going to increase my bids and budgets on those to deliver more against them. And then the ones that are kind of dragging down the campaign, I can either turn them down or turn them off completely. And then now I know where to focus and I can actually use that as a feedback loop to inform my strategist on what to do next. 13:38 What type of creative, you know, even product stuff, it feels like the platforms, not only black boxing, you know, the, the, the targeting, but really just dumbing down the ads managers themselves to where like almost anybody could, you know, set up or set up a campaign has really to some degree been purposeful. 14:02 But it is also to been to maybe relegate the you know, the, the ad buyer who knew how to not manipulate the system, not game the system, but knew how to drive efficiencies and claw back some of that increase in price and cost for on behalf of their client. 14:20 Do you think they are trying to make the media buyer irrelevant or less effective so they can kind of slowly increase costs and, and, you know, make it harder for brands to, you know, optimize their spend essentially? 14:38 Speaker 2 Yeah, that's a, that's a great question because here's a paradox. They they always try to make the platform easier to manage. But what I've been seeing for the last five years it that the platform became so much more complex from what it was before. 15:00 I can, I can see it in, in the time that I'm spending on average because I'm tracking time with each of the clients. So I can see, I, I checked, I think last week, I looked at my schedule, how, how much time it took me today versus 3 years ago and it, and the difference was like almost 2 times more. 15:23 So that's, I'm, I think that they actually bring more complexity to it. And it's not necessarily coming from the platforms because we had we had the privacy issues, the the cookie policy, all, all, all of those things affected the platform. 15:46 I know Meta removed some of the targeting features because because of that. So a lot of things happened, but overall, it became much more challenging than it was before. And I think that the marketing platforms are rolling this extra work on us, the advertiser, because if, if, if think about the, the, the fees we're charging the almost the same fees plus the maybe the inflation, but the work became much, much more harder. 16:18 It's, it's not just going into let's say Google Ads platform. You need to go to client's website and you need to check that the consent is turned on and then it's synced with Google Analytics and Google Ads and only then you have the data. 16:35 But it's not the whole data 'cause there is a discrepancy. There's so, so much more complexity now. 16:42 Speaker 1 Yeah. I think the other thing that I think is interesting is, you know, you mentioned more work and more complexity being pushed off on either the agency or their brand of the brand directly. And then also I think in the same time you see like a, a push to offer less service. 17:03 So, you know, people who used to have Google reps don't have them anymore. People used to have Google ad OPS folks that help them, you know, QA, Pixels and install things. Those people are gone. The account managers are gone. You know, maybe if you're lucky, you get a call from some guy in Brazil once a year, like you said, to just tell, maybe to say, hey, you know, you should try all this new shit so we can test our test, to test our test our systems. 17:31 And you know, I've, I've gotten those calls and followed their, you know, their suggestions because I thought that they were the experts and, you know, just ended up spending a bunch of money. No big deal. You know, that's the business that we're in. But yeah, I think it's interesting you mention more complexity, more time, but also less support. 17:51 AI Creative Optimization and Google Policy Violations So I'd be curious if you had a point of view on the creative optimization, because I think that was a thing that, you know, raised a little bit of red flags with brands and agencies. You know, I think we all have the expectation that AI is being applied at the placement level and it's hopefully helping us, you know, find the right people at the right moment for the right cost and the right conversion. 18:15 But, you know, brands get a little bit more, I just think a little bit more uncomfortable with the idea that, you know, things that they wrote are going to be changed. You know, pictures that they uploaded might be manipulated or changed, videos might be changed. 18:35 But, you know, some performance marketers say, well, if it's in service of a conversion or a lower funnel metric, then who cares? But then, you know, Coca-Cola might have a really big problem if you change their Coca-Cola red to 5 shades different. 18:53 Even if it was, you know, in service of a conversion, they're a lot more brand sensitive. So what's your opinion on the creative optimization piece that seems to be happening more and more often? 19:06 Speaker 2 So for creative, we have good news and we have bad news. The good news is that there is always an option to opt out from most of the settings that alternate the the images of the creatives, auto apply settings, things like that. 19:28 The bad news is that most of them get switched on by. 19:34 Speaker 1 Default when you. 19:35 Speaker 2 Create accounts and if you don't know where to look, you, you'll be in trouble. And I need to tell the the viewers that everything that is created within your account within the brand's account is the brand's responsibility, not Google's. 19:57 It's written in their terms and conditions, which is very, very important to know legal wise. So you need, you need to know where the settings are and how to switch them off. 20:12 I had a case. Which was kind of long time ago before, before AI was introduced, a client had auto apply settings enabled that kind of let's Google scan the website and take any images Google likes and insert it into that. 20:38 So what they did, they found an image that contains bunch of text in it. And the moment they added to the this image to the creative, they flagged the creative because you cannot have text on the image. 20:55 And they paused the creative and this happened automatically. All all all of these steps were done automatically without clap even knowing. And kind of annoying thing in in Google is where when you violate specific policy they not always notify you about it. 21:18 You need to go into the settings and check policy manager all the time and monitor it 'cause otherwise you can even get your account suspended because of that. 21:28 Speaker 1 Yeah, that's crazy. Yeah. I think the, you know, you mentioned something important especially for regulated industries or even claims, which is, you know, Google's not responsible, you are. So if they change something and put a financial services company or a pharmaceutical company in violation of any kind of regulation, you know, the the advertiser is ultimately responsible. 21:54 These settings are default on. Some have even suggested that when you turn them off, they magically turn on occasionally. So you have to kind of keep an eye on that part of it as well. But yeah, so I think it's, you know, it's buyer beware. It's like just, you know, that's. 22:09 And I think as people continue to say, Oh, well, AI is going to take jobs in the space and things of that, you know, it continues to be I think humans in the loop for this QA process of making sure things are working the way that they should. 22:28 Makes more and more of a case for why agencies and having third party companies advocate on behalf of clients across all platforms that they're buying against, not just Google and Facebook and some of the ones we've talked about today really is the place of the media agency and why it continues to be important. 22:48 So awesome, man. Well, appreciate your hot take on, you know, that topic. So maybe we can spend a little bit more time talking about you and your specific work at what you guys do at online labs. 23:02 Top 3 Hidden Leaks in Google Ads Accounts So you've got this philosophy called fixed and Scale in your book. You advocate for it. What are the top 3 hidden leaks in Google Ads accounts that brands must fix before throwing more budget at a campaign? 23:19 Speaker 2 Well, first of all, the conversion tracking. Conversion tracking is kind of the most boring thing to do in the account and I I personally don't like it. I know that my colleagues hate it as well, but if you don't set it up properly you won't be able to measure the results correctly. 23:43 And from from what I see, a lot of times agencies just don't do a proper job if you want to track let's say the results. Let's say you advertise the Shopify store selling products. 23:59 Usually they track purchase the end goal and they track add to cart and check out kind of the the main events, let's say. But you can do so much more on top of that. There are so many, many events that are part of users buying journeys such as newsletter sign up. 24:23 If they initiated chat conversation, did they, Did they call before? Did they let's say applied coupon there? There's so many events that can help you send like the first signals of buying intent. 24:43 And when I audit accounts, I audited dozens of accounts throughout my career. I go into conversion settings. There's always something wrong not set up properly if it's the click through window, if all of a sudden you have all conversions set up as a primary instead of just one conversion being primary. 25:07 So that's, that's the first thing. The second thing is what we talked about previously, which is policy violations that and The thing is that you can have a clean account running proper ads and you can still have policy violations. 25:30 How if you have an old inactive campaigns that have inactive ads showing something that might violate Google's policy, even if that was created like 10 years ago, it will trigger account violation. 25:50 And not, not many advertisers know about this. But if you before you go to policy violations, you need to filter out and select all campaigns, including the paused ones to see those violations that trigger the account. 26:11 It's it's kind of like under the radar, but it affects the account. I wouldn't say necessarily it will cause a account suspension, but it affects behind the scenes. Google doesn't say it directly, but they do have a scoring system for accounts to see the quality of the account. 26:30 So if if your account has many policy violations, it can cause a higher cost per click. 26:40 Speaker 1 Yeah, 'cause you're, you're, you're, you're essentially a higher risk account and they have a general assumption that the things that you're putting out are, you know, risky. So it should cost more and it potentially creates bad user experience for their users. 26:59 So, you know, why not make it more expensive and kind of just not kill you outright, but kind of kill you by 1000 cuts essentially? 27:11 Speaker 2 If you're in a very competitive market, then these things affect your, your ad positioning compared to other companies and affect your results overall. So I, I really like to keep the policy violation window clean without any violations. 27:32 And the last thing will be, in my opinion, product feed, especially for e-commerce, because product feed gives Google ads the data it needs to display the products in the proper way. 27:51 And nowadays it uses this data to target the right people because if we look at key Max or even shopping campaigns, we cannot target keywords. How does Google know who is the user who is going to see the ad by the information it gets from the product feed? 28:14 And it's not just the title and the description which are important, but there are little things that called attributes. And a lot of store owners from my experience, they, they really neglect Google Merchant Centre, which is a separate platform, but that platform directly affects the, the performance and the results that they see in shopping campaigns and Google ads. 28:44 So products feed is kind of very neglected. And it's, it's a good opportunity because not, not a lot of advertisers are doing this properly. So if you take your product feed and you optimize it well, you will have a competitive advantage for sure. 29:05 Speaker 1 Wow, those are great. Those are great foundational things to think about. So thanks for those hot tips. 29:13 Campaign Strategy for Luxury vs. Impulse Purchases All right, you've helped scale high ticket jewelry e-commerce brands, you know, 1000 to $10,000 AOV to massive ROAS. How does campaign strategy and our audience targeting differ when selling luxury items versus impulse CPG products? 29:33 Speaker 2 Well, first of all, the sales cycle, it's much longer. It kind of reminds AB to B industry. But in B to B the deal from the moment the user clicked on an ad until he closed the deal, it can be even a year. 29:55 So for for jewelry, it's usually a few weeks to a few months. And that is why your sales funnel is much more complex. If we compare it to a fashion jewellery, because fashion jewellery the client needs just a little gentle nudge and he will purchase the item for $100. 30:19 If we're talking about $10,000 ring, then the consider consideration window is much longer. He also wants to check all the alternatives, maybe use some AI to compare the options. So the sales cycle I I make it as as long as possible. 30:43 Technically speaking if we go into conversion settings in in Google ads, the default click through window is 30 days and I like to put it to the Max 90 days. If there was an option to make it longer I will do so. 30:58 And server side tracking is very important because server side tracking saves the data about the user or on on our server. So even if the cookie get lost somewhere or he switches the devices, it will still be there. 31:23 It it makes it makes the tracking much more accurate. 31:30 Speaker 1 Yeah, that's interesting. So having deeper targeting windows, whether that is how you store the data so you don't have any loss of signal when things happen that normally will cause a brand to lose signal and having the longest potential attribution windows are key for that category. 31:56 Evaluating Cross-Channel Performance Amidst Privacy Changes With privacy changes and 3rd party tracking depreciation, multi touch attribution gets messy. How do you evaluate true incrementality and cross channel performance when the primary analytics platforms disagree? 32:13 Speaker 2 Wow, that's that's a very good question. It's it's such a common problem for for everyone. When I talk to my colleagues, there is no one-size-fits-all here. 32:30 We use different, different methods for first of all, we use additional third party attribution tools like for example, N team that can give us an X additional source of data so we can compare it with the platform. 32:49 Because we know that platforms like Google ads or Meta ads, they're kind of subjective when it comes to attribution. You can see one sale being attributed to both of the platforms and they kind of fighting with each other who who gets the credit for for that client. 33:07 So you need some kind of a third party tool to help you kind of make a proper decision with that. With that said, there there is no 100% attribution. You should always take into account that some people use ad blockers, and when it comes to fine jewelry, there are actually plenty of cases where the the the fiance chooses the ring for herself. 33:41 Speaker 1 On. 33:41 Speaker 2 Her device and then she shows to her boyfriend. That's the ring I want and then he goes from his laptop or device and buys the ring. So we have two two different users on the website and we are not able to connect between them. 34:01 Speaker 1 Yeah, unless, unless you somehow have household ID, it's hard to make that attribution case. 34:09 Leveraging AI for Tedious Tasks, Not Decisions Yeah, you build custom AI tools to support your manual search work floor flows. What specific tedious tasks have you been able to automate? And what should media buyers never delegate to AI? 34:26 Speaker 2 So my, my rule of thumb is use AI as a assistant that can help you read the data, analyse the data, never make decisions and God forbid never make changes in the account. 34:45 That's why when I, when I see some kind of SAS tools pop up that offer you a full automated Google ads management for, for a few dollars, I, I get very, very skeptical because AI is a great tool. 35:06 I'm not against the I, I have bunch of tools and I will, I will tell you about them in a bit. Just I just want to highlight that it's not, it's not currently able to make proper decisions. 35:24 It makes a lot of mistakes. When I look at the results, it sometimes can find things that I may miss and that's why I use it. In terms of use cases, if I have account data and I'm looking at all the search terms that triggered the ads in the last year, sometimes I can see about 280,000 different search terms. 35:51 Before AII was never able to to go over the this list and even if the list was a few hundreds, it still took me few hours to kind of understand what's the logic, what do they share in common. 36:09 Now I can do it in minutes with AI. It can analyze the data and from this 280,000 variations, it can find some common patterns of of the search term variations that kind of shared among all the customers. 36:27 And based on that, I can take it and build a brand new campaign that will give those keywords, let's say a higher bidding. So that's that's one use case. Second use case kind of obvious is the the reporting. It can analyse the data and it can prepare the reports for clients externally and also internally. 36:50 So when when I start working at the at the beginning of the day on client accounts, first of all, I already have AI analysis ready and I use that analysis to see if there's something very urgent that meets my attention so I can prioritize my day. 37:14 If I see that there was for example a policy violation triggered, then I know that this account needs deserves my immediate attention. So it really helps me with prioritizing tasks and also helping me find things that I might have missed otherwise. 37:33 Speaker 1 That's awesome. So it's I love that idea of, you know, kind of like the prioritization, you know, it's running 24/7 so it can kind of organize your task list and help you force rank the things that are most important. And then, you know, also finding like the missed opportunities. 37:52 Yeah, I think as a business development person, you know, I generally am really good at top of funnel, you know, getting to meet people, but the follow up piece is hard. And these agents keep you on task and make sure that every, you know, positive conversation you have, you're taking it to the logical conclusion. 38:12 So I think that's smart. 38:15 Speaker 2 Another interesting automation that I have regarding the clients is a full final analysis. What I mean by that is that let's say I manage Google ads for for the client, but at the end of the day people click on ads and they land on clients website. 38:37 I want to make sure that everything on the website is working properly. So we were not wasting money on on on the traffic. So for that case I'm monitoring the client's website and I'm using different methods for that. 38:57 First of all I'm checking that the website is actually live and there is no like server issue with it. And then I check the website speed, make sure that it loads fast. And then I let AI imitate a user that landed on the specific page from an ad. 39:20 And it tries to go through the whole process until the payment page. And if it encounters some kind of issues, let's say the button doesn't work or all of a sudden you have some kind of a pop up and it prevents you from going to the next step, it will notify me. 39:36 So this really helps me make sure that not just the Google Ads side is covered, but also the website. The main thing is working properly. 39:48 Speaker 1 So you essentially use it as kind of a ongoing QA assistant as well, just making sure everything's being in your funnel, yeah. 39:56 Speaker 2 Exactly. 39:57 Speaker 1 Very cool, all right. 39:59 Optimizing Branded Search Without Cannibalizing Traffic Defending branded search keywords can often burn budget or artificially inflate ROAS. How do you advise clients on structuring branded search spend without cannibalizing organic traffic and has this changed? Now that people are spending more time in chat bots before they come to websites. 40:24 Speaker 2 Brand search traffic is usually not that expensive as as people may think. With just a small budget, you can cover all the people who already searching for you and even if you have competitors targeting you, Google will still give you the priority and cheaper clicks because they know that's that's you, that's your brand name. 40:49 On the other hand, if you avoid brand search at all, then competitors might steal your customers. They might appear in the in the 1st result in Google search before people scroll to your website. 41:05 So I, I recommend always having a brand search campaign when it comes to measuring the results. If, if a client tells me to exclude or exclude brand search traffic, I already know that this client has the knowledge and experience in the platform. 41:30 We we keep that campaign, but we keep it as a separate, We don't measure it with, with the rest of the campaigns. And that's how we can keep the reports clean and transparent. Now Speaking of settings, if if you have AP Max campaign, you need to make sure that the brand terms are excluded and from other campaigns as well basically. 41:59 But in P Max it's kind of neglected. If you don't do it then kind of P Max competes with the brand search. Usually it will still the attribution from the brand search cause Google kind of tries to give more attribution to P Max 'cause that's their golden child let's say and and you end up with inflated results. 42:27 And when I take over the account, I, let's say I take an existing account and I start working with client, I exclude that traffic from the P Max. And then all of a sudden the the client sees a drop and gets spooked because he thinks we, we, the results are getting bad. 42:46 But in reality, the results are just getting clean. We we're cleaning the mess. 42:52 Speaker 1 Yeah, that makes sense. All right. 42:54 Paid Search Evolution with AI and Chat-Based Discovery As search interface evolves with AIO reviews and chat based discovery, how do you think paid search buying adapts over the next three to five years or just in the immediate future? I know it's hard to predict even one year out at this point. 43:11 Speaker 2 I think to answer that, we can go just one or two years back in time when everyone online was saying Google is doomed because we have now ChatGPT and nobody wants to use Google search anymore. 43:28 Everyone has ChatGPT, it's much easier. And then Google had a very, very nice comeback with with Gemini and they proved that first of all, they can, they can create a very good AI technology and the search is not going anywhere. 43:53 It just evolved. We have AI overviews, which which is very nice thing. It's kind of like a free Gemini in the search results. And Google's revenue, if we look at the Google's revenue from this first quarter this year, it grew by two digits. 44:19 Speaker 1 Yeah, it's, it's interesting. I think if you think about search behavior, like how people shop, you know, what's the most important outcome for doing a Google search? It's like, what are you going to buy? So, you know, I don't buy a lot of stuff because my wife tends to buy most of the things. 44:37 But every once in a while, if I'm like not patient and I don't want to wait for it to get done, I'll, you know, I just bought some JBL headphones that are up here somewhere, you know, maybe right there. I just, I just went on to Google. 44:52 I typed in headphones and I found the one that was the cheapest from a brand that I kind of knew and I wanted to buy directly from the brand versus a third party site. So, you know, I bought these JB LS and you know, I bought some scissors the other day because like somehow like all my scissors went missing in the house. 45:11 Same, same, same basic process. Like we're spending so much time in AI, but it's for information and research. And when we're actually ready to buy, it seems like we default to the behavior that we know, which is Google something, right? 45:27 We don't buy and chat. We don't buy and ChatGPT now 1 interesting thing, you know, in the last I've only been kind of in the agentic, you know, work world for, you know, coming on less than a week. So I've been really heavy on Brockbot over the last week and it's really revolutionizing things from a business perspective. 45:47 But when I noticed that was interesting is, you know, Brockbot came back to me and it said, hey, you're trying to accomplish this task. Basically what I wanted to do is I wanted to interface with Grokbot through an e-mail. So I wanted to have like Dylan Conroy, assistant at my G at my e-mail. 46:04 And I wanted to be able to just like in the emails, be like, you know, my assistant will send you times for the meeting and then have Grokbot be the assistant in the background. And then it told me the way to do that was to install this particular piece of software. 46:20 And you know, I basically was able to buy the AAS directly through the Crockbot interface. O If you think about it, that was my first experience with a jet to commerce was Crockbot. I allowed an agent to surface a piece of software that it thought that I needed to perform a task. 46:38 I gave it my credit card and it went and bought it and then registered for the software. So, you know, for me, agentic is hat agentic commerce happened in the business world before it happened in the professional world. But I imagine as people like me, professionals who are using agents in the work world and get used to handing it their corporate credit card to buy software on their behalf are going to be a lot less threatened by the idea of handing your credit card to Meta AI to go buy toilet paper, right. 47:08 So I think, I think, you know, that maybe a Genta commerce starts to grow faster, you know, at a faster clip. I think my other my other observation with Google recently is that I was trying to complete a a presentation and I needed you know some nano banana credits or whatever in my Google account. 47:32 So I upgraded to Google AI on my personal account. It immediately credited back the $2.00 a month that I was paying for storage. So I basically traded $2.00 a month for Google storage for basically like Google AI and then and then it also gave me ads free YouTube. 47:52 So the next time I went on my YouTube account, I no longer am seeing ads. So, you know, Google is pushing AI on its customers and then allowing customers to opt out of advertising as part of, you know, subscribing to these freemium or low lightweight versions of their AI suite. 48:11 So it's kind of interesting that subscription revenue is cannibalizing advertising revenue from the perspective of like who are their early adopters with AI first, You know, so as more and more people say, Oh yeah, $2.00 a month for Google AI and I don't have to see ads on YouTube, that was just a a benefit. 48:29 So I wonder, you know, as they're looking to, you know, get more people, convert more people to paid users, how does that change, you know, their their economic model with customers and how does that change, you know, their advertising business or their reliance on AD dollars versus direct to consumer subscription dollars? 48:49 So yeah, it's a brave new world. 48:52 Episode Wrap-up and How to Connect with Dan Well, Dan, appreciate you answering the call. When we put out this thought series, it was great to talk to you in the pre show and get to hear a little bit more about your speciality and get some advice in your take. If folks want to learn more about what you're doing, following you with your thought leadership, where do you want some people to stay in touch? 49:14 Speaker 2 They can reach me at online labs dot IO. 49:18 Speaker 1 Awesome. Well, thanks again for spending some time with the audience today. For those of you at home, if you're seeing the show for the first time, don't forget to give us a subscribe or a follow. If you're listening to the show, a honest review on whatever podcast platform that you're listening on is how we find a larger audience organically. 49:34 So don't forget to do that. And last but not least, if there's a a e-commerce marketer out there, maybe specifically somebody in the high ticket space, somebody that's selling expensive things online and need somebody with deep industry of speciality and understanding how to build out larger attribution windows and help convert those types of consumers down the funnel. 49:58 Or maybe somebody in media buying land who wants to get deeper on this topic of AI optimization inside the walled garden ads managing buying ecosystem of Google and Meta. Or maybe just somebody who you think would enjoy the show that I had with Dan today. 50:16 Don't forget to share. And with that, I will see you guys all again shortly. Thanks again, Dan. Appreciate your time.