Subscribe
Share
Share
Embed
Insider Secrets for Digitial Marketing Success
This is Performance Delivered insider secrets for digital marketing success with Stefan Horst.
Steffen:Welcome back to Performance Delivered Insider Secrets to Marketing Success, the podcast where we explore the strategies, trends and technologies driving today's marketing results. I'm your host, Stefan Horst, and today we're diving into a subject that's reshaping how advertisers approach search AI and its role in paid search strategy. We've all heard the buzz AI is taking over, manual bidding is dead, predictive models will change everything. But how much of this is hype and how much is a real lasting shift? To help me break it all down, I'm joined by two incredible guests.
Steffen:First, Nella Hopkins, a marketing strategist who has led digital growth initiatives for both startups and global brands. We had her on the podcast before where we talked about how to grow small businesses through digital advertising. Make sure to check those episodes out as well. She currently works at Microsoft where she collects and disseminates feedback from the advertising community to the product team, as well as shares accessible insights on new Microsoft ad features. Joining her is Shawn Walker, a performance marketing expert with deep expertise in navigating platform changes, evolving bidding strategies, and integrating AI into campaign planning.
Steffen:Shawn is the Chief Marketing and Strategy Officer at Symphonic Digital. Together, we'll explore how AI is redefining paid search strategy, what's truly disruptive, and what's just adding efficiency and what advertisers need to prepare for the next twelve to twenty four months. Now, before we jump into our discussion, let's frame the conversation. AI has been steadily moving from an optional enhancement to a core part of campaigns, especially in areas like predictive bidding, creative testing, and budget allocation. Today's topic, The Future How AI is Redefining Paid Search Strategy.
Steffen:We will cover critical questions such as: Are we witnessing a fundamental shift to AI first campaign management or is this just another trend? How reliable is AI powered forecasting and predictive modeling in shaping campaign success? And as platforms evolve, how should advertisers rethink budget planning and strategy? So let's dive in. Nava, what trends in AI are truly disrupting paid search versus just adding efficiency?
Navah:So I think the the biggest trend to to focus in on beyond bidding, because we're all gonna talk about bidding, is the idea of asset first creative. In order to be successful in the AI first world where creative needs to thrive in any sort of placement, what creative you put forward has to actually be able to be useful as a standalone as well as the whole creative. Now I'll shamelessly plug Microsoft launch and what we're currently piloting with Mercedes Benz are showroom ads. So there are some AI specific creatives, but for the everyday person, not that showroom ads aren't for everybody. I really would drill into the idea of creating assets in your responsive search ads, in your PMax ads, in your demand gen creative that can stand on their own without having to have any of the other assets with you.
Navah:The reason why again, is so important is we wanna make sure that we can take full advantage of all possible inventory that would be useful for us, and also shed some of those biases around what inventory is not as useful as consumer habits change, how they wanna consume information, and what sort of information they're accepting of or not accepting of out of creative being there.
Steffen:Sean, any thoughts on that?
Shawn:Yeah. I'm gonna tag along with the asset part. So I see as part of the assets being landing pages. So the one thing we're finding is AI has to scrape something. Right?
Shawn:It doesn't come from nothing. And we're seeing that landing pages are becoming more important because it scrapes landing pages to not only determine intent. That's another topic in itself where keyword intent isn't as important as just intent in general. So I'll just give one example. I like using pet stores for some reason.
Shawn:It's just easy to follow. And I got so you can say, I wanna buy cat food, let's say, and you type that in. Or someone might just type in cat food, or someone might just type in cat, let's say. Nowadays, it's more important to understand behavior, how people browse sites, even credit card information, who they're calling. There's a lot of signals to where the keyword is less and less important.
Shawn:It's the intent of all the signals that determine what AI is going to use. On top of that, going to the website, your website has to be good. Right? Quality in, quality out, crap in, crap out. That's how it works.
Shawn:So unless you have a really good landing page with solid content, you're gonna get some random results. So it's more important to have a really good landing page. And then also the creative, if you don't have good creative, if it scrapes a website and pulls creative off your site and it's not up to par, it's just gonna give you bad results. So, really, overall, it has to do with all your assets. Nava?
Navah:I'm so glad you brought that up because one of my favorite ways to actually audit landing pages is to throw the landing page into a dynamic search ad campaign and see what are some of the categories that will come up. And what's actually really fascinating is that if you don't love some of the creative that gets suggested even in a PMax workflow of putting in your landing page and seeing what creative comes back out, that's a really good sign that your creative on your landing page is not a good fit. So if you're looking for a go do, off of Sean's really useful and insightful advice, Try putting the domain into either Dynamics or Chad campaign or into a PMax campaign and just see what comes up. I'll also shamelessly plug the free tool Microsoft Clarity. You're able to see how people are behaving.
Navah:So it's just another really useful insight because if users are getting lost on your landing page and aren't quite sure where to go, there's no possible way that AI is gonna be able to scrape, crawl, and parse and understand your content. So definitely be empowered to lean into those behavioral analytic insights, but then also the free resource of what do the app platforms understand your site to be.
Steffen:Yeah. Now, when we talk to clients and then kind of identify that landing pages or the landing pages they have available are not really useful for the campaigns that we are tasked with running. We go and suggest, you know, let's build custom landing pages for that. There's always this apprehension from clients to spend money on that. Is there a way to kind of quantify what the impact with having proper landing pages is versus using landing pages that don't have the content, that kind of aligns with keywords intent.
Steffen:Imagery is not aligned in everything else.
Navah:Yeah. Sean should definitely go first since I won first last time.
Shawn:Yeah. So we just went through this process with a client. We had tremendous results. So mind you, they had a landing page that was really subpar, I wanna say. We developed a new one that had all best practices, which includes hero images, calls to action, frequently asked questions, peer reviews, things like that.
Shawn:And I don't remember the exact number, but it was very high. We're talking a 300% difference. I think they had under a 1% conversion rate, and they ended up getting to five. So it was a really heavy increase. So to answer your question, huge, huge difference.
Shawn:We're talking about the same spend being 300% more effective because you took the time to focus on landing pages. And to our first point, you can't ignore that now. Before, I think you can get away with it, but because you use AI for productivity and speeding things up and it scrapes all of this, you can't ignore it. And I'm also a fan of Microsoft Clarity, by the way. I'll give you a shout out there because there's one thing in analyzing data that comes through versus using your human eyes to see what people are actually doing, sometimes it makes a difference where these tools, while they spit out data, they don't give you the exact feel of where are people getting lost.
Shawn:Like, maybe it's the color of a button, or they don't even know something is a button. So these things are so important to assess right from the get go before you even launch.
Navah:I'll throw out there also that, for what it's worth, not that people take advantage of it. Microsoft thinks it's so important to have a good landing page that we actually offer a free landing page tool. Kind of one of those hidden features. It's a smart page builder. Yes.
Navah:Landing pages are very important. But one thing that's actually even more important is partnership with your SEO teams. For a really long time, there was kind of this divide on the landing page where PPC would kind of go into the our corner, and SEO would go into their corner. And you would make a really hard line between traffic being directed towards paid pages, paid traffic versus earned traffic, organic traffic. And what's really fascinating, at least about how Google has evolved, if you notice, ecommerce tends to have quite a number of amazing resources and tools.
Navah:And by and large, people will put their organic pages that are indexed. It's you're not dealing with no follow, no index. You're not dealing with this a really pared down content landing page. You're getting the full experience, and ecommerce has had a golden age of tools, whereas lead gen tends to stick towards those really simple templatized landing pages. So I think we all kind of have to check our biases of, yes, it's okay to be a little bit more robust.
Navah:Yes, it's okay to put more trust signals. Yes, it's okay to have a navigation bar to help not only the system, but the user understand the full context of who and what you are. But you also wanna apply some of your core transactional principles to those organic pages. So if it's a big wall of text, the average user is gonna really struggle with that. They're gonna wanna be able to have shorter paragraphs.
Navah:They're gonna wanna be able to engage with modulized content. The other thing to think about is accessibility. There are certain colors that are going to be really hard for everyone to read and process. Same thing with serif font versus sans serif font. So I think the question is less about what do we do with our page strategy at this point?
Navah:The landing pages need to meet the muster for both. It's more about how do we speak to humans, and by speaking to humans effectively, we'll speak to AI really, really well.
Steffen:I think we could probably continue talking about landing pages. Rohit, I'm just afraid we're going away from the AI topic itself. Let's move on to the next question. So with automation being on the rise for years now, AI now integrated into nearly every platform. Many advertisers are wondering if going AI first is simply the next logical step or just a passing trend.
Steffen:So is the shift to AI first campaign management required, or is it just trendy? Sean, you wanna start?
Shawn:Yeah. Sure. So my POV, it's inevitable. It's like the Internet. You can't ignore it now.
Shawn:Just the same as when you go to grocery store with self checkout, while we don't ride horses into work now. We actually drive cars. This is the era we're in. Now do I think AI works all the time is 100% foolproof? Absolutely not.
Shawn:But you need to be able to prompt correctly, make sure you have quality data coming in, but you can't ignore it. Right? And I think the big thing for me on a daily basis is productivity where if you're someone not using AI, you're going to be a tenth, you know, of the speed of someone else. Right? And that's gonna change if you're an agency or client doing in house advertising and you're going that slow versus someone else doing this 10 times faster, you're just gonna fall behind.
Shawn:Right? And I think that's the biggest issue. So I don't think it's trendy. I think it is going to be a commodity if not already. And you can see that with everything that's 's changing now where a lot of people focused on ChatGPT, buzzword.
Shawn:And, yes, we use it too, but Google is going to end up winning this, I believe. They have AI mode, which is basically their search engine results page, but kinda like ChatGPT where it gives you more robust answers. And a lot of people are switching to this kind of behavior where regular search, you have to do it's almost sad
Navah:saying
Shawn:this, but you have to think more, right, where AI does a lot of the thinking for you. Again, productivity, it moves a lot quicker. I think that a lot of people are gonna start moving this direction where you're gonna use some version of search engine results pages, but in AI format.
Navah:So it's interesting. I both agree and disagree. I agree that, yes, AI is becoming ever present, and it probably feels odd to have someone who works at one of the ad platforms with their own AI say that it's both will be all in, but then also choice. I really like that you picked the self checkout example because in almost every grocery store, yes, there's self checkout, but then there's also a cashier and a human. And even if there is a huge benefit on time, on efficiency, what have you, of going with the AI path or the automated path, there's still always gonna be some people that trust interacting with and interfacing with a human more.
Navah:And there's a certain generation of marketers where those who come from ten, fifteen, twenty years of working in the industry will probably lean more towards human and control, and where those that are up and coming will lean more towards the automation. For what it's worth, I don't see a world in which manual control is ever fully taken away. But I do see that those that opt for that solution, to Sean's really great point, losing efficiency gains. That those that lean into automation and AI are going to have bigger benefits. They'll probably have better performance.
Navah:They'll be getting out of their own way on biases on where ads serve. Like, there there'll be a bunch of benefits for it. But the platforms that win and the solutions that win are going to at least maintain even if it's an illusion of choice, the mechanic of choice, that there won't be a world in which you are fully forced to opt into AI. Now I also really love that you said AI doesn't force you to think. I could not agree more.
Navah:AI's whole mechanic is enabling us to just act. To think about, though, is that AI also has hallucinations. AI will get things wrong. There isn't always consistency. And so the ones that will win are not the ones that fully delegate out their amazing brilliance to AI.
Navah:The ones that win will be the ones who partner with AI to improve their efficiency. To put this in in real context, I joined Microsoft at the June. You have to know a lot. Like, you have to process a lot of information when when you join Microsoft. I leaned on Copilot, and within a month, it was like I'd been there for, like, three, four years.
Navah:I was able to interface just as well. Those are the start sorts of hacks that I think AI will be really meaningful for. And even in account management, able to quickly adapt tone for different creative, able to adapt for seasonality, able to adapt for visuals. But you still need to be the brilliant human marketer that collaborates with the AI. If you fully delegate, you'll end up with kind of average results.
Navah:An average might be okay, but the people that win will be the ones that are still somewhat partnering.
Steffen:Yeah. It's a little bit like when Google released PMACs. Right? Everyone was a little bit on the fences because there was very little control and an insight into what works and what doesn't work, and people, you know, like we need more information on what works, which channels, etcetera. And over time Google give us more information on on better optimizing campaigns, right?
Steffen:But I'm pretty sure still there are companies that use PMICs now fully, even rent terms are thrown in there. Right? And everything runs out of there. And then there are the ones, as you said, that I don't wanna say they take it more seriously, but they have a slightly more advanced approach where they take burn terms out. They take certain parts out of PMax because they see your results improve when they have manual parts of their campaign instead of having everything run through PMax.
Navah:And even just thinking about the example of PMax, PMACs takes a step behind exact match, exact syntax. So the machine will respect that if you picked something, that's gonna be the winner. What's interesting is that while Google has more of a stacked ranking, Microsoft, with the exception of, again, that exact match, everything defaults to ad rank. So there are some subtleties that go into all automation. So that's another area where I think people make mistakes is they apply the rules of one platform's automations and AI to another, and that is really, really dangerous.
Navah:So make sure that before you interact with any AI or any platform, you really understand how it's going to be making decisions, whether it treats signals as hard targets or as suggestions, things like that.
Steffen:Mhmm. Now looking ahead, we know the pace of change isn't slowing down. Platforms are pushing more AI powered tools every month. So the real question really is, what does the next twelve, twenty four months look like for search advertisers as platforms evolve? Nava?
Navah:In all honesty, I think the next twelve to to twenty four months is gonna be search first marketers having their moment where they stop being search first marketers. The world of living by the SERP and living by keywords and living by search is my core focus are are done. At this point, you need to be able to speak to customers at all stages of the journey and account for the fact that not every customer is going to wanna interface with text. Some of them will wanna interface with video. Some of them will wanna interface streaming audio.
Navah:Some of them will wanna interface with native sorts of placements. So I think the idea of a search first marketer is really going to erode in the next twelve to twenty four months as the socialization of search really comes to fruition.
Shawn:Right. And I would say the speed of things is what's gonna change in twelve to twenty four months because there's so many new campaign types. I mean, every day, I look up something new. Like, for example, they have this chatbot in Google that helps you set up campaigns, which is completely wild, but it's basically having someone sit next to you to help you. Now I wanna jump on the griping chain with you, Naval, because there's something that I think people are gonna realize.
Shawn:Right? The barrier to entry is so low now. You're gonna have more advertisers. It's almost like, to me, the music industry, you know, where you can just anyone who could have a app set up can create a song and put it out there. Now you have to sift through the quality.
Shawn:I think what's gonna happen is people are gonna jump in here and realize the things that AI cannot help them with, and I think people are gonna have to become more knowledgeable on those things. Like, for example, just couple things that I've dealt with recently is billing inquiries. This is 25% of my day. AI will not help you with that. You know, trying to get invoices and solving discrepancies.
Shawn:You have to verify advertisers and platforms. That's very manual. Building in bulk, which is difficult, I will say, with AI right now. And it doesn't do great with unindexed pages also. So if you have a page that isn't indexed or there's some kind of block block on it, you're in the wind.
Shawn:It's not gonna help you. So you actually have to use your brain to figure things out. People are gonna get used to bidding settings. You click the wrong button, you thinking it's doing the right thing, and it does the wrong thing. Just because you give it a CPA target or a ROAS target does not mean it's gonna hit it.
Shawn:And when you start going, why is this happening? You have to go back in your search knowledge to figure out what am I doing here. Is it actually doing the opposite of what I want it to do? And then the other things that we get all the time is connecting CRM. Right?
Shawn:So we do a lot of b to b. AI is not gonna help you with that. It is painstakingly manual to connect, like, let's say, Salesforce, HubSpot, troubleshooting, like, forms, things like that. And I think overall, a lot of people are gonna have analysis paralysis. There's a lot of information, and you're gonna have to sift through and figure out what's important.
Shawn:And you can't ask AI to tell you what's important because it's gonna tell you what it thinks is important versus what you think is important.
Navah:I so love that those were the things you picked because the agentic future is coming. And there are some people that are super skeptical about it, and there are some people that are really bought into what it's gonna represent. For what it's worth, I believe in twelve to twenty four months, you actually could have an agent if you trusted it and if you consented to sharing information with it that could do all those things for you. And you would not need to do them day to day, but that requires your consent. It requires you trusting the privacy and clean room data sort of settings of the platform handling those agents and the personalization of the agent as well.
Navah:So for what it's worth, I've gone through the campaign creation on Google, and it's okay. It's not I've had these feelings before joining, a different platform. And the reason why they're just okay is that they do a very good job of exactly what Sean said, crawling the site. Here are some keywords that seem like they're correct. Here are the images that seem like they're correct.
Navah:It doesn't actually take the time to understand your business unless you take the step to partner with it and give it a more thorough explanation. So I think if we're thinking about the agentic future, the agentic future means that you've taught an agent enough about you, your business, your strategies, that it's like you have a clone of yourself working alongside you. So if you do one click checkout or one click sign in, that's an example of an agentic interaction that you have consented to sharing your information, and an agent is doing that on your behalf rather than you having to manually type that in. So the question I think everyone needs to kind of ask themselves is, am I willing to consent to sharing that kind of information with an agent to make my life exponentially easier knowing that the price is I'm sharing that information? If the answer is yes, you are going to be in for the best time because that is where everyone is going.
Navah:That is like the race to get to agentic. If that is not exciting to you and you think that that sounds horrible, to be honest, where you'll succeed is in leaning into script. Scripts can do a lot of that kind of hybrid approach, but then also really maintain your technical knowledge of how the platforms work. Because the amount of people that will truly knowledgeable will go down with this feature because not everyone is going to need to know all the things.
Steffen:But it almost I'm kind of glad that you said that, you know, yes, you can lean into that. Right? But this is it going to create the ultra perfect campaign and results? Probably not because that still leaves room for agencies like us to help clients get the extra percentages out of their campaign, basically. What?
Navah:Well, what what'll be interesting, Sean, I'm super curious on your take on this too. It's really helpful for agencies to run tests of what would the AI do versus what did I wanna do, and how did the results differ. And that's the best way to test this. So for example, Google's AI Max, best way to test it is not to turn on AI Max in an existing campaign, but carve out part of the the campaign and start a fresh one. Or if you were thinking about Microsoft and PMax and audience ads, Carve out a portion and do it that way with Advantage Plus with Meta.
Navah:So I'm super curious actually, Sean. Do you agree that there's value for, like, the testing agency versus the AI?
Shawn:100%. And we actually did this. So there's two things here. In an article Google published, they said AI Max performed 14% better with the cost per KPI on AI Max versus not. So we did a short test, and we saw the opposite.
Shawn:And I think part of it is it takes a lot of budget in the beginning to feed enough data. It takes about three to four weeks to get there. And then on top of it, you know, we have a subsidiary at Symphonic where we work with smaller clients. And, unfortunately, unless you have enough data, we always talk about statistical significance. If you don't have 30 conversions per thirty days per campaign, everything falls on its face.
Shawn:And we've when we did the shootout, we had one client where everything was on max conversions. In some cases, it was, like, $60 clicks. Right? And we said, this is not working. We don't have enough conversions to feed the system.
Shawn:We did better with manual adjustments and cutting things out, like, just running on desktop, excluding certain, like, income. You know? And and I think it's different per business, but, like, when we do a lot of b to b, it's so niche that I don't think AI keeps up with that and the bidding keeps up with that. So we've kinda found a way around it by introducing more of a upper funnel conversion that gets more volume. But this is a mixed bag.
Shawn:So AI bidding doesn't always work how you want it to. You really have to be smart about what you're doing and continually test. Don't just trust and say, oh, turn this on. You know, flick the switch on, and everything's gonna be great. Because we've already learned that doesn't work all the time.
Navah:And that's where also bid floors and bid ceilings can be really, really helpful. It's worth noting that when you put a bid ceiling, you should not put a bid ceiling that is gonna create a pragmatic math problem for you. So if your budget can't fit at least 10 clicks in your day, odds are you're not gonna be successful because a 10% conversion rate is really, really great for non branded. And if you can't get at least 10 clicks in your day, you're being kinda better than 10% conversion rate. So don't set yourself up for an unwinnable math problem.
Navah:But the bid floors are also really helpful because sometimes the AI, if it doesn't or the automation, if it doesn't perceive that there's enough value, it'll actually underbid, and it'll cause you to underspend. So those floors are just as impactful as the ceiling. So, yes, we obviously want as cheap clicks as we can get, but sometimes those cheap clicks are in less than ideal inventory or they're super outside the target persona. So just bear in mind that, yes, anytime you add human intervention, you are limiting the probability of AI success overall. But if there are really clear signals that you know for your business are needed, absolutely intervene.
Navah:Just know that in an ideal state, the human does not touch AI. The automation is able to cook, and you have this physical significance, you have the data. But that requires a budget that accounts for that. If you don't have the budget for that and you need to make mitigations, then just bear in mind, you will not get the best possible outcome out of AI that you could have gotten had you left it alone, but you'll get the best possible outcome for you and your business accounting for the winnable and or unwinnable math problems.
Steffen:That's a wrap for the first part of the first episode. We initially had planned this to be one episode, but since Nava and Sean had such great conversations, we're actually going to have a part two to how AI is redefining paid search strategy. Now, Nava and Sean, thank you for joining me on the Performance of Web Podcast and sharing your knowledge on how AI is redefining paid search strategy. As I mentioned, we'll be back next week with part two. For everyone else, thanks for listening.
Steffen:If you like the performance of our podcast, please subscribe to us or leave us a review on iTunes or your favorite podcast application. If you want to find out more about Symphonic Digital, you can visit us at symphonicdigital.com or follow us on x at symphonichq. Thanks again and see you next time.
Steffen:Performance delivered is sponsored by Symphonic Digital. Discover audience focused and data driven digital marketing solutions for small and medium businesses at symphonicdigital.com.