Use Cases
    Capabilities
    Roles

How Advanced Auto Parts Leverages Generative AI for Digital Marketing Success

Nov 27, 2024

Watch or Listen on:

Episode Description

Learn how Katlyn Edwards from Advanced Auto Parts reveals how they’re using generative AI for PPC and SEO.

She shared the tools, tactics, and strategies that are helping her team stay ahead in a competitive market.

This episode is packed with insights on using AI for scale, building customer-centric campaigns, and preparing for the future of search marketing.

What You’ll Learn:

  • How to integrate AI into PPC workflows for scale
  • Preparing for Google’s AI updates for SEO
  • The role of human oversight in AI-powered content
  • Future trends shaping paid search and digital marketing

Episode Takeaways

Leveraging Generative AI:

  • Generative AI is used to streamline repetitive tasks such as generating product descriptions and titles, enhancing productivity without fully automating creative processes.
  • Human oversight remains crucial to ensure AI-generated content aligns with brand voice and maintains quality.

Impact on SEO and PPC:

  • In SEO, generative AI assists in creating content like meta descriptions and blog outlines but requires careful review to meet Google’s quality standards.
  • For PPC, generative AI can rapidly produce ad copy and campaign content, allowing for quick adjustments and testing.

Future of SEO and Generative AI:

  • The shift towards more interactive AI-driven search experiences, like Google’s Search Generative Experience (SGE), may reduce traditional blue link SEO’s impact.
  • SEO may need to adapt by providing richer data to feed AI systems, ensuring content is recognized and correctly summarized by AI-driven search platforms.

Episode Transcript

Frederick Vallaeys: Hello and welcome to another episode of PPC Town Hall. My name is Fred Vallaeys. I’m your host. I’m also the co founder and CEO at Optmyzr, a PPC management software. So for today’s episode, we have Katlyn Edwards from Advanced Autoparts AAP, and she has been working in SEO, PPC, at agencies, in house, and of course generative is part of the thing she works on.

So we’re going to hear from her today. how generative is changing the game in the organizations that she works at, how they’re using it, how they’re leveraging it, but also really think about what is the future of SEO and PPC when generative is changing the whole consumer experience leading up to purchasing something.

So with that, let’s get rolling this episode of PPC Town Hall.

All right, Katlyn, thanks for coming on the show. Good to meet you.

Katlyn Edwards: Yeah, good to meet you too, Fred.

Frederick Vallaeys: So Katlyn, tell us a little bit more about who you are, what you do at Advanced Auto Parts, and your history in digital marketing.

Katlyn Edwards: Yeah, so at AAP right now, I focus on paid search, mostly on paid shopping, but that actually isn’t my predominant background.

I’ve only been in PPC for a little over a year now. Actually, prior to that, my previous life was in SEO. And I worked in both B2B, B2C, both in house and also in an agency. So I had experience on, on both sides of the world.

Frederick Vallaeys: Great. So I think you’ll bring out a lot of different perspectives here from the various hats that you’ve worn.

And I’ll certainly want to hear about how SEO and PPC kind of interplay these days. But but let’s start with the big question that’s really on everybody’s minds. These days is how is generative helping us do our jobs better? So the fear is obviously that it does the jobs for us, but I think you find some ways to actually leverage it, right?

So let’s talk about.

Katlyn Edwards: Yeah, for sure. There definitely is. And I think it’s a valid fear that eventually I is going to take over our jobs and then take over the world. But as it is now, I think it’s more of a helpful tool than something that we should be afraid of. Like just for me on the day to day, it’s.

It’s obviously a great tool just for doing anything as simple as generating product titles and descriptions so that you don’t have to be going in there and typing out thousands and thousands yourself. But there’s also for any copy if you need that. You will obviously need to go through and read it.

I’m not sure if you’ve seen all of the LinkedIn posts about like, here are the easy to spot generative AI, especially from chat to PT terms, but like, we can’t just be cutting and pasting that in and then for strategy ideation as well. It’s great for that too. Especially if you’re able to properly communicate to the AI, you know, here’s the customer that I’m looking to target.

Here’s the voice that I’m looking to master. I think that’s probably the biggest challenge in using it is making sure that you’re putting in the right information so you actually get something quality out that’s not just too general.

Frederick Vallaeys: Yeah, makes sense. So let’s unpack that. You kind of covered three areas there.

I’m going to start with the first one, which was you use GPT for text generation. Now, we’ve all gone to chat GPT. We can easily do that. But in the types of accounts that you manage, there’s obviously an element of tremendous scale. There’s so many auto parts. And what does that look like? I mean, are we looking forward to that?

Sort of like boilerplate components that are used throughout, or do you take it to that next level and you scale it up for all of the different components that that you sell or parts that you sell? And how does someone do this at scale?

Katlyn Edwards: So I can’t speak to the tools that we’re using specifically here at AAP, but like, theoretically, if we’re looking to do anything at scale, we You might use something related to like the feed management that you’re working with.

That way you already have access to all of that product information. It’s a very different beast if you’re working at, like, an enterprise size company versus if you’re just contracting and working for a site that has less than a 1000 URLs, because then, like you said, you’re working at scale. So I can’t really give too much more detail there, unfortunately, just because.

We like to keep our cards a bit more close to the chest, but I hope that answers the question.

Frederick Vallaeys: So and then that second one you said there’s some markers that you pick up on that tell you it’s written by GPT. For the people who haven’t seen those LinkedIn posts, tell us more about what those markers would be.

Katlyn Edwards: Oh, my gosh, I knew I knew you were going to ask me this and I was like, I should go back and look at those LinkedIn posts. So I know exactly what those words are because I couldn’t remember them off the top of my head. Unfortunately, I don’t remember them off the top of my head. It’s words like decisively or things that you wouldn’t normally use in human conversation where you look at it and you’re like, wow, this person must be really smart because they’re using this, this large word.

But in actuality, it’s just because AI doesn’t know exactly how we talk and it just likes, it’s. Using big words.

Frederick Vallaeys: Yeah. One that I always see is delve into.

Katlyn Edwards: Yes. Yes. I like

Frederick Vallaeys: that. Now, I mean, so what strategies do you use? So you said there’s human review, right? To make sure it’s not using those words, but have you used custom GPTs or custom instructions to sort of proactively prevent those words from even occurring in the first place?

Katlyn Edwards: Yeah, so, so again, I can’t speak to like the specific tools that we’re using. But typically, we will rely more on human review once we’ve generated the Like if we’re doing product descriptions, then we’ll do like spot review for it. And obviously that puts in a lot more legwork. So if you’re worried about job elimination, there’s always going through and reviewing the output of the AI.

Frederick Vallaeys: And then that third component that you were talking about was, was the strategy ideation. So walk us a little bit through the process that you go through with do you use a custom GPT? Do you use just chat GPT? Have you tried Entropiq, Claw, Claw, Gemini? What do you like for strategy?

Katlyn Edwards: So, I usually prefer chat GPT.

Jim and I, I have found to be, just from personal preference, I, I try to explore outside of Google products, and I really like encouraging chat GPT to grow because it’s nice to see that taking up more market share. I have worked some with Oh, my goodness. And now I’m totally forgetting the name.

I have worked some with other AI generators as well. But chat GPT has been my like comfortable go to from the beginning. So I haven’t reached much outside of that. When I’m doing strategy generation, usually the way that I will want to approach it is by making it really clear, like what I’m trying to achieve, because there’s the there’s this great engineering quote of garbage and garbage Garbage out.

So if you’re not specific, like, here’s exactly the customer voice. Here’s an example of like previous copy. We’ve used. Here’s exactly the target audience and then giving a lot of detail and information there. Then you’re going to get something extremely generic and like, usually the 1st output you get from that prompt is not.

What you want to end up applying because really, you’re looking to, like, shape it down. And like, like, we already discussed, we want to make sure that we’re moving all instances of dialed in or like, in the case of a strategy, anything that’s like, clearly, the AI has just gone in and applied because it thinks it looks smart.

But doesn’t work for an actual human.

Frederick Vallaeys: So I want to delve into, delve into? I just said it. Oh my God, I’ve turned into a GPT right now. So let’s unpack that third one a little bit more, right? So when you talk about the strategy, as an agency, you’d work on strategy for multiple clients. As an in house, you mostly work on the strategy for your own company.

How do you make that scalable? Like what’s your starting point with a GPT? Do you kind of start a new conversation every time or do you go back to an existing thread that already has the memory? How do you like to do that?

Katlyn Edwards: Usually? Well, if I’m, unfortunately I left agency before AI came out, that would have been really nice to have, especially like starting out because then I wouldn’t have to be doing all of those manual tasks.

But if I were an agency now, then I would just keep. Conversations open for each client that way because you want to be able to work from like previous memory that she That gpt has so that you can say like hey We’ve already talked about like the customer id for this specific product and then let’s just work from there Here’s like what I would like to achieve building on what we’ve already discussed.

Frederick Vallaeys: Yeah Yeah, one little problem that I ran into recently was The memory seemed to have crossed more broadly into the whole GPT as opposed to the individual conversations. So when I wanted to start a new thread and sort of like reset everything that I’ve already told it that I wanted it to lean into it wouldn’t.

It immediately came back and it said, well, I don’t think you’re going to like this answer because you’ve already told me. And that was problematic because I wanted to, you know, have a fresh set of eyes come into a problem with. For a strategy, for example and it just wouldn’t do that for me. So, so I think these are the little tricky things.

It seems like you may have found a solution to that.

Katlyn Edwards: No, I was going to say, I think that’s just because of the new update they rolled out for GPT, right? It was like the 4. 0 update and they were trying to make things, things better. Like that was supposed to be a solution, but I guess it’s just adding more bugs into the code.

Frederick Vallaeys: Right. Yeah. And I don’t know if it was a bug or by design, but I think that’s one of the reasons why I really. Strongly recommend agencies to look into building custom GPT so that you have a GPT that’s customized to each client. So you’re not building on top of the, you’re still building on top of the generic GPT framework, but you’re basically given a set of custom instructions to each of your GPT assistants, and then they can wear the hat of that client.

They can be informed about those, like you said, what’s the historical attics they’ve ran, what are their typical goals when it comes to, is it CPA based? Is it ROAS based? I kind of try to keep those things separate so that it doesn’t start taking the behaviors of one of your clients and deploying it for a different client.

Or, or I guess if it’s within an in house team, you may have different strategic initiatives and you kind of want to keep those separate at some level. I mean, you still want to cross pollinate, but you don’t want too much of it to influence one another.

Katlyn Edwards: Yeah, exactly. Like for AAP, for instance, we have a B2B and B2C side of things.

And the way that we speak to the customers will sound. Different depending on which one we address. So exactly what you’re saying.

Frederick Vallaeys: Yeah, that makes sense. So are there any ways that you’re using GPT for SEO? And I can guess that somebody answers, but share what you’re doing.

Katlyn Edwards: Yeah, it’s pretty similar to what we’re talking about so far, especially for, like, if you’re doing mass title tag and meta description updates, then GPT can be great there.

And again, it’s always more useful for, like, larger enterprise sites, because when you’re working with smaller sites, you wouldn’t want to do it manually, but it’s not as helpful and especially for, like, generating outlines. You wouldn’t want to use the copy that GPT generates necessarily, but if you want to make things a lot faster, it’s just much easier to generate that content.

And, of course, Google is cracking down, supposedly, on more AI generated stuff. And it’s not necessarily AI generated, it’s just, you know, content that isn’t helpful. But if you’re having a lot of delves into, then you might get flagged. So it’s always, you know, I’ll use it to give me a starting point and then work from there.

Frederick Vallaeys: Right. And I think There’s always that challenge with how Google perceives something, because I think the underlying thing they say it has to be helpful content. We all understand that, right? And if you think about it from a programmer’s perspective, 90 percent of the work you do as a programmer is like writing the lines of code to do the thing that you decided.

Very helpful. And so there’s all this tedious, manual, repetitive work to achieve the outcome. And I think when it comes to writing and producing SEO content, it’s the same thing. Hi. I read an article. I have a great idea about it. I can connect it back to the industry. I want to say something about it.

That’s going to be helpful to people. But then putting them to paper is the thing that takes up so much time. And then, and then you have these requirements. Oh, you know, if you write a blog post for search engine land, we want it to be 1200 words at least. I’m like, yeah, but I could have said this thing in 400.

Right. So kind of helped me run the gaps. Like maybe let’s give an explanation. of what a performance max campaign is. I don’t, I don’t have to write that for the umpteenth time. So that’s where generative I think is really helpful. So I think, and I’d love your perspective but should we be afraid about leveraging generative In that sense where it’s really helping us fill in the gaps in an article and do some of the actual writing.

Or is the, is the risk more if we just say, Hey, I need an article about carburetors and how they work and like generate it and we’ll just post it.

Katlyn Edwards: No, I mean, I’m always leery of taking directly from the AI to post onto the site because Google has been very specific for, for SEO for paid. I don’t think Google really cares.

But for SEO, then that is just seen as much lower quality content. I guess it depends on how well you’ve trained your AI model and how comfortable you feel with it actually being unique and capturing your brand voice and meeting that customer need. But typically I would think that the AI does not.

Ultimately, no better than you, like you should be the one that’s best first with what your customer is looking for, how you’re trying to answer this query, what competitors are in the case of organic ranking well, and how are they accomplishing that? Why is Google deemed them the ones that are best suited to answer this question?

Like those things, it can be a lot harder for AI to keep in mind, depending on how well you’ve trained it.

Frederick Vallaeys: Yeah, here’s a big question. So the purpose of SEO was to get ranked on Google. But now a lot of consumers are starting to just have more of the generative experience. Search GPT obviously is in the works, but even on Google, you can have the search generative experience where you don’t actually look at the blue links on the page anymore.

You just have a conversation with Gemini to get to where you need it to be. The ads, I think, are still tremendously useful because a lot of these Gen AI assistants, they don’t link you directly to the place where you can do the transaction. And that’s going to come soon enough, right? So, but for now, it’s not doing that.

But so, in that world where SEO maybe is not the best and easiest way for the consumer to get the information they need, what do you see as the future of SEO? What, like, what are you focusing on? Are you focusing more on producing something that generative can understand and incorporate?

Katlyn Edwards: Yeah, I mean, that’s a fantastic question, right?

Like, ever since, like, back when it was called AI overviews was introduced, it was always the question of, like, is this going to replace SEO? Like, how is SEO going to work with this going live? And I think that’s been in the back of the mind of every single SEO since it was first launched. And there’s not really a great question yet because people are still trying to figure out how.

Search generative experience works and like what’s fueling it. And that’s kind of a never ending question right now because Google is obviously continually updating it because when it was first when S. G. was first rolled out to all users earlier this year, I’m sure you saw it was really big. Really bad and like a lot of us years.

I’m sure breathe the sigh of relief because they’re like, okay I’m definitely not gonna get replaced by something that tells people to glue cheese onto their pizza And I know that some of those SGE results were like faked just for you know, fun social media stuff but like It, it was telling people to do like medically dangerous things.

And it was answering like your money, your life queries or health related queries that like AI really has no business answering right now, which is a long way of saying that, like, I think that SGE is the future because what Google wants to do in my opinion is serve that like almost no click experience where it knows everything about you.

It knows, like. Where you’re driving on Google Maps, it knows what you’re looking up on YouTube. So when you look up something about like, what should I get my dad for father’s day, then it already knows you and it knows your dad and it knows what your dad would like. And then it serves you that SGE experience where it says like, Hey, here’s, here’s what your dad should get.

And in that future. I’m not quite sure how SEO would work right now. The way people are dealing with it is just by, I’ve heard working with schema. So adding more schema to the page to make it easier for Google to process that information. Obviously, if you look at like the people also ask boxes Google has said that it uses a similar AI that pulls information into this.

People also ask boxes as it does for SGE. So seeing like what ranks there, am I already ranking there? If I’m ranking there, do I also rank in SGE? But with the way that Google keeps moving things around and changing the metrics with SGE, it’s really is hard to say like how exactly SEO can continue to thrive.

But with that said, I do think that most SEOs are persistent enough to be able to figure out kind of what’s going on, but it is difficult to tell since Google isn’t always the most straightforward.

Frederick Vallaeys: Yeah. Well, I think Even if Google wanted to be the most straightforward, the reality of machine learning and AI is that it’s a bit of a black box and it’s like looking into human brain, right?

We know that a human is going to be able to identify a cat, but identifying the exact process which it goes through to do that that’s a mystery to by and large still. So, and it’s the same with these generative AI where it’s the weights and the biases of the system that eventually come to some conclusion, some answer but we don’t know how.

And that’s tricky for anyone working in this space. And so you, you alluded to some. examples where the AI was dangerous. And I think that’s still a human in the loop sort of problem, right? I mean, we can all by and large look at that thing about putting glue on the pizza and probably deciding as a human.

Like, I don’t think that’s a good idea. That’s gonna hurt me, right? But then Google did make the shift and they started, It seems like incorporating more of the top ranked organic results and summarizing that through generative. And that’s super helpful, right? Because I don’t want to open up 10 different links to to get the information that I needed.

If the information that’s already been deemed most relevant and helpful by Google is summarized, that’s great to me. Did you see a shift when Google transitioned from kind of like doing a more generic summarization in SGE to actually leveraging. Their database and their index of the highest ranked SEO articles.

Katlyn Edwards: You mean in terms of like what was showing up in SGE?

Frederick Vallaeys: Yeah, I mean yes,

Katlyn Edwards: yes, yes and no. SGE, it’s still pretty hard to track like how things are ranking there, especially when back when Google wasn’t including any links to the things that they were referencing, which definitely had some people up in arms about, but as it is now, I’ve seen a lot of SGE instances where they’re actually just pulling from the featured snippet that’s already at the top of the page where you would have like the answer for SGE and then right below that is the exact same thing.

Same thing in the featured snippet. So at that point, then there would be no change, obviously, because you’re just featured a little bit higher up on the page and duplicated. But other than that, I have not seen that many changes myself.

Frederick Vallaeys: Yeah. Here’s the thought for me, and I’d love to hear your perspective on it.

So the purpose of SEO is basically to get these blue links to show up on the page. That got me thinking about books. So I’ve written a number of books. I’ve talked to some other authors. And we all sort of complained that to put something on a bookshelf, they want it to be at least, say, like 150 pages, if not 200, 250 pages.

But the key points of any book is usually about three key points, right? The big, the big things that you need to take away. So, as authors, you write 240 pages of books. On top of the 10 pages, I could have said what you needed to say to make the publisher happy. Then the person picks up the book and they’re like, Ah, now I need to read these 250 pages to get to three points.

So what should you logically do? Well, you should go to generative AI, get it summarized. You get the information out. So it’s like all of this information that’s been put in the middle. That’s kind of useless. But then I started thinking about it from the perspective of, well, it’s actually not useless because now I can go to a Gen AI assistant and I can have a conversation about the book.

And so it can tell me the three key points. But because there were 200 additional pages with examples, with case studies, with anecdotes, I can say, hey, can you make that more relevant to me? I’m I’m male. I’m in my 40s. I live in California. Like bringing examples that would resonate with me. And so it’s able to pick through that book and have a more useful generative experience.

And so what I’m thinking is, You know, when the consumer just wants to get the transaction done, SEO is just sort of been this thing that sits in the middle that at some level has been useless and it’s kind of making it slower to get to where you want to be. But then at the same time, the landing pages have great information that could help me figure out like which one of these 10 solutions is the best one for my situation.

And so is that sort of the rethinking of SEO where we’re just like overloading the system with even more information than in the past, because it is better Distilling it down for each individual human. So would SEO actually expand as a result of this, even though we’re going to get fewer blue links to click on?

Katlyn Edwards: I mean, I think that kind of lies more in the hands of Google to decide than an SEO necessarily. I do think that is exactly the direction that Google wants to go in. And from like a PPC side of things, you’re kind of all set because you’re just Putting up ads and assuming you’re serving a quality experience and you’re actually giving the authentic product, then Google wants to make money from that ad.

So they’re of course going to continue putting those ads out in front of you and prioritizing them from an SEO lens. Like I hate to say that I agree because I have been on the SEO side of things, but yeah, there’s definitely a lot of really bad content out there. That’s so redundant. Like if you scroll to the bottom of any, like.

Product landing page category page. You can see they have this enormous FAQ section and it’s like this is clearly just been put here so that when Google crawls it, they’re like, ah, yes, this page is selling break pads because they said break pads 72 times down here in the FAQ and that’s not useful for anyone.

And I, I think that Google is also going in that direction in terms of encouraging helpful content. And so that seeing that SEOs will then do like you’re saying, try to serve a more optimized experience toward the user. But I’m not sure how well they’ll be able to do that since it requires a lot more AI resources than most companies typically have.

And it’ll usually fall in the hands of Google to serve that just from the search experience as a whole than from a company serving that from the site experience.

Frederick Vallaeys: Yeah. So you having been an SEO and PPC. Which field would you recommend people focus on more these days?

Katlyn Edwards: Oh, that’s, that’s such a hard question.

I, I feel like If you have money, then PPC, but you can’t do anything. If you don’t have budget with PPC, obviously like SEO is free and the two ultimately should be working at tandem where SEO can cover much more top of funnel results. Then PPC can be capturing those leads that are at the bottom of the funnel.

And SEO can be helping with bottom of funnel as well. They had like some amount of budget. Then I would try to set up PPC first, but then focus the majority of my resources into establishing SEO. And then once I have like at least an SEO platform, then I would start splitting my attention between the two, but that way you can be driving leads, getting revenue from PPC first, while you’re focusing on SEO and establishing rankings.

Frederick Vallaeys: Right. But SDO at the end of the day is not free either, right? It’s, it’s still the human resources that you put in and the time that you spend on it, it’s just a little bit more disconnected. And, and BP C’s very clear. You put a dollar in, you get a click out, right? And you actual conversion rate for that click.

And hopefully you’ve done a good job with that. The other part that we haven’t really talked about, but influencer. Advertising or influencer, branding, social media, like, where does that fall in the spectrum?

Katlyn Edwards: That is outside of my realm of experience. Other than any work that I’ve done in LinkedIn, I don’t have a lot of experience putting any ads on social or doing any organic social.

Obviously, it’s incredibly important, I think, in terms of building backlinks for SEO. If you have a well established social platform, then putting up your newly published page is a great way to get eyes on it without having to have it rank first or even be indexed. But beyond that, I don’t have a lot of experience there.

Frederick Vallaeys: Now, you also pointed out some future Gen AI trends. Do you want to cover those a little bit? What are your thoughts on the future of Gen AI?

Katlyn Edwards: You mean in terms of, like, SGE, Chat GPT, just everything as a whole?

Frederick Vallaeys: Anything that you think would be most interesting?

Katlyn Edwards: Okay. So my, this is my tinfoil hat theory.

Okay. As far as chat GPT, because I find it very interesting. I really, I both like personally, I like professionally, I dislike what the CEO, Sam Altman is trying to do with chat GPT because it would ultimately eliminate our jobs as paid search workers. But with chat GPT, his ideal experience. That I understand is by offering an ad free experience that is totally customized to the user that you pay a monthly subscription for.

So, it’s basically offering what it It’s offering Google, but minus all of the ads and it’s customized to you. So that way you can go in and say, Hey, this page is here. Not because some SEO came in and shoved in a bunch of keywords or because someone paid a lot of money to put this out in front of me, but instead, because it’s the best product that suited for me, according to chat, CBT, so that I find very interesting and I’m curious to see, cause that’s offering something that Google can’t possibly do because from the beginning, Google’s whole thing is that it’s free.

So I’m definitely watching that. It’s, it’s a long road ahead for chat GPT to compete with Google because there’s such a giant, but it’s a very interesting product.

Frederick Vallaeys: Yeah. Well, my take on that, I think that’s going to fall flat on its face. I think enough, enough companies have tried to offer. A paid sort of Google search experience under the premise that you’re saying, which is it’s not influenced by commercial interests, but the average consumer does not want to pay 30 a month.

And so I think when it comes to the typical consumer who’s got a white collar job, their company probably pays the 30 GPT subscription. And that’s why we have it and that’s why we get to use it for personal stuff and actually be more productive when we’re planning a vacation or looking for a personal purchase, but anyone who doesn’t get it paid for by their job, I really highly question that they’re going to pay that subscription fee.

I mean, you even look at Netflix, like how many people, Signed up for the ad supported version, even though the defer the the non-ad version is not that expensive. And then the other point is that for as much as you can guide the consumer to like the right solution for you, for example, in travel, like would it be better for you to fly directly from A to B as opposed to have a stop?

Would it be better for you to have more leg room? And would it be better for you to be able to bring a bag? Yes, probably, we can all agree on that, but ultimately the consumer, the typical consumer will take the ticket that’s 10 cheaper, even if it’s less convenient. People tend to value money more than inconvenience or convenience.

So I think those are the reasons why I really don’t think ChatGPT Or search GPT is going to be able to be an Atlas platform.

Katlyn Edwards: Yeah, and you might be totally right. It’s just something that I have my eye on because I find it interesting. As far as other generative AIs, I also find the circle to search that Google recently released.

really interesting as well from like a PPC’s perspective, especially from a B2C perspective. And I’ll be curious to see how they’ll integrate that into or how we as people working in PPC will integrate that into our ad market. That’s something else I’ve been watching.

Frederick Vallaeys: Tell us a little bit more about what that is.

Katlyn Edwards: So it’s essentially a function where you open up your Google camera. I actually, I can’t remember which camera it is, but it’s a camera through the Google app. And you look at a product and you can just circle it. So say if I, I’m on my laptop right now and I looked at that, circled it, then Google would be able to serve me products saying, Hey, is this the product that you’re looking for?

And it just cuts down on that whole, like, user journey like that. So if you have a user that’s out in the wild and say, let’s talk about car parts since I work for a car company, that’s like, hey, what car part is this tire? And then like, bam, like that, Google can serve that ad to them.

Frederick Vallaeys: Yeah, so the Google Lens experience, which I think they now introduced it in the Chrome browser, so any tab you’re on, you can basically send that back as an image and then do the circling thing and then that generates some responses, both commercial and non commercial.

And I think that’s super interesting. I’ve talked quite a bit about this, but like the meta AI glasses with the little cameras built in, it’s super useful. I mean, I can imagine running a, I mean, hopefully not running a flat tire, but walking into my garage and it’s like, Oh, my tires are a little bit bald.

I need to get a new one. So now the process is I run to my computer, I’m like, okay, what do I need to look for? It’s like, well, you need to look for these three numbers on the, on the, on the sidewall. And then I run back to the garage and I write it down and I hopefully get that right. With the Meta AI glasses, you take a picture and you’re like Hey Meta, look at these tires and tell me what they are and where I can buy them.

Fantastic, right? But, generative assistant becomes really useful is like, I don’t really understand much about tires. So like, what is the benefit of having a all season tire versus a winter tire? And, and, and, and just the exploration process in traditional search is so cumbersome and so slow and so bogged down by SEO that you almost don’t want to do it.

But if the, if the assistant is sitting in your glasses and you’re literally having a. Two minute conversation. Oh, well, it opens up new ways for me to think about it. And maybe I’ll buy a tire that I hadn’t considered because I actually got, got to inform myself better in a fraction of the time that I would have otherwise taken.

So I think it’s super fascinating how these things are going to work. And then the other thing it serves as a memory, right? So if you have these assistants on you, like in the form of glasses and you’re in a store and you see an interesting couch and like, Hey, take a photo of that couch. You remember it.

And later on, the generative assistant, I can tell it that I want to buy that couch. And it literally knows what couch that was because we had a conversation about it. It knows where I was, what store I was in, what it looked like. And so now from a keyword perspective, like I don’t have to define the brand, color, the size.

I just have to say like, buy the couch. And it knows what that couch is. And that’s so fundamentally different. And so I’ve,

Katlyn Edwards: Oh, I was just going to say from a retargeting perspective, it’s also great, right? Because like, say you wanted, you looked at that product and then it knows and it remembers. So then you take off your glasses, you sit down in front of your computer and you’re just scrolling on the internet.

Now you’ll get hit with retargeting ads for that same couch. And it will just, it’s great just from capturing any lost leads.

Frederick Vallaeys: Exactly. The retargeting thing, how that can change from having visited a website to just you having seen a thing. Through your AI glasses. That’s pretty amazing. But that also raises an interesting question, I think about first party data.

So really, that’s third party data because it’s meta. It’s Facebook. In that case, who’s getting the data through the glasses. But I was in the store that belonged to, say, IKEA. So, and I had that interaction in the IKEA environment. So how is that not really their first party data? Which is tricky, right?

Because every company is struggling right now, or is vying to get as much first party data as possible, given the way that cookies are changing on the internet. And so. And that’s where, like, right now you go to any website that does commerce. First thing they ask is, do you want your 10 percent coupon?

Give us your email address. Okay so you can imagine there’s going to be generative experience where I’m in Ikea, I’m looking at this couch, but the first thing that comes up is like hey, by the way, Seems like you’re in Ikea. Do you want to sign up for the email list? And now it becomes like a first party interaction through the GPT action platform.

And the GPT action platform is basically it’s able to use APIs to whichever company. To do the transactions, but also could use it for first party data generation.

Katlyn Edwards: Yeah, it’s really interesting. And it’s also I feel like using the UK and Europe as a case study for that also works really well too, just because they’re so I don’t want to say selfish.

They hoard their cookies, right? Like every single site you land on. They’re like, Hey, Like they have to ask you, like, can we take your cookies right now? Like, is that okay? And just looking at them as a case study from a PPC and from an SEO lens, just in terms of gathering first party data and user data to see, like, how are they performing there?

What marketing techniques have worked well there would be useful as well.

Frederick Vallaeys: Oh, that’s interesting. So you use, you use the data from the more restrictive markets to inform your strategy for the more the markets where there’s fewer privacy regulations.

Katlyn Edwards: Right, like in the U. S. Since we’re all expecting, even though Google pushed back the end of cookies deadline, but that access eventually going to come down.

So just using like, how have marketing strategies been set up in places where data has already been so restricted and then emulating that here just so we’re prepared for when we have more restricted data.

Frederick Vallaeys: Yeah, that’s actually really good advice because I talked to a lot of people and they say, well, hey, we always do things first in the United States because whatever privacy, right?

So they can do anything they want. And then we’re going to figure out how to fit that into more restrictive model that is more complicated, right? And to tie this back to GPT a little bit, I was watching a master class and it basically said, listen, if you ask a generic query to GPT and you don’t put a lot of restrictions in place, it’s going to give you pretty generic answers that you’re probably not going to be super happy with.

But the more restrictions and limitations you put in place, you’re forcing the system to be more creative. And I think that’s exactly what you’re saying for humans. Like, don’t take the easy path out, like actually spend time to strategize. What would you do in this more complicated market to operate in?

And if you can make it work there, then you can bring that to the United States and then be ready for success. In the long term, as opposed to have a short term game.

Katlyn Edwards: Yeah, completely agreed. And it also gives you a testing ground where you already have to apply all of those regulations and a place where you can both test like in the EU and the UK, and then you can bring that back to the US where you have more open restrictions and then compare with A B testing.

Frederick Vallaeys: Very cool. Okay, so a couple more really quick questions here. But how do you stay informed about changes in the industry? What are your favorite resources?

Katlyn Edwards: Oh, I’m obsessed with search engine land and search engine journal. I know we’ve already talked about. I always get the two mixed up because the acronyms are so close, but I love their they send out daily updates over the weekdays and I read those.

Every single day. I have an hour blocked out every morning to go through and read them, especially for following with changes like the antitrust trial with Google, like those are so important to keep pace because everything’s always changing.

Frederick Vallaeys: We’ve talked about generative quite a bit, but is there like one cool example of something you’ve done maybe in your personal life, like a unique way?

Of using generative

Katlyn Edwards: not well, this isn’t this is something that I’m still exploring, but I really like the new offerings to create like generative videos and generative images, especially looking at things, not from a personal, but from like a paid perspective. It gives you a lot more reach into like P max because you’re able to actually generate those assets.

Instead of having to wait for creative,

Frederick Vallaeys: Is there one thing that you wish you would have known about marketing digital marketing before you started a career in this field?

Katlyn Edwards: I feel like starting in the agency was really good because it gives you exposure to. Such a wide range of companies, all sizes, all industries, B2B, B2C.

That plus the corporate red tape is everywhere. I’m sitting in front of a wall that’s washing me with red. So that. This is a physical representation of the corporate red tape, and you just have to

Frederick Vallaeys: be patient. Okay, Katlyn anything else you want to make sure people know or how can they get a hold of you?

Katlyn Edwards: LinkedIn is the best way to get a hold of me that, and yeah, I don’t think anything else comes to mind. I really appreciate you having me here today, Fred.

Frederick Vallaeys: That was great having you on. Thanks for sharing all your insights and thanks everyone for watching. Hope you enjoyed the episode. If you want to see the future ones we got into works, please hit the subscribe button at the bottom.

And of course, if you need a PPC management tool, check out Optmyzr. We have a two week free trial. And with that, thank you for watching. See you for the next one.

 

More Episodes