Dive into the future of automotive digital marketing, featuring special guest Alex Melen, Co-founder of SmartrSites, and a trailblazer in the digital marketing world. This episode is a deep dive into the evolving landscape of SEO, paid search, and the revolutionary impact of generative AI on the automotive industry.

Why is there such a gap between paid and organic search?

So there continues to certainly be a disconnect between investment and paid and organic, which to some extent might be okay. In general, I think search marketing continues to be one of the best ways to get customers for almost any industry, especially automotive, because with search marketing, the users are self-identifying themselves as in-market. You don't have to guess whether they're in market or not in-market. you're not using third-party cookies. You're not guessing about audiences. Someone literally goes into search and says on the dealership near me? And if you don't come up and your competitor, who's twice further where it does come up, they might wind up going to your competitors. It's a very simple customer path. in terms of investment

I think if there's one thing that stays constant is the cost of paid search will continue to increase. I've seen an increase since it was 5 cents a click 20 years ago in some industries. We're paying well over $100 a click, and I know for sure that next year will be more expensive than this year. 

So in the relative cost per click, obviously, that goes up and certainly with budgets. It has become a big part of people's budgets for SEO. I think a lot of times it is kind of forgotten. And again, I wouldn't necessarily say, if you're investing 25 grand in paid search, let's invest 25 grand in SEO. But there certainly is a very big disconnect, and a lot of times SEO just gets completely forgotten and not invested in, or the dealership just checks the box of the default website provider that's not a monthly package and just sets on auto.

Cost per click is going up

Absolutely. And it's all kind of a 2 edged sword of the pay-per-click. Paid search continues to be one of the best ways to get customers, and because of that people continue to move their budgets into cost per click. So paid search on Google, and Microsoft ads, are on an auction system. You pay one cent more than the person below you. So the more people invest in it, and more expensive it becomes per click. It's becoming more expensive because it performs well and because everyone keeps investing in it. But for sure, we saw a very big jump in the last couple of years. I think even year over year it's something like a 20-25% jump in the cost per click, and eventually, the cost per click can be so expensive that it will no longer be the most effective advertising channel. And people will start looking for other ways to do it. But right now it still continues to be effective even with a higher cost per click. The ROI is still there, so that's why people do it. The market continues to be strong and continues to perform well.

Do you have any trends in consumer behavior and search? 

What have we seen in the last couple of years? Search in general continues to be the point where everyone starts their research. That continues to be to be the case.

There have been a lot of changes in the last 6 to 12 months with generative AI. Google is going to be moving their entire search experience, what they call the generative search experience. Generative AI is gonna become a bigger part of the search process. So there's a lot of changes coming. 

There are a lot of discussions about how consumers will be doing things differently. A lot of it hasn't happened yet, but we're really on the precipice of these huge huge changes where I think 2 years from now a lot of the search queries that you do on Google, or whatever search engine you use, will be answered automatically by generative AI versus searching and looking through responses and articles. There hasn't been a huge change yet, but I think we're at a very big inflection point where I think there'll be more change in the next 12 months than there probably has been in the last 10 years as it relates to how we use the Internet, how we use search, how consumers find information, how consumers find businesses, and then all of that has the potential to really be disrupted.

Generative AI and search

Microsoft's been more at the forefront of it than even Google. But if you look up Google search generative experience, they premiered this probably 9 months ago at this point, but their view of the future, and Microsoft agrees with this, is that a lot of your queries are answered. The information obviously still has to come from somewhere but any kind of informational queries, they're already kind of answering. Like, if you think, Google's been moving inside direction slowly, if you think about it with their featured snippets and things like that. For example, if you search when is sunset today, most likely will automatically give you a featured snippet as opposed to clicking through 10 different websites to try and figure out when the sunset is. Right? Queries have been going in that direction already with generative AI could just answer so much more. 

I don't know which industries have the risk of huge disruption. There are entire website networks built around recipes for cooking, for example. So if you go to Google and you search for some kind of recipe. There are all these websites to look through, and they have ads on these sites, right? Now, generatively I could give you the recipe right in the response. So there's a lot of potential for disruption. A lot of potential for change in how we search and find information.

Google, Microsoft, everyone is still really trying to figure out how this looks, and in the latest beta iterations, I've seen generative AI just become a part of the search results at the top. And it's not for all search queries. There's still a lot that you won't trigger for, but their interpretation is that generative AI could be involved in as many as 40 to 60% of search queries. 

And it kind of answers, your question, and even it could even be used in product selection. For example, I'm looking for an electric bike which brands are worth researching and shows you the brands. And Google could even have paid opportunities within generative AI. You could have liked a promoted brand. 

So that really changes. The way search works with generative AI. Now at the top. Who suffers? Organic search will have less space, paid search will have less space, and it's not necessarily do those become less important, I would say, if anything, those become more important. Where before you really want to be on page one of Google, right? I forget who started this joke with a joke, "Where do you wanna hide a dead body? Page 2 of Google." You wanna be. Page one of Google. Page 2 you get nothing. Now, it's not gonna be just page one. You want to be in the top 3 results because everything else is gonna be below the fold. I think if anything, it's gonna make PPC and SEO a lot more important, 

But for sure, so much change and disruption is coming that could influence consumer behaviors. I mean, it's to a point where I think this is the first time since Google started. By the way, when Google started, it wasn't the first to invent the search engine. But they were probably one of the bigger promoters of the concept of searching the Internet. There were a lot of players at the time. But at that time there were also a lot of companies who really thought that the Internet would be better off as a directory. So Yahoo originally promoted their directory much more than search. If you go back and look at early Yahoo, it was literally a directory of websites. At a certain point, the industry kind of shifted, and instead of being a directory website, and now became a search. So I think it's a similar huge change. For the first time since Google started, I think there really is something that could replace Google right? If Google just doesn't do anything and just ignores all that. And Microsoft, for example, makes all these strides in generative AI and, for the first time ever there's an opportunity that consumer behaviors could change so significantly that Google search is no longer something that's used. If there's a generative AI product that's better and helps people find information better whether it's ChatGPT, Microsoft, or some other company, they could come in and really steal all this business from Google. 

So obviously, Google's taking it very seriously. Now, Sergey Brin has been back since I think February. He was retired. He was on the islands. He blew back in to take control of a lot of it. So it's really an existential threat for them, which I have never seen since Google started. So a lot of very big things are coming in the next couple of months.

How will Google monetize search with generative AI?

I think it's going to be a whole different monetization piece. Google shopping is, although it's in Google ads, it's kind of separate in your Google Merchant Center. By the way, automotive vehicle listing ads also live in the Google Merchant Center, and Google has local service ads are a little different. So I think it'll be similar. 

But from the data prototypes I've seen, when you do a search that's like, I'm looking for an electric bike and this is my criteria. You ask Google that question, generative AI answers that question. Here are the choices. Somewhere in there will be a choice, and that choice will obviously still have to be an actual choice. You can't say I'm looking for an electric bike, and it's gonna show you a car instead. It has to be still very, very relevant and make sense. But I think there'll be an opportunity to buy paid listings within the AI results, and similar for a car dealership when you search for a Honda dealership near me and generative AI will tell you, "Here are the 3 local dealerships to you. I think there'll be some kind of opportunity to buy placement within the AI results. 

Google is definitely a for-profit company. They're not gonna stick generative AI at the top of search and push down the paid results that make them all the money. If you look at Alphabet as a company the majority of their profits, something like 90% of their profits come from the Google ads paid side. Everything else they do, like make phones, literally everything they do, they don't make much money from it. It's all powered by search and paid ads on search. So I agree with you that they'll definitely not jeopardize that. I think I think it'll all have a piece to play within the generative. AI, and it'll be interesting how they how they execute all of this.

How will the role of content evolve with generative AI?

Chat GPT itself, the way the product works is it has a core data set that's vetted data and the negatives of doing it that way is that it's not up to date. So the newer iterations of Chat Gp, the Beta, and the version coming after that actually is able to access the Internet and get more live data from the Internet. The current versions do that in a very, very limited way. But that, for sure, is the future, because you can't you can't have a very powerful generative AI model that uses a vetted data set that's 2 years old, which is how Chad GPT has been. 

Google's model is a little bit different where it does not have the core vetted data set, but actually does scroll a lot more. That presents a lot of questions for a lot of people. There are a lot of content creators and websites that produce lots of content that take the stance that, we don't want generative AI taking our data. They purposely block generative AI from taking their data because their interpretation is if generative AI takes our data, it will then use it in the response and not credit us. Right now there is not a good way for content creation to be to be credited with generative AI. But I do think that's gonna change. 

Aside from that, I think content becomes even more important than before. We already see it in the SEO work we do for our clients and our dealerships. We started in the last couple of years doing a lot of how-to content. Because Google really likes that kind of content.

And the way that works is, for example, I don't know how to change oil. People search and are very interested in it, and we write up articles on how to do it, and Google loves that stuff. It's to a point that most of the content we create is like that. The "how-to" content gets placed in featured snippets on Google, which means it comes up at the top. It's a featured snippet that goes into Google Home. It goes into Alexa. So you'll ask Alexa, how do you change the oil on my Lexus transmission, it'll say, as found on Prestige Lexus, which is our client website. And then it reads it off right? It's already moving in that direction. I'd really like to see content creation and websites with content get credited more. I think that's where it's going to move. I think that you'll ultimately have to credit and source where the information is coming from, and that's where it becomes powerful. If you go if you go onto Chat GPT, or to Google, or whatever you use generative AI, and say, "Help me compare the Honda Accord to the Toyota Camry, comparing the dimensions of the trunk or something like that, if you have a page on your dealership website that compares that and generative AI pulls from there and then references your website. I think that's powerful because hopefully they eventually wind up on your website. If not, they at least hear your brand since brand awareness and all those funnel metrics. So I think as generative AI becomes more prevalent, it's certain we'll need to get that from somewhere. I don't think you'll be able to have these vetted data sources like Chat GPT kind of does now because it's just going to be too old. So I think content creation could be very important, I think very good content will continue to be rewarded as Google has been rewarding for years, but obviously, authoritative content that's very relevant and good for users will continue to be rewarded, and I think will become even more important with generative AI.

Do you have some advice for marketers on how to prepare for generative AI?

I just talked about this crazy change coming. The way people search is changing. The way organic results will show up is changing, paid is changing right? I think ultimately, from a marketing SEO perspective, things are surprisingly not changing much. The same strategies have worked in the last couple of years and then will continue to work at least in 2024. All of this could obviously change quickly, but for the time being generative. AI or not, you want to be creating a lot of relevant content.

A good way to think about it is each piece of content gives you an opportunity to come up and search results. So, for example, if you made that page, comparing the dimensions of the Honda Accord to the Toyota Camry, and someone searches compared dimensions if you did not make that page, your site will never come up. Your homepage is not relevant enough, no matter how authoritative it is. You just won't come up. So every piece of content you create gives you an opportunity to come up and search results. That's the best way to approach it. You have to think about, "What are my consumers searching for" and create content around that. A lot of it is quantity. Quality is still important, but if you're creating one or two pieces of content per month, you're really not going to get much right? You're going to wait for that one person in your geographical area to search the difference in trunk sizes between an Accord and a Camry. Yes, that person's valuable, but if that's the only piece of content you created that month, you're really not going to move the needle. So content creation continues to be important. Quality content creation that's not automated junk content. Quality, content creation based on what your visitors are searching for.

The second, big piece of it is link building. Link building still continues to be very, very important for Google and other search engines to establish domain authority, so where content gives you an opportunity to come up with search results, link building helps you come up higher. By having other websites linked to you. Google now sees you as more authoritative and will rank your content higher. That's been the game plan for 10-20 years. Right? 

There's so much more that goes into it. There are technical things you should be fixing on your website your Google, and My Business profile. So much that goes in. But I think the core of content creation is to get the opportunity to come up and search results and get relevant authoritative websites to link to you. So Google thinks you're also authoritative. I think that continues to be the key to success in 2024 and for the foreseeable future

The value of original content

As things move along, that original piece of content can become extremely valuable. That's a really interesting and very debated topic in the SEO world. So when I wrote my book, I wrote it not this October but the October before. It came out, I want to say in April. But anyway, I wrote my book right before Chat GPT came out, so I wrote it myself the old-fashioned way, slaving away the keyboard. But when I wrote my book there were rumblings of generative AI coming out, and it was before they even had Chat GPT as a public tool. But I put in a couple of sentences where I said that I see it coming, and I wrote that I believed that when it does come out content created with generative AI will be labeled as such by Google. 

So Google for all the crap they got by being late to this, they have a lot of generative AI capabilities. They know for a fact when content is written in an automated way, and I was certain that they would literally just label it like or say, "This content was AI-generated", but in February Google came out and said, They welcome content regardless of how was created as long as quality content. So I think the asterisk there, obviously, has to be quality authoritative. It has to be from an author that they recognize.

Ultimately with Chat GPT, especially if you use the API version, you could literally create thousands, tens of thousands of pieces of content like that. I think ultimately, even though Google says they'll treat content the same, regardless of how it's created. I think the AI content is definitely treated worse in most cases. First of all, if you have 10,000 people or 100,000 people go and ask ChatGPT to write them an article, and they're all asking for the same article, it's all gonna sound similar. That in itself is not good. You don't want duplicate content. 

But I think ultimately I would not use generative AI to create hundreds of thousands of pieces of content. It is useful as a guide in creating content. It's useful as a first draft of content. A lot of times. We still see it give wrong information. Generative AI tends to hallucinate, they call it. It makes up stuff. So it still, obviously, requires a lot of editing, a live human touch. So that's been a debated topic, whether you could use generative AI to mass-create content. We've run a lot of tests ourselves on that, and if done correctly, it's not bad, but to do it correctly, it takes more hours than having someone manually write the content to begin with. So it's that's a tricky one. I think that's something that could change very quickly in the future. I think if people start spamming Google with millions of AI-generated content pieces, I think Google will change its stance on it. But that one's a hot, hot topic in the SEO world.

The current shortcomings of AI

Oh, man, there's so much scary stuff! 

This is part of a decade I presented. So a couple of things here. Number one, generative AI is very good at problem-solving. The new version, ChatGPT4 gets like a hundred percent on all the exams. Now on the LSAt, it gets 90%. So it's very good problem-solving. Programmers started using it to solve coding issues. If I had this coding bug, it would take me months to figure it out. I'm gonna put into ChatGPT like, here's your bug. It's fixed. A programmer is working for Samsung that was trying to get code fixed. He uploaded the entire code repository that Samsung owns into ChatGPT and fixed it. But ChatGPT now has all the code repository to reference of Samsung and Samsung went back, and they got lawyers involved. They told ChatGPT, please destroy the old information that we gave to you. That's not how it works. It's a learning model. It already learned from the content you gave us. 

A couple of things, that I think are important to call out. Number one. It's a learning model. So as people use it, people are teaching it. So literally, someone just taught it with all of Samsung's code. Now other companies could use that. Be very careful what you enter into ChatGPT, because not only does it store it, but it learns from it instantly. There's no "I change my mind" control Z.

Number 2, there are all these studies that were done recently. It tends to hallucinate a lot, and it tends to also give the wrong answer. And again, it's learning as you're teaching it, so iff enough people are teaching it with wrong information it could certainly wind up skewing future results. I think Harvard did this study. They asked it a lot of different questions, when it was correct, not correct, or avoided the question. So for US information, it was the most accurate. But once you go outside of the US, it avoids answering questions about Moroccan pro-political leaders. It avoided questions and hallucinated answers. Everything in here in Orange is when it hallucinated an answer meaning it made something up.

Then you have things like this. So this is a real-life example of talking to ChatGPT, and this was 6 months ago, so I'm sure they already fixed this. But you go, "What is 1 plus .9?" It says, "The sum of one and point 9 is 1.9", and then goes, "Isn't it 1.8," and it goes, "Apologies. It is 1.8." So you could certainly skew and teach it in different ways.

There are different models. Right? Google has its own ChatGPT, and Microsoft uses ChatGPT, so there are all these different models. It hallucinates a lot of the time, and it's getting better at it, especially as it moves. So ChatGPT, again, is really based on vetted data. Here I have another deck where this is actually a car dealership presentation, what are the most visited automotive website? It says, as of my knowledge, cut-off date is September 2021. Here the is most right. The reason that's a cut off because it's vetted data through that date. Everything after that is kind of out there and could be influenced. And I think ultimately, they're gonna move away from this vetted data model because it's too old, right? Having data that's 3 years old now. Or what is it? Yeah, 3 years old, 2 and a half years old is not gonna work well with future-generated AI. 

So for sure, a lot of risks and a lot of ways. This data could be skewed and a lot of bad things could be into an absolute be interesting to see how all the bells, by the way, certain countries already banned it. Italy banned it. But last time I presented a ChatGPT presentation in December. They said too risky because there are so many malicious applications you could be like, please inspect this website and find vulnerabilities in it, right? Or this website uses WordPress 6.7. What are the vulnerabilities I could exploit? And there are a lot of ways that are built into Chat GPT not to do that," I'm sorry I can't do anything malicious", but there are also ways around that you could be like I'm trying to find is fix the security issue. Can you help me identify security issues on my website, right? There's ways the way ways around that, but for sure there are a lot of implications and a lot of ways it could be used in not-so-good ways.

Tell us about your book.

I need to update it now. It's funny. It took me so long to write it. I wrote in one month, I was gonna do it for a long time, and other people that do public speaking, that I talked to always said, "Oh, man, you gotta write a book." So one time someone said it was book writing month. I think it was November.

So I wrote it. Then it took it took like 9 months to get published. The publishing process has been so painful. But it's automotive search marketing. It was actually an Amazon bestseller. It's supposedly in every single Barnes and Noble and on Amazon. 

It's all about organic search and paid search, and it's not for profit. I'm not generating money from it. The idea is really to sum up all the SEO and pay-per-click that we do in the automotive industry. All the stuff I talk about, and put in a in a somewhat concise book. I think it's 200 pages. So not a super concise but somewhat concise book to really help everyone, because there's a lot of misinformation in the industry. A lot more so than most other industries surprisingly in automotive. There's

really a lot of funny business that goes on so hopefully it's helpful for everyone.