Episodes / AI in the Business / No. 666
SEO vs AEO: What Every Business Needs to Know
We explore the evolving landscape of search engine optimization (SEO) versus answer engine optimization (AEO), and how businesses can adapt their strategies to stay ahead in AI driven…
Listen instead and the audio keeps playing while you browse the site.
In this episode
The write up
Ryan Deiss came into this conversation with a question that he had from a client call. He said that an SEO agency had pitched one of his clients, and the pitch was that if you optimize for search engines, you're also optimizing for the AI tools. Ryan said he had to plead ignorance on whether that was right or not. Then I got on and we started talking about it. My position is that it's not the same thing. Most of the show is me trying to explain why it's not the same thing, while Ryan tries to make it practical.
Where we landed is that the game has moved from rankings to recommendations, and nobody really has it fully figured out yet, including us. The order that you're going to do this in is going to depend on where your customers are coming from today. Ryan's advice to his client didn't change at all as we went through this, so I think the reasoning behind the order is what's important for a business owner to think about.
Rankings were the old game, recommendations are the new one
We're going from an old game where the game was position. You wanted your site to come up first, the competitor's site to come up second and the other competitor to come up third. That's really changed with ChatGPT, Claude and Gemini. What I argued was that we are optimizing for recommendations, not rankings, and that we're optimizing for selection, not position. One answer is going to get selected as the best answer. The second answer is going to be considered and the third answer might get mentioned.
Search rewarded relevance. My opinion is that recommendation rewards evidence, meaning who can prove the most that they are the best answer. It's all of this data, results, testimonials, third party recommendations, original proprietary content, and being cited and referenced by other people who have credibility. Reviews I would put in the third tier, which is why I don't agree with a lot of the SEO advice that I hear, which is just get more reviews.
Ryan asked, when's the last time that you did a normal Google search? I said I still use it, but not for research, and not for lists of options. I use it after the AI has made its recommendation to me, to find the site of the direct source. Ryan tied that into the levels of awareness. He was saying that search engines are going more towards not being top of funnel discovery, but being how do I find something that I already know I want? I said it's like direct type in traffic, except you don't know what to type in.
The paid search play that falls out of that
Ryan's rule is that if you're on a fixed marketing budget, then the first bit you spend should be branded search, bidding on your own name. Ryan said it is frustrating to pay for your own name in Google, but if you don't, a competitor probably is. And it's not a lot of money to do that.
I took it a step further. If a competitor is getting the AI recommendation, then people are going to go to Google and type in something to find the site of that competitor. What I would do is bid on the terms that they're likely to type in to find that company. My example was a made up firm, Ryan Deiss Financial Advisory. If I can pick up a click on financial advisor for $70, but I can pick up a click on Ryan Deiss Financial Advisory for three dollars or fifty cents, then I'm just picking up that traffic that the AI created at seventy, eighty, ninety percent off. I think that's a fun strategy.
Ryan said you can bid on competitors' terms all day long. There are sometimes limits on putting trademarked names in the ads themselves, but he said it's worth a test.
Start with 30 questions, then build the evidence
Ryan asked, where does the business start? I said to pick 30 questions that you want to be the obvious answer to and reverse engineer from there. Ryan stopped me there, because to him it's such a big shift. You're not thinking about keywords anymore. You're thinking about the whole sentence question that the person is asking the AI.
Then I go through evidence, authority, assets, distribution, freshness, uniqueness and trust. Is there third party data that says you're the best at this thing? Are your reviews visible and pointed to? Have you built tools or proprietary content? Is it recent? Is it a unique point of view or is it just the same answer everybody gives?
Underneath that is what we're calling the authority graph. When I started out in SEO, I built a private blog network, and I had servers on multiple networks linking back to my site. The modern day version of that can live on your own site. You've got nodes that are your assets and signals, the connections between the nodes are how each one relates to the others, and the strength is the authority and trust. Schema helps with this, but I don't think it's the whole story, because the AI reads much the way we read.
Ryan brought up the AI slop thing. He said people are mass producing this AI content hoping that AI finds it. I said it's kind of like the skyscraper strategy from SEO. Let's find out who is winning, work out why they're winning, and then beat them in as many categories as possible. He called it reverse engineering the ultimate key. His version was basically to play customer: go to the tools and ask your 30 questions, see who gets recommended, ask for the sources, and work out what they have that you don't. My hypothesis addition to that was to look at the sources that the AI is citing and work to become the number one answer on each of those.
Where Ryan and I differed on priority
If your business is falling apart, and you can't fulfill the customers that you've got, then you've got more important things to do. But I think that everyone I know that was dependent on SEO is at half or worse in traffic, and I think it's only going to get worse. So I would move my SEO budget over. If I had no budget for either, and I was starting from scratch, I would start here, because it's easier to build assets that I know are going to help.
Ryan was more cautious. He scores things on impact, confidence and ease. His view was that the impact could be massive, but his confidence is low and the ease is low because the playbooks don't exist yet. He agreed completely with me that if SEO or local search or reviews are your primary driver, you have no choice. For everybody else, Ryan said it was branded paid search first, then the obvious paid channels, then organic social with an embedded influencer where the founder is the spokesperson, and only then this.
His example was a client named Gary, who runs an environmental construction business that puts large fuel storage tanks in the ground. Ryan said Gary's number one source of leads is LinkedIn videos of him walking job sites, with posts that get around six comments. Ryan would not tell Gary to stop that. As Ryan put it, "people like to follow people, not brands."
My point back was that Gary is already building an asset that helps here, because that content creates authority and trust. The next step is to connect it all in one place so the AI doesn't have to piece together a website, an article and a social account on its own. Ryan agreed and summed it up as no wasted effort, just a question of what you prioritize first. He would move this ahead of organic social only for a business that is search driven.
Run your own tests, because ours are anecdotal
I said on the show that this is an anecdotal observation from us as to what we think is going on, and hopefully it's somewhat intelligent guesses. I haven't seen any of the AI companies say exactly what they do. Ryan said that neither of us sell those services, and that nobody has it figured out. He said he was doing SEO in 1999 and never had that figured out either.
Here is our own benchmark. We picked our 30 terms and then we asked, how often are we showing up? We found that Scalable showed up in two of the 30, honorable mention. But we're going to do a test now of eight things at one time in a two week test, because I would rather know something in the group works than have to spend 16 weeks testing one thing at a time. Ryan added that once you know that's working, you want to pull away pieces one at a time and see which one it is.
Ryan gave us some relief there. He said he was not first to Facebook ads, and that there's something to be said for being second and letting other people make the mistakes. But I also see it as a gold rush going on, and the first people in are likely to become the incumbents that are harder to dislodge, just like what happened with SEO. Both can be true. So pick your 30 questions. Look at who's getting recommended. Run your own tests. And if we got something wrong, tell us.
Written from the recording of this episode of Business Lunch.
Show notes
In This Episode of Business Lunch: We explore the evolving landscape of search engine optimization (SEO) versus answer engine optimization (AEO), and how businesses can adapt their strategies to stay ahead in AI driven search environments. They discuss practical tactics, the importance of evidence and authority, and how to leverage paid and organic channels effectively.
Connect with me on social:
- TikTok: Check out my TikTok Here
- Instagram: Check out my Instagram Here
- Facebook: Check out my Facebook Here
- LinkedIn: Check out my LinkedIn Here
- Subscribe to my YouTube 👉 Here
Resources:



