Ep 227: Prompt Engineering For AI Content Creation with SEO Steve Wiideman

Jaryd Krause:

SEO Steve, you're back. Thanks for coming back on.

Steve Wiideman:

What's up my friend? Good to see you again.

Jaryd Krause:

Good to see you too. You're on podcast episode 136. This is around 230, this podcast episode. So it's been almost two years because we released one.

Steve Wiideman:

Where do we find the time? That’s amazing.

Jaryd Krause:

It's crazy. It's crazy.

Steve Wiideman:

So much fun. Thanks for having me on, Jaryd. I appreciate it.

Jaryd Krause:

Yeah, thanks for coming on. There's so much to dig into today. And yeah, typically, our podcast is around the half an hour range, and we've booked out, we've got an hour. And I did just get some feedback from some people saying that they get to the end of the podcast episode and they're like, “Damn, let's get a bit more.” There's more that we could have covered. So let's try and cover it all today.

Steve Wiideman:

Let’s do it.

Jaryd Krause:

First and foremost, this fear that has set in across the industry around AI, I think it's going to be good for us to just bring it up and talk about why we shouldn't be frantic around these AI changes and SEO. But where are you at? Let's just start this at a complete open, clean slate. When people talk to you about AI, what's your feelings and what are you normally speaking to them about AI?

Steve Wiideman:

Oh, man, I'm so excited. I’m kind of a nerd, right? I love any kind of change that transcends an industry, and I know there's a lot we'll talk about with concerns and what we worry about. But as somebody who’s basically built my career and my life off of helping businesses appear as often as possible for the right search terms to attract their target audiences. So with everything happening in the AI world and being able to automate a lot of things that used to take human manpower, critical thinking skills, and research skills, being able to create prompts to automate a lot of that is really exciting stuff.

You have two perspectives you can have, right? So you can have this fear mentality of “Oh my God, the sky is falling” and panic and find a career that's less stressful. Or you can embrace it, fall in love with it, master it, and really, really kick the tires and enjoy the process. And even then, at the end of the process, if you still don't find yourself on the other side and doing well, that's when you sit down and start to decide, "Hey, maybe it's time to look at something else. But you don't do it out of the gate. You do it after you've embraced the change and done what you can to adapt to that change, before you start hitting the panic button.

And in many ways, it's a sign of character, right? What kind of character do you have? Do you have a fear mentality or an opportunity mentality? And right now, my perspective is that this is all really new and exciting, but that doesn't change my client's need to show up in search results. It doesn't change the restaurant chain's need to scale to appear when their users or customers are searching for a burger restaurant near them. That's not going to change. What's going to change is our process and how we adapt to the way that people are going to search now. Thanks to AI and the 200 or so developers that are working on Google's new search engine.

So I'm excited. I don't think I'm nervous, but I think that might happen at some point when the new search engine launches, but you're going to find me red-eyed after a few days of really kicking the tires to see what shows up and trying to understand why what's showing up is showing up. But my pulse is high, and I'm excited. I think it's going to be a lot of really interesting things that are going to happen.

Jaryd Krause:

I love that you said that. The only thing that's constant is change, right? And if you think about your career in the SEO industry, it's always changing and evolving. And if you're going to hold on to things like, “Oh, it's changed. What am I going to do now?” and the fear around it, you're just going to have a really tough time having a full career in SEO or in an online business in general. Because the only thing that's constant is change. And I think where people get really worried is that something comes out and it gets floated, and then they're like, “Oh no, we thought we had a set and forget approach.” When reality is, it's always constant change.

Steve Wiideman:

Not this industry, yeah, for sure. SEOs who are listening to this, if you're an SEO practitioner and you have this kind of fear about, “Oh my God, what am I going to do?” Now is the time to make sure your certifications in ads are up to date. Because the only way you're still going to be able to see success as you're learning the new search ecosystem is through paid ads. So you might have to take on some paid search clients for a little while until you've really cracked the code on how to make sure that your clients are doing what they can in organic search. So that would be something that I would do to give yourself a little peace of mind that you've got some financial stability as search evolves more than it has since the first search engine displayed blue texts or blue links and black texts.

So that's something that you could do if you're in that kind of mode of, “Oh my God, what am I going to do?” Just make sure you're up to date on all the new cool things that you can do in paid search. It's a completely different Google Ads than what it was when it was Google AdWords. It's completely different from 20 years ago, when there were Overture and Yahoo Ads.

Learn it with a clean slate and an open mind, as if this were a brand new way to advertise. Don't think with the old mentality of keywords. Go into it like a beginner, and I think you'll enjoy all the cool things that you can do now in Performance Max and the neat little ways that you can use audiences.

So I would make sure that on the paid side of search, you're up to speed. Start your own campaign for your own brand so that you get acclimated to all the different ways that you can create campaigns. And you'll have a way to sort of survive through the change until you've mastered being an SEO in the AI SEO world.

Jaryd Krause:

Yeah. And especially for non-SEOs that are listening to this who own their own online businesses and get the majority of their traffic from search. The reason that they would be worried is because they don't know how to operate in the new landscape that is coming. But the reality is that if you're an adaptable business owner, you're going to change anyway because you have to adapt to it, and you are literally in the same seat as everybody else who needs to learn as well.

So there's nobody who really has more advantage than you anyway because you're hearing about it at the same time and you get to use it at the same time. If anything, it's a bit more exciting that you could be on the cutting edge than people who are less adaptable, right?

Steve Wiideman:

You know the first thing I asked Bard was when I got access to it? I said, “Bard, tell me as a client how I would rank my website in your search results.” Are links still important to show up in Bard? I started basically asking algorithm questions, and Bard was happy to share answers. And that would be the place I would start right now. Go to Bard and start asking those questions about the algorithm, how important links are, and what are some ways I would show up?

Ask specific prompts because the search engine engineers haven't gone into Bard to exclude answers to questions about how its algorithm works. So have some fun, ask questions, and Bard will tell you everything it knows. It's really not that complicated. And you're going to find that the change really isn't that significant. The principles are still the same.

Have helpful, relevant, user-friendly content that's original. Don't just be unoriginal by taking AI and just dropping it on your page. Come up with helpful examples, videos, pictures, and other things. Make sure that you're still earning links from other websites, because those are still going to be important. There's still going to be a Google bot crawling the web to look for new content. And make sure that when people do find your results in search, your results are the most helpful, the most compelling, and maybe the most entertaining in some cases, so that you're always the one that gets selected. And when the question comes up, “Was this a helpful result?” users will always say, “Yes, this was a very helpful result.”

So pay attention to search experience, pay attention to off-page SEO signals, and continue to make sure that your pages are performing better and better every month, not just by bounce rates, but by conversion rates and by the things that you pay attention to and how users interact with your content. As long as you're doing those three things and getting better at them every month, whether it's Bard, Bing Chat, or whatever happens to be, you're still going to be there. You're still going to be prominent.

They're still going to see you as helpful. Then you won't have to worry about these changes. It's when you're complacent, rest on your laurels, and have that kind of fear mentality that things start to fall apart.

Jaryd Krause:

It's the same right now as it was five years ago and ten years ago. For a business or a blog that creates content, you just need to continue to create helpful content. That hasn't really changed.

Steve Wiideman:

Not at all. And I feel like it's a cliche, and it's exhausting for a lot of SEOs. It's like I can't just put content up and it'll suddenly rank. You're right, it won't. You have to make sure you're using internal links to show the search engines that those pages are important. You have to make sure your brand is mentioned in conjunction with those keywords on other websites so that the co-occurrence of your name and the brand is prominent. You have to do offline advertising so that people are searching for you by name, sometimes including the keywords that you want to be semantic for.

There are a lot of things that search engines are going to pay attention to. And as long as you're addressing all of those different areas, even just a little bit every month, you're going to be okay. You're going to get through. And in a year from now, you're going to be like, “I'm glad we kept paying attention to those things every month while our competitors decided to try some shady things to game the new search results.” We did what Jim Rohn said. We do ordinary things extraordinarily well, and that's all it really takes.

Jaryd Krause:

That's all it really takes. Learning how, and everybody listening should know how, to be a better business partner with Google or the search engine to give the search engine what it wants. And if you give them what they want, you're going to give the users what they want, and everybody's going to win rather than playing a game against your business partner. Imagine how long your relationship with your business partner would last. Just as long as the next Google update if you're playing games.

Steve Wiideman:

17 years ago, I was challenged with the task of ranking for the keyword "seo expert" by my boss. And it was to show him that I understood search well enough that I could reasonably give enough confidence to the folks at Disney Parks & Resorts that my SEO skills were going to increase non-branded search results. And the funny thing was, at the time, no one was really searching for an SEO expert.

It wasn't that challenging of a keyword to try to rank for. And I held it for some 12 years before my peers said, “Hey, the reason you're not getting invited to events, they think you're bragging about being an SEO expert.” And I'm like, “Well, I kind of have to get people to search for me and use that phrase for me to be able to hold that position.”

And so eventually I said, “You know what, I need to be a better part of my community.” And I just killed the page altogether. And it was so worth it. Because I have so many great relationships now, the stigma of “he thinks he's number one for SEO experts because he ranks number one for SEO experts” is gone.

What I would want to rank for now if I were pivoting into the new world of chat, AI, and large language models is to try to rank for the keyword prompt expert, chat prompt expert. You want to be the number one prompt expert in today's world of search. That's what I would try to rank for. And then I would practice what I preach, become a prompt expert, study it, learn it, and look at all the different examples people are using. Create Chrome extensions with some of the prompts that you want to test with and templates that you want to use. And just master it.

Get some coffee, stay up till one in the morning for two or three weeks until you're just exhausted to death, and then give yourself a few days break and take a vacation. But become a prompt master, and you'll have confidence that you'll be able to leverage the right prompts to create the right contents and SEO and other roadmaps to really use prompts as the catalyst to becoming better at what you do—not what's going to replace your job, but what you're going to be using to improve your marketability in the marketplace.

Jaryd Krause:

Mic drop. Well, it's not that I think, but I know that that's just so damn juicy because that is a philosophy that goes across all verticals of all life – prompting. If you think about what Tony Robbins says, he says, “The quality of your questions determines the quality of your life.”

Steve Wiideman:

I love it.

Jaryd Krause:

And I knew this when I started using Google to try and find answers. If I worked out the more specific and better the question was, I would get a better answer. And people already know that without even being business owners, right? So if you adopt that philosophy that you should have in life and apply it to using these chats and these AI softwares, you're going to get far better results in and out of your life, right?

Steve Wiideman:

Absolutely. And you can have fun with it. Remember the movie Patch Adams with Rob Williams?

Jaryd Krause:

Yes.

Steve Wiideman:

Do you know how he was able to get through college? He had to make it fun and funny. And so one of the things that I do as part of my career is do some public speaking, and some event coordinators will call me and say, “Hey, we've got a room full of private investigators that want to understand different ways that they can search Google to find information about people they're looking for or things or facts and stuff that they're looking for. And I'm like, “Well, how do I make that fun? Am I just going to do a bunch of slides with search operators on them?”

And so what I ended up doing was finding a story. I said, “Here's a story. Rowan Atkinson was in an interview, and he got offended by a question that somebody asked, and he left and he disappeared. Now, our job as a private investigator is to use Google search operators to find out, first, if the information was true that he was being accused of in the interview, and second, to find out where he is so that we can interview him ourselves and find out from the source what was going on.

And so I use those different slides as I talk through the search operators, and I went through every single search operator that you could use. And I included these funny stories and memes about Rowan Atkinson, Mr. Bean. And it was the funniest, coolest, neatest, and most laughable presentation that kept people engaged because they didn't know what Mr. Bean was going to do next. But they also learned by using some of the search operators that were introduced in those slides.

And some of them were shockers. I was able to do an URL search for .gov and then in quotes I put, “Not for public distribution”. And I showed them all sorts of nuclear content that was not supposed to be in the search results. But because the folks that were putting the information on the website didn't realize that that website was getting crawled and indexed because they didn't understand search, all they understood was that I put a piece of content on a website that was public to the world. And so the private investigators were like, “Oh my God, this is amazing.” And so sometimes it becomes something that's not just funny and amusing, but in some cases, shock and awe.

But that's my advice. Create a story. Create a fun, engaging, goofy thing that you use, and then explore every possible way that you can get to that endpoint using the different ways that we can search now.

Jaryd Krause:

Yeah. And prompt engineering is going to help you get better stories. Last week I decided, all right, and I've told this story before to a bunch of clients, that I wanted to use the ChatGPT to help me write some content, some emails. I use my email for BOB like somebody would use Instagram, because I don't really use Instagram that much. And I put videos. Well, not so much videos anymore, but images, GIFs, and stories about crazy stuff that happens in my life. Like shark attacks, crashes, all these different random things, falling down the Great Wall of China, and some wild stories that I tie back into business and philosophies.

And the way I tell the stories, it's very junky. I start the story in media res. I started like a James Bond movie, in the middle of the action. Think of it, get the people's attention, and then move into different parts and break it up.

Steve Wiideman:

And then give them the backstory.

Jaryd Krause:

Yeah, afterwards, right? So I understand copy, and I understand engagement, and all that sort of stuff. Trying to get ChatGPT to do that is very hard if you don't understand prompt engineering, and you still need to put your spin on it. These AI tools are great because I think they help prevent writer's block. That's the way I feel. It's like one of the best things. But then still taking it and putting the human element on it because the human element is not yet replicable, I believe, in a way that creates connection, awe, shock, and stuff like that, that you can do through your content.

Steve Wiideman:

An AI can't display video interviews of your customers talking about you or before and after pictures of what you did for a client or a customer. It can't really enhance your branding visually or what the experience is when they visit your page, versus just giving you a lot of text and maybe some ideas or video scripts to use to help build that. And you can actually say, “Read this article and then write a page with the same voice.” I know you can actually have it create a page using your writing style. But then again, you'd want to go back through and really look at it. Is this the message I'm trying to convey to my customer?

And yeah, you can go back in and say, “Yeah, you didn't really hit it on the nail. Make it more compelling. Be more empathetic.” And you can guide it to the place that you want it to get to. And then you're still going to read it and feel achy and be like, I don't know if this is really authentic. Let me draw some inspiration from what it's saying and kind of put it in my own words. Read it maybe two or three times, then close it, and then write your own with your own voice.

And then go back in and say, “Edit this and remove any typos, brevity, or anything that feels misleading or confusing.” And then you'll see the output of that and go, "Okay, they just took my page and made it so much better. And I took their ideas and incorporated them into what became my ultimate paper.

So there's ways that machine works with human who works with machine who eventually have human write the final page. So I think there's a lot of that collaboration that can happen between us and these large language models to create amazing stuff. We don't have to feel like if I use it, I'm cheating. No. Use it. I use it. I wrote, I don't know, some 30 meta descriptions for a healthcare client over the weekend because we had a team member out. And I'm like, “I got us.”

I'm always covering for our team members, and I used ChatGPT to help me write some titles and meta descriptions. And I did end up touching up almost all of them, but I wrote the right prompts with “Write a convincing 160-character meta description for this client who offers this service,” and I put “using empathy and making sure that we've addressed the intent of the target audience more so than the competition.”

And it took a few rounds. And do it again. Give me another example. Give me another example. Give me another example. Until I finally got it. And I'm like, "This is great. And I let it know, “This is great. This is what I like. Now do it again for this keyword. Now do it again for this keyword.” And within a few minutes, I'd literally written some 30 rows of titles and meta descriptions. And again, I did a little bit of my own editing and put my own spin on it, making sure the right keyword is used and removing the brand from titles that waste real estate. But I got what I wanted out of it and it looked fantastic.

And then we present to the client, and we say, “Here are some ideas for titles and meta descriptions that you could use. We encourage you to put your own voice on these, but this will give you some inspiration to start with. Or if you just want to start with them outright, please at least read them to make sure that they're making the right statements and claims and so forth before you just deploy them. But this could be a good starting point for you.”

And if they come back later and say, “Yeah, I'm not really getting the clickthrough rates I want.” great, let's run some tests and do a few more. Give me five meta descriptions with different call-to-actions and different value propositions to see what we can do to get our clickthrough rates up. And I'm pretty sure Google's not going to penalize you for doing that, especially if you're getting higher clickthrough rates and users are getting the content they were looking for.

Jaryd Krause:

Yeah. By the way, those meta descriptions have been touched so many times by a human, not just written, but also consciously thought about, like, “Oh, maybe we could change it this little way?” I like what you mentioned around getting some good ideas, getting it written, and getting your really good prompt engineering in place. There's one really cool prompt that you mentioned, which is “make this better than, more conveying, or more valuable than my competitors.” Really, really smart little prompts that I heard when you mentioned that. And then take it away, shut it down, and rewrite it in your own words.

I remember one of my mentors saying, “Write something one day, come back the next day and edit it with fresh eyes.” And then you can edit it. And then like you said, put it back in. And you're right. It's an amazing tool. It's an amazing tool that we should use. And if everybody's using it, why would you not use it? You're not cheating; you're just getting better and better at your craft. It's like changing from a hammer with a really small head to a hammer with a massive head. You're going to hit it more times than not.

Steve Wiideman:

Yep. And again, don't rely on it as an exclusive resource. We had a client that did that with a few other category pages. They just literally went out to ChatGPT, and they said, “Rewrite this category to be more helpful” and whatever. And they just dropped that content verbatim. And they saw like a 20% drop in performance for that page, and they had to revert after a couple of weeks.

Yeah, don't do that. This isn't a replacement for what we're doing. This is an enhancement to our process to speed things along. And you could easily do some keyword research using Google Keyword Planner, free tools, or paid tools like Semrush or Conductor, and then jump back into Bard and ChatGPT and say, “Help me build a silo of supportive content underneath a taxonomy.”

I did this for a lawyer who works on truck accidents the other day. And I said, “Build me a list of supportive pages underneath my truck accident lawyer page to provide more helpful information about what to do after a car accident.” And it gave me 24 really good topics that were great, and they incorporated the keywords that I found from the Keyword Planner, which took five minutes to pull. And now I've got so much more supportive content underneath that truck accident lawyer silo that there's no way they're not going to look at that page as being more helpful or more authoritative based on just having that supportive content underneath it. It was super easy to do and took just a couple of minutes.

Jaryd Krause:

I want to bring up the detection of AI content with Google, Bard, and all that sort of stuff. I know that when people are looking to buy businesses, because that's what we do, we help people buy businesses—where we're detecting, we're looking for AI content. Most of it is written by an AI generator, whereas some of it has been touched up and worked on by humans. And then there's also the question of how accurate those tools are in detecting them, and there's also a question around Google.

Steve Wiideman:

False positives.

Jaryd Krause:

Yeah, exactly. There's that. And then how will Google respond to that type of content that is clearly just written by AI? Like you said, the person who created those category pages lost a lot of traffic because of that.

Steve Wiideman:

Just two posts. I’m sorry. Go ahead.

Jaryd Krause:

So what are your thoughts on the detection? I mean, this might be speculation.

Steve Wiideman:

I'd like to start with the guidelines. They put out two posts. The first one was specifically around the use of AI at scale. They said, "Please don't just use artificial intelligence to deploy thousands of pages of content. They made that blatantly clear; they want you to do that. And then a couple of weeks later, they did a post that said we love AI, we encourage the use of it, and you could absolutely write content and do things with it. We're not going to penalize you for that.

But they did make it clear in the first statement that if you try to do it at scale as your primary means of trying to game search rankings, like, “Hey, I can just create a bunch of keyword rich pages and game the search results,” that stopped working in 2011 with the Panda update, right? We were just buying content through content sellers. And some of them were extorting people for their content at one point, too, if you remember that. And they got sued for it. But I agree. I think that's where we should start.

And then the second point is looking at the detection piece. And what was funny to me was, I think, in February, when the whole sort of ChatGPT just blew up and hundreds of sites like Jasper’s and similar type products that you can create content now through automation, just hundreds of them, just all over the place. And now, here in May and at the end of April, hundreds of AI detection software programs are coming up. It's so hilarious to watch how people try to monetize and capitalize on the changes and trends that are happening and search more than ever because of AI.

We're kind of loyal to them right now, even though there are a lot of false positive issues, but we've been using Originality.AI. And what is the other one that some folks are using? There's a lot of these different tools. There's one that I've been in talks with the folks that have Originality.AI because I teach, and I'd say about 15% of my students now are submitting assignments that are essay driven. And 15% of them are using AI to do their essays, and I've got to kick them back. And they can't just expel every student who does it. So they've got to figure out a way to manage it right now from the education side.

So I reached out, and they said, “Yeah, there's a site with a partner we're working with called passed.io or something like that. And they're supposed to be helping teachers see if their students are using AI on their assignments instead of writing them for themselves.

So my challenge with Originality.AI was that, yeah, I would get a lot of these weird false positives. Instead of getting red and green, I'm getting weird yellow, even though it's saying 50% AI and 50% original content. And so what I'm asking students to do, and I hate that I have to do this, but I have to ask them, please make sure that you've run your content through an AI originality tool or an AI detection tool because I'm using them. And if I see anything that looks suspicious, I'm going to ask you to redo it.

And I know in some cases you've written it from scratch, and you feel like this is unfair and I'm sorry, but this is the changing world of what we're having to work with. And a lot of what I'm looking for is original content. So maybe you did hand write it, but you didn't put enough effort in creativity into it that these detection tools are finding it to be original content. So really get creative, put your voice on it, get weird, get silly, get funny, get original, get controversial. That's the way that you're going to pass these originality tests, these AI testing tools, is by just being real.

But it sucks. You're right. It's not fair to students. It's not fair to content creators. It's not fair to the folks that are getting banned from Reddit for doing original designs in Photoshop that are getting flagged as mind journey stuff, right? So it's an unfair world for a lot of creators. But you've got to adapt. You've got to run your stuff through those tools now to make sure that the world knows that you created it. And that might require you to think a little bit outside the box and change your original idea of what you wanted to create. But that's the way the world is right now. It might change later, but if you want to adapt, that's what you have to do.

Jaryd Krause:

Well, I guess that's the one good thing about AI is it's forcing us as humans to create better content, right?

Steve Wiideman:

Isn't that amazing? That's what I'm telling you. It's an exciting time in our world, in our industry. It's really going to challenge us as individuals to be individuals in a world that's becoming standardized and unoriginal.

Jaryd Krause:

Yeah, exactly. It's so cool. It's so cool. I find it hard to get my emails detected as AI content. Because that's right. You've just got to make it. You've got to make your content so damn engaging, so good, and so out of the box. It's really out of the box, isn't it? Because AI is very linear in how it would write a story. When you ask it to write a story, it says once upon a time most of the time. If you don't have any good prompting at the start, it's going to start the story once upon a time.

Steve Wiideman:

It's fun, though. We did one where we said Siri and Alexa were Romeo and Juliet. Rewrite the Romeo-Juliet story so that it's Siri and Alexa. And it's fun the way that you can play with the technology to create really neat scenarios and then have some fun with them. So I think there's a lot that we can do that is helpful, useful, usable, and won't affect organic traffic or have any impact on it, such as using it to help write a resume or for content editing.

I'm writing my next book right now. I wrote a textbook called SEO: Strategy and Skills, and I didn't have any sort of AI to help. I had to hand write that stuff and get yelled at by an editor because I'm not that great of a writer. I'm really good at what I do, but writing hasn't really been a thing that I'm a master in. But now, thanks to AI, I can send it through AI tools and say, “Please give suggestions, feedback, and editing on this content.”

Sort of like the Hemingway app and Grammarly would do, but also in a way that can help me look a little bit more original and make sure that I'm explaining things that make sense. And AI does a great job of that, and it's still my content. They're just giving me editorial feedback without me having to pay for it.

Jaryd Krause:

Yeah, I love it. I love it. So how Google reacts to all of this is going to be fascinating. And I think what they're doing or what they will do is that they're going to do what's best for humans because humans are the ones that are using that platform, and that's where they make their money from us and our eyeballs looking at ads and whatnot.

So what's interesting is—and I feel that when you said using AI to create content at scale, Google will detect, “All right. You've just posted 2,000 articles this week. That's not natural.” This is an assumption by me that they will penalize a site like that. I would like to highlight how else you would believe that Google could penalize people or hurt sites for using AI if they're trying to play a game against their Google partner or their partner being Google.

Steve Wiideman:

I think we're already seeing that. I think with the latest review update—I don't know what you want to call it—they're really trying to give sites that have real user feedback and social proof a little bit more credibility. We see that with some of our restaurant chains, where we've got location pages that are appearing pretty prominently. And then, after this big update, we're seeing that the delivery service providers and review websites are starting to flank us on many queries simply because they have ratings and reviews and unique content that's not static the way that some of our location pages are.

So I kind of feel like they're going to detect it. Well, we have to go back to sorting out the three basics of SEO, right? Is our content the most unique, original, and helpful? And how do we determine whether it is or not? It's not going to be words and H2 tags and keywords tied into paragraphs, entities, and topics that we research. Those are going to help us get indexed and shown for those keywords to see if we're going to be a good search result. But that's not ultimately what's going to keep us on top of the search results.

But that's paramount, right? We're going to do our keyword research. We're going to do our entity analysis. We're going to think about the ways that we can create the most helpful, unique page. Second, we're going to look at ways that we can get other websites to want to share and talk about us and maybe even link to us, hopefully organically, because we've got so much great supportive content underneath that silo below that lower final page and upper final content that everybody finds to be helpful, linkable, shareable, and mentionable.

And third, we're going to be paying attention to our search behavior signals and how people interact with our listings. At the moment, while it's still blue links and black text and some image and video carousels, the way that we do that is through using structured markup – to markup our images so they become image thumbnails; to markup our FAQ so that we get two questions under our results, to markup our ratings and reviews that we're going to be adding to all of our product pages now, especially after this last update, so that we have stars in the search results. It's going to be making sure that our listing is the most compelling, stands out in the search results, and gets clicked on.

So the way that I think AI plays into that is as we look at kind of the big picture of things is AI is going to give you a base of text that if you try to scale it across millions of pages, Google is going to see, “Hey, you've got artificially intelligent driven language, large language model driven content.” But I don't see anything original in that content as it relates to a customer testimony, a customer rating, an expert review of the page, an image, or a video that was created by your company and uploaded that's not already in the binary of my database somewhere. It's not a stock image; it's real, helpful, unique information.

And I don't see a pattern of over time more and more websites linking to, sharing, curating, and syndicating some of the content that they find or curating that content on their own pages. I feel like you've just dumped some content on there. I've watched it appear in my search results. I've seen people click it, not be very interested in it, and come back and choose a final URL that wasn't you. And I'm going to infer that that content wasn't very helpful.

This isn't anything new. If you had a directory ten years ago, 15 years ago, and you had empty categories where people got to a page because you hadn't fully populated your database yet and 900 of the 1,000 pages you deployed were empty, people would click back and choose a different result. And over time, Google says, “Maybe this website isn't as helpful as I first thought it was because it had all these great URLs with great optimized titles, descriptions, and H1 tags.” But eventually, it figured out through watching how people interact with that listing and the search results, that that website wasn't very helpful. And just like you're seeing in all the posts that people are sharing now, there is this massive decline.

That decline isn't because they're calling your page and finding the content, they warned you about in that initial warning. It's because they're watching the user behavior of the user going back and saying, “Yeah, this obviously wasn't as unique and original and helpful as some of the other pages that have more rich media and more compelling expert-reviewed type content. So I think that's the mindset I'd love to see digital marketers adopt as they think about scaling with AI is that mindset of creating a directory with thousands of pages in it. And if every page isn't the most helpful page and the most original, unique, and media rich page with all the great markup to help define the meaning of words like schema.org/thing where we talk about this page being about this or menti, we refer to Wikipedia and what that is. We're not considering all those things, and we're just deploying at scale.

They're going to look at their clickthrough rates and the stay rates of what happens to users, and they're going to go, “Yeah, overall, this site isn’t very helpful. So maybe I'll stop showing it as often in my search results.” So us as website builders, we stop deploying content at scale and we work on one page at a time to make sure every single page is amazing, optimized, helpful, expert-reviewed and as markup and schema rich as it possibly can be before we move to the next page. And that's the way we've always done things as white hat SEOs. And that hasn't changed since the directory days.

Jaryd Krause:

I love it. Absolutely love it. It just comes back to quality over quantity. And when I look at businesses for sale, people submit their businesses to us to look at their due diligence, and we review it. And I was on a coaching call with somebody just today, and he had a couple of businesses. We went and looked at the listings and stuff like that. And there was one business that had 800 posts and had a domain rating and a domain authority score. And it was around eight.

Steve Wiideman:

I never look at any of that stuff. Do you really look at all those metrics?

Jaryd Krause:

Yeah, I look at the metrics, and then I tell people to go look at the quality of the content, because typically the metrics are going to tell me what the quality of the content is like anyway.

Steve Wiideman:

I don't know. I feel like those metrics can be manipulated quite a bit by SEO. So I use my intuition when I look at a page, and then I ask, is this a page that I feel is going to attract links of its own? Is this a page that I feel is really going to attract quality referral traffic? Is this a page that's likely to be seen by tens of thousands of people because it's on a newsfeed or a blog feed on a really trustworthy domain? Or is it just some little community that submitted a subdomain somewhere that someone's selling a link for $600, right?

So for me, I'm a little bit more critical, and I guess I'm spoiled that I work with larger brands and get to be a little bit more critical, whereas smaller brands are more shrapnel and you just get whatever you can and sort them out later.

Jaryd Krause:

That's where it is. I'm in a fortunate position as well, where I can afford different size businesses than people starting out. But for people starting out, it's definitely a thing. And you've got something that's got a very small level of authority and a lot of content. And it's just like, "How is this at all valuable? I mean, you've got 800 pieces, and you can barely get them in traffic. So the content stinks. It just tells you that the whole site stinks. Then you look at something else that's got 100 pages in its entirety, and it's got way more traffic compared to this other one. And yeah, it comes back to quality, right?

Steve Wiideman:

A really easy thing you could do too is use a service like Amazon Mechanical Turk, take one of your AI created pages, and ask maybe 200 users which of these pages is the most helpful. Don't tell them which one is yours. But give them the URL that ranks number one in the search results and number two in the search results, and then throw yours into the mix.

And then ask them to look at all three pages and ask which page they would buy from, which page they would want to interact with, which page was the most helpful, and why. And then, what didn't they like about each page? And then go through and theme those 200 answers and run a pivot table to focus your time on the attributes, the aggregate of your feedback culminated in.

Jaryd Krause:

I love that process for content creation. So cool, yeah.

Steve Wiideman:

So nerdy, but yeah.

Jaryd Krause:

It's nerdy, but that's what people are doing. That's what we need to do to create awesome value for the internet. These AI tools and the direction Google and Bard are heading, are making the internet better. And it's an awesome thing, really.

Steve Wiideman:

I remember the first time we created about a $5,000 page. We were working with an e-commerce brand that specializes in Rolex watches. And it was a consignment site, and they have to have a balance of buy and sell, right? If people aren't selling their watches to them, they have nothing to sell back to them. And it was fun because, when we started, the site was just nowhere to be found in a search. It was brand new. And it was like, “What if one day we could rank on the first page for Rolex watches? Wouldn't that be amazing?”

And I remember looking at the inbound marketing page at HubSpot as my sort of bar. This is the bar in terms of what great content looks like. It's got a lot of white space. It's simple to use. It's got all the different processes, and it kind of has the puzzle pieces that match everything up. And you can go back to archive.org to follow the history of that page in HubSpot. Just search for inbound marketing and look at it.

Because that inspired the Rolex watches page that we created on this website. We put a very expensive, high-definition image as a hero. We introduced video into different tiles on the page and images and tables of data that had pricing information and lots of text in a way that flowed well and wasn't just there for SEO. It actually enhanced the page, the experience, and the helpfulness of it.

And when we got to the first page, I think it took almost three and a half, four years to get there. And again, it was like $5,000. That page generates millions of dollars of revenue. And I'm like, you look at the ROI on spending $5,000 for a page that generates millions of dollars of revenue to be—I think at the time it was like number four, number five for Rolex watches. That's amazing.

So I don't know. I think you're right. I think if you really want to make the most out of SEO, spend as much as you can afford to make every page absolutely original, helpful, and unique. Look at the search results, study the pages that already rank, look at the attributes and focal points, and look at the type of media that's used. Use a GPT for Sheets plugin in your Google Workspace and pull in some entities from the search results that you want to incorporate into your outline of things that you're going to write about on that page. I think there's a lot that you can do to invest in.

One of our clients is a prominent name in housewares, and she was working on a particular product line that revolved around a type of winter jacket. And we have like eight different documents in the page folder. And they're like, “Look, we just wanted an optimized page.” What are all these documents?

Well, one of them is the competitive analysis, showing every single page that shows up in the search results and what their snippet looks like on mobile. Since, by the way, 80% of your traffic's mobile. Here's what their snippets look like and how we want them to look. And here's a mockup of what we want our snippet to look like if we're going to beat these other guys. Here's the HTML title description. Here's the schema they used.

We took all that insight and started to create an outline of the content that we wanted to have on the page based on the top performing pages. And then we did an off-page plan for this page. Then we did a technical plan for the markup that's going to be used and the image dimensions and formats like WebP, the alt attributes, the file names of images, all of those different technical attributes.

Wow, this is a lot of work. This took you over two weeks to put together, and it was super expensive. And I’m like, “Yeah. But now you have the absolute ultimate product category for this particular fashion product. And it can't be beaten.” And if they take that and act on it and create that page, no one’s ever going to unseed them or unroot them from that top listing. Because they really took the time to invest in the research that went into the planning and in the page itself and the media that the competitors just won't put the time into in terms of photos, videos, and expert-reviewed content.

Jaryd Krause:

It comes back to if you give me six hours to chop a tree down, spend five hours sharpening the ax, and actually pour money into the resources for sharpening that ax, right?

Steve Wiideman: I love it. Absolutely.

Jaryd Krause:

Well, because in my space here where we talk about buying, most people are starting out, they're buying blogs, and they're learning from so many people out there from YouTube videos and other podcasts to go at velocity and just create as much content as possible.

Steve Wiideman:

Your main services pitch, your main category.

Jaryd Krause:

Yeah, exactly. Exactly. Start at the top and build those out to be amazing. And you get to the point where a lot of bloggers are like, “I just want to create a lot of blogs and get a lot of blogs on the site.” Why not instead of creating 100 blogs, put the same amount of money it would cost you to create 100 blogs into ten blogs, pour most of the money into competitive research, and have less content that actually gets you more traffic and makes you more money. It's easier to manage on your own damn site, right?

Steve Wiideman:

And I remember I was mentioning the SEO expert example, that site that I ranked for SEO expert had five pages. It was top10seotips.com. If you go to archive.org and you want to look at those pages, it has five pages. And I ranked number one for SEO experts with it for 12 years up until 2015, 2016, or something like that.

But yeah, it's about the quality, not the quantity, when it comes to content. And yeah, there's a lot that you can create as supportive content underneath the silo. Like I mentioned earlier, don't just put everything on the blog; isolate it in its own site structure and its own taxonomy. Nest that evergreen content underneath that lower final page that you want to rank so that that page has enough supportive content to show Google that you're not only providing more helpful content, but you're also solving more problems that your customers might have and answering questions that they might have.

Whether you've sourced it from the search engine tool you're using or an enterprise tool like Semrush, Question Filter, or AnswerThePublic, which Neil Patel recently purchased, or even his old tool, Ubersuggest. Use those tools to get ideas for questions that people have about your product and nest those underneath the category, underneath the main topic, underneath the services page, and underneath the practice area page.

Don't just throw it up as a blog post. By all means, if you want to create a blog about a great but canonicalized version of it. So the evergreen version is what shows up in searches, what gets found, and what gets linked to because blogs eventually kind of fade away and get buried in the RSS feed. So use the blog for company news, seasonal things, and things that are happening right now in your industry. Don't use the blog for evergreen content. By all means, curate it. You can have evergreen content on your blog if you want to. But make sure that that really good, solid, evergreen stuff stays on a page that supports your lower funnel content. That's the way we build silos and taxonomies.

Jaryd Krause:

I love it. I love it. I had a friend who sold his business, and then the company that bought it sort of did nothing really good with it for a couple of years. And then he had the opportunity to buy it back from them, but said, “Look, I'll buy it back from you in equity, and you pay me a wage.” And it worked out to be this really awesome deal. And they just spent a bunch of money on getting the whole site redesigned. And he showed me the front page, and it was just like the most recent. On the front page, it was really, really well laid out, but just had the most recent posts.

I said to him, “Look, dude, people don't care when you come to your web, your homepage, they want to go to the most recent posts. They come to your page to find the best help. So give them the best help with your best pages on the homepage.” He's like, “Oh, that makes so much sense.”

So many designers and stuff like that don't actually understand that, right? And then, because you've got so much authority going to that main domain, that page authority linking from that homepage to that other page of yours has far more authority now because it's linked from the homepage.

Steve Wiideman:

Yeah. And you can see that right in Google Search Console. You can go straight to your links section within Google Search Console and see your most linked pages. And if your most linked pages aren't the super competitive, keyword targeted pages that are on your website, then you might want to shift your internal linking strategy to focus on those. I'm with you 100%.

Now, we do still leverage the homepage as a tool when it comes to expediting the ranking of content we want to appear. With an auto repair franchise, for example, they said, “How do we get our new locations to rank faster? When they do sign up with us and become a franchisee, how do we get them to start showing up quickly?”

We say, “Well, it's linked from the homepage to their location page so that they can pass that page rank as quickly as possible. And then once they're up there and they're in the top search results and the clickthrough rate is there, then we can rotate in another location.” And that worked really well. And it works as a really good process.

It’s funny. I have a client who every time they call to troubleshoot a page, the first thing he says is, “And yes, before you asked, I did link to the homepage.” I always ask, “Did you link it to the homepage?” So it's a kind of neat little internal linking trip because most websites that link to you tend to link to your homepage. So it carries the highest page rank.

Jaryd Krause:

Correct.

Steve Wiideman:

It's kind of a neat little trick of the trade. But well, we really broke out of AI.

Jaryd Krause:

We did. People are going to kill me if I don't stick to the AI route. So, yeah, tell us how you see AI playing out in the next couple of years with Bard and Bing, and how you think it will evolve for us? And I know that some of this might be speculation. You might actually know things that are actually going to happen, and it's not speculation. You can see a linear path to it. But I think people are hungry to know.

Steve Wiideman:

I think movies have really been a prediction tool for where the world’s going. They started with Star Trek and phones, and there are movies like Her talking to their operating system via their Bluetooth, conversational search, and voice search. We've used that. In fact, three years ago, when we were really trying to push this thing, we called it VRMMM (Voice Readiness Mobile Markup and Measurement) as our focus for where we should put our time and energy.

And then the pandemic hit, and the focus was to stay alive—keep the kitchen lights on, you know. But the voice isn't going anywhere. And there are still voice devices that are being purchased every day. People are getting Nest devices for their homes. They're using Siri. Now they're starting to use some of these ChatGPT plugins to be able to use it like a voice assistant, ask questions, and get answers.

So I think the way AI is going to evolve how we search is now finally going to happen, where we're doing more voice queries than ever. I hope it happens that way, because that's a much easier way to search and get the answers you want. And if you didn't get the answer you're looking for, you're going to say, “Give me more results.” So I don't think it's something to worry about when it comes to “Oh my God, our rankings are number nine right now. And now with the way that search has changed, we don't show up at all.”

Well, you still weren't showing up because no one was clicking on number nine anyway. They're clicking on the top results. And if they wanted more, they're going to ask for voice search in the same way that they would've used scroll to give them more results. So I don't feel like there's going to be this “Oh my God, I disappeared because now we're in a voice search world.” So I don't think that's something to worry about with AI.

I also see that there are likely going to be a lot of websites that are going to help create ways to leverage AI in your business. I'm hoping to see this, and I'm hoping it's going to be affordable so that plumbers, florists, and doctors can go to these websites and look at ways that they can leverage AI technology to improve, streamline, and automate their businesses.

And I hope to God that they're not going to be extorting businesses by forcing them into their system of doing it. But instead of giving them the tools they need that they can own, manage, and run themselves without having to pay a recurring fee to have access to, those businesses can grow and improve.

So I hope that communities evolve to help their industries use the different AI tools to improve business and streamline things that a lot of us are spending redundant, exhausting, and time consuming hours doing. So that we can focus our energy on what we do best in our business and worry less about repetitive tasks like writing 800 meta descriptions for all the different practice areas.

So I think that's my hope. That's where we're going with AI. Standards, processes, and things are built so that whatever industry you're in, you know where to go to get access to the techniques and the prompts to help improve your business. But unfortunately, just like we found with all of the different automate-your-content-now tools and then later verify-your-contents-not-AI-written tools, I hope that it doesn't become this big cesspool of pay-for-use tools for AI. I hope instead that we can be transparent and open up these tools so that we can just get better as a world by making our businesses more streamlined, more effective, and more efficient.

Jaryd Krause: Making our products and services more valuable to others.



Steve Wiideman:

Right. Absolutely.

Jaryd Krause:

I love it. I love it. Steve, thanks so much for coming on. Guys, if you haven't checked out the first podcast that we did at 136, it is about how to double your traffic with SEO. There were so many good things that we talked about. This is a very different episode. So I'm glad that you came on.

Steve Wiideman:

Of course.

Jaryd Krause:

Where can people find out more about you? Go to wiideman.com.

Steve Wiideman:

Sure. So, yeah, well, God, we need to practice what we preach and get our new website launched. But no, I think the best way to find me is on any social network; I'm SEO Steve. Our team here, if you want us to take a look at a ranking issue you're having, would love to help the industry in general. As a teacher, I love to help people. So our tag is just Wiideman. We're everywhere.

And I'm pretty comfortable if you want to email me. I've got some folks that help me work through my emails. My email is just [email protected]. Send your question, your idea, or your concept. Send it over to me. I love to talk to people. I love to help. And wherever we can add some value, collaborate, and support each other, I'm all about it. Let me know how I can support you.

Jaryd Krause:

Love it. Thanks so much. Wiideman. It's Wiideman, I apologize.

Steve Wiideman:

Wiideman, yeah, right.

Jaryd Krause:

It's a good way to remember it. I apologize for incorrectly pronouncing it.

Steve Wiideman:

No, you're fine.

Jaryd Krause:

Cool. Guys, there'll be links to all of that in the show notes. I'm sure we're going to get you back on again, Steve. This is such an awesome chat.

Steve Wiideman:

Anytime.

Jaryd Krause:

Cheers. Thanks.

Steve Wiideman:

Thanks for having me.