We all know we’re supposed to skate where the puck is going (thank you, Wayne Gretzky) – unfortunately, many people in the digital marketing industry move in packs toward the next big thing, often without understanding why.
My guest today, Andy Crestodina, is one guy who always remains ahead of the pack, which is why I’m excited to welcome him back, for the second time (the first time was episode #238). Andy is not just the brains behind Orbit Media; he’s a guru in making content work hard and smart. In our chat, we dive into how AI is shaking up the way we find and consume content online. Andy breaks it down: what’s happening now and what might come next. It’s a sneak peek into a future where Google just might take a back seat to AI apps for quick answers.
But it’s not all about AI. Andy gets real about the content game—how to stand out, why being genuine–and having opinions–beats any AI-generated fluff, and the secret sauce to thought leadership that sticks.
If you’re into making your brand shine in the digital space or just curious about where content and SEO are headed, this one’s for you. Andy’s got the playbook, and he’s laying it all out. So without any further ado, on with the show!
In This Episode
- [02:35] – Andy Crestodina talks about the potential shift in user behavior towards AI apps for simple information queries.
- [10:54] – Andy shares insights on the impact of 10x content, such as the annual blogger survey, and its ability to produce original statistics and differentiate content programs.
- [16:16] – Andy discusses his full embrace of Google Analytics 4 despite its limitations and the need for technical skills. He emphasizes analyzing data for search traffic, engagement, and conversion rates to drive bottom-line impact.
- [20:59] – Andy explains the simplicity of the subscriber mechanism of LinkedIn newsletters and the potential reach for B2B audiences. He shares strategies for promoting content through the newsletter and engaging contributors for added value.
- [26:58] – Andy elaborates on the importance of repurposing content and the value of choosing the opposite angle.
- [30:16] – Andy challenges the perception of declining webinar effectiveness, urging a shift in focus from live attendees to the total number of views and the repurposing potential of webinar content.
- [35:59] – Andy reassures marketers that changes in user behavior, such as AI-driven referrals, should not be feared and stresses the importance of maintaining a strong online presence through helpful and useful content.
- [42:16] – Andy emphasizes the importance of creating a synthetic version of the audience using AI to personalize content and enumerates his preferred AI tools.
- [45:05] – Stephan and Andy discuss where to find Andy’s valuable content, including his articles on orbitmedia.com, his active presence on LinkedIn, and his informative videos on the Orbit Media YouTube channel.
Andy, it’s great to have you back on the show.
Stephan, I love the invite. I would never have missed the chance. Thanks so much for having me.
Yeah, a lot is happening. The timing is divine; it’s all perfect. We were just talking before the recording started about topics, and AI, of course, is always a hot topic, and how that’s going to shake things up in the SEO industry. Maybe we should start there. You’re mentioning there’s a Gartner report that has some predictions, so let’s start with that.
Sure. This is pure speculation. But if you’re looking for a simple answer to a question, you take out your phone, fire up a browser and a search engine, scan through search results, and then click on a webpage, you probably downloaded more than a gig of data and dozens of JavaScript files and lots of tracking code. You’re dismissing pop-up windows; you’re accepting or declining cookies, blocking notifications, and scanning past a bunch of noise, ads, and videos just to get to your short answer, which is not good. These are dark days for visitor experience, especially on mobile.
AI does not have strong opinions, but people do.
Open an AI app and do the same thing, and you mostly just get the exact thing you wanted. It’s easy to imagine that in the future, for simple information intent queries, user behavior will shift toward AI apps instead of browsers and search engines. That’s why Gartner predicted a 25% drop, which might happen or might not. We really don’t know; it’s all TBD, so stuff in the future is definitely unclear.
Tough time for content marketers. Let’s expound on that a bit. What’s in store for content marketers? What do they need to do or shift in their approach?
I would predict a decline in website traffic to articles, especially if the articles are just giving people basic information. Very long-form, very detailed articles that are basically long-form guides probably will still be successful for those readers who need lots of step-by-step instruction, for thought leaders who stick their neck out and take a stand, for content programs that have a strong email component, or for those of us that have moved to publishing directly in the platform, like anybody who’s got a LinkedIn content program or a YouTube content program. That obviously is living where your audience is, and you circumvent the Google algorithm.
Then there’s the YouTube algorithm. What do you think about how AI is going to impact all the cruft that is out there and massively increase the number of faceless YouTube videos that are AI-generated, and there’s not even a human speaking in the narration? It’s just like ElevenLabs or something like that. Where do you see all that heading?
If tools like Fiverr create a flood of low-quality content, AI will make that even worse. There’ll be tons and tons of super low-quality content. But the algorithms, in a way, are on our side here as viewers that will protect us from that; if the total number of videos on YouTube doubles tomorrow, do you think that you and I would even notice? Or if the total number of articles on the internet doubled tomorrow, I don’t know that we would notice.
AI can help you find triggering topics, but don't trust it - trust yourself. Share on XPeople’s ability to subscribe, click the bell on YouTube, subscribe to email, or search results are really still just showing you at the top, the tiny percentage of things that are on page one or page two. I think that it is tough for the platforms and a problem for big tech more than it is for consumers, readers, and viewers.
Yeah, but the thing is, this is a constantly evolving space. If the low-quality component that we’re used to seeing with AI actually now is a year into the future, two years in the future, and markedly better than most podcasters and YouTubers out there doing it manually, what do you think is going to be happening then?
If you’re competing with AI-generated content, you’re probably not going deep enough into quality. I’ll give you lots of ways to do things. Jay Baer said this: The problem is not that AI can do what marketers do; the problem is that marketers keep doing what an AI can do.
A quick example is point of view. AI does not have strong opinions, but people do. I made a video on YouTube recently that said there are 13 things you should remove from your website. You should definitely not have social media icons in your header. Those are exit signs. You should not put dates on your blogs because that just makes your content look older and faster. This has 500 comments, and YouTube decided to show it to 170,000 people; by far, it was my greatest success on YouTube, but it’s because it’s a strong opinion piece.
YouTube sees the engagement and triggers the algorithm to show it to more people. AI could never have written that piece. AI can’t make that video that says you should do this, but don’t do this. That’s a bad idea, and I strongly believe in this. It doesn’t have beliefs. It should differentiate thought leadership content even more.
Yeah, that’s great. How many views do you have on that video so far?
It’s a thousand a day right now, but it’s maybe 170,000. My average video gets maybe 800. I’m used to 800, maybe a thousand, a few thousand, maybe a good one. The reason is clear. Take a stand, draw a line in the sand, and come out for and against something.
By the way, Stephan, this is interesting; AI can help you find triggering topics. This is creepy and weird. But if you write a prompt that says what counter-narrative opinions are least likely to be covered on big blogs in my industry, it’ll tell you. What things in my industry do people think to be true but are really false, or vice versa? What very mundane things do people get very triggered by in a certain industry?
It will give you ideas, and then you can choose for yourself, as we all should. Don’t trust AI; trust yourself to decide which of those topics to explore. I think it’s a great time for people who are a bit courageous in their content.
Take a stand, draw a line in the sand, and come out for and against something.
What is your process for ideating a month’s or a week’s worth of content for your blog, YouTube, and various other platforms?
I only write one article every two weeks, and that’s been my frequency for 15 years. I don’t push myself to hit the calendar that hard. That’s enough for me. I’m always going to prioritize the questions that are relevant to my prospects, so writing for the bottom of the funnel whenever possible. I’m going to do things that require really detailed answers, lots of screenshots, and step-by-step stuff.
My most recent articles are about How to use GA4 for SEO, How to Research Keywords, How to Write Truly Great Headlines, and lots of analytics topics. All of my best articles came together into a book. I’m talking to one of the great authors here, so it’s funny to point it out. How to launch a LinkedIn newsletter? I believe that, even if there’s no search volume for it, people should do this.
The best practices for website navigation, website footers, website about pages, homepages, the structure of high-performing service B2B lead gen pages, and things that are almost visual. It’s a pedantic strategy in a way, but it performs extremely well because these are things where it’s not that hard to write a 10x piece because most people don’t go into that level of detail on those types of topics.
Yeah, that’s awesome. I know there are some 10x articles, research, and linkable assets that you’ve created over the years. Some of them, you’re probably still doing. Some of them, you’ve given up because they just don’t work anymore. I’m curious, for example, if the annual blogger survey is one of the things that you keep doing because it is still working as a technique, or is that no-longer-do pile?
Thank you for mentioning it. That’s another way 10x content gets 100x results. If the average article takes us 12-15 hours total, that’s pretty typical for us. Again, just one every two weeks takes a hundred hours because we have to do a ton of outreach to get a thousand bloggers to answer 20 questions.
There are tons of chances and opportunities to do something different, cool, weird, and memorable. Those who choose to publish utility content will have a better chance at that than those who don't. Share on XWe still do it—we’ve done it for ten years. It produces new, original statistics that are highly differentiated. For example, my website is the source of key stats, such as the average blog post taking four hours to write or the average blog post being 1400 words. The average blog post is 70% longer than eight years ago with stats like this.
If you’re looking at the content strategy or any domain, you could just ask, is this domain the primary source for any new information? If yes, it’s a categorically different content program. I definitely am going to keep doing it for now because even though I’m exhausted by it, quite honestly, Stephan, it’s so much work. The AI question is in there now, and I’m very interested in seeing the longitudinal data on that. Are more people using it? How are they using it?
There are lots of surveys on it now, but I thought about just giving this whole program to a friend, marketing profs, or content marketing institute because they’d be almost better suited. They’ve got giant lists and can get the results and survey responses much faster.
What stopped you from handing it over to marketing profs or somebody like that?
Your listeners know very, very well the benefit of having that data on the site, the linkable asset, as 2500 websites have linked to that one URL. If I chose a partner, even if we both published together, that number would be far lower over time. They’re going to get a lot of the link love from that, which is fine. There’s more to life than that, and I have other ways to go to the market. It’s a little fear, a little greed, it’s a tradition, it’s just a habit. If you do something that many times, it’s almost hard to stop doing it.
It consistently shows that effort correlates with results.
Yeah, that makes sense. What’s something from the blogger survey that really stands out that you want our listeners to understand about blogging, its future, the ideal length, structure, topics, or anything like that?
It consistently shows that effort correlates with results. It’s simply the things that fewer people do that get the best results. The bloggers who report the highest performance are the ones who either write very long content, publish very frequently, collaborate with influencers and subject matter experts more often, or add more visuals to their content.
The takeaway is that, at least in one or two elements, you should be doing much more than your competitors or readers expected. My average blog post has 15 images and almost 3000 words. It just shows you every single thing if you go through it. Bloggers who add videos to their content more often, bloggers who podcast. We added the audio question and got better results.
If you get to a moment in your content where you ask yourself, I don’t know if I have time for this, or I think I’ve already done enough, keep going. That’s how you win: push harder, go longer, get up earlier, write more, and make that thing. Don’t stop until you have looked at it and said, if Google was the judge and my audience was the jury, this is the best content on the topic. I’m not holding back. I got to make closing arguments in a legal case. This is the number one URL on that topic, in your professional opinion.
Are you podcasting as well? If so, how do you find time to do that because you’re busy?
I don’t accept as a co-host for Chris Carr of Farotech. Chris Carr has a fantastic show. I’ve been the co-host for about a year. It’s called Digital Marketing Masterclass. It’s very heavily focused on AI. There’s, again, a content strategy. There’s more to life than just having your own, writing for your domain, or publishing on your YouTube channel. So here I am; I’m podcasting with you right now. It’s a lot of fun for me, and I try to just say yes to every single invitation.
Wow. That’s a big commitment. Are you podcast guesting a few times a week, every day, or what?
No. On average, I’d say one podcast a week and maybe one to two webinars a week. That’s plenty. There are lots of people like me, and we don’t go look to see if the show has lots of reviews. It’s a chance to meet someone new. It’s a chance to reconnect. I’m really always thrilled when you reach out. I thought this would be great; I get to talk to Stephan; it’s been too long.
Yes, it has. Let’s talk about Google Analytics 4 because you brought that up earlier, how that ties in or can tie into your SEO strategy, and what your thoughts are about GA4 compared to universal analytics. I can’t find a single person who likes GA4 better than its predecessor, but maybe you’ll be the first. I don’t know. I’m curious to hear.
I don’t look back, and I don’t complain about things I can’t control. I have fully embraced GA4 and use it for hours daily. I can do almost everything I want to do in there. However, I am quite annoyed by the lack of certain things. Why did they get rid of annotations? There were so many useful things about them. The data retention is far too short—14 months. What? I can’t.
Now, we have to use BigQuery. It pushes us to be more technical, but no alternative. It’s just so useful. Once you’re comfortable with it, you can get whatever you want as soon as you want it. If you go back, rewind, and count the clicks, it was six clicks to get anything you need. I also don’t do much reporting with it; I’m mostly just doing analysis.
Push harder, go longer, get up earlier, write more, and make that thing – that’s how you win.
In the context of search, yeah. Just go to traffic acquisition and set a date range. You can see search traffic before or after at the top level. If you go to engagement landing pages and then create a filter to only show when the medium exactly matches organic, now you can see the URL performance in search, every URL up or down from search.
If you have that report customized to show the conversion rate, you can scroll over to the right and see the likelihood that a visitor from a search to any given landing page will turn into a lead or will subscribe to your newsletter. Those are all just extremely useful. It’s a bit absurd that SEOs think not you, but there are SEOs who really just focus on rankings. That’s totally insufficient to drive bottom-line impact. They focus on rankings, then traffic, and think that’s enough. We need to go deeper; we need conversions, and we need marketing qualified leads.
Those few reports I mentioned will help you prioritize your time and focus on SEO efforts to improve the rankings of the exact URLs that are most likely to convert visitors into newsletters, subscribers, or leads. Now you know what to do; now you can prioritize. It solves the big problem in digital: there are so many possible actions. The key to success is prioritization.
How do you prioritize? What’s your process for using GA4, Google Search Console, or a tool like SEMrush or Ahrefs? How are you finding the opportunities for the links to go after the topics to write about all that stuff?
I’ll share with you my most reliable SEO win: to take a page with some striking distance key phrases, in other words, a page that ranks high but does not yet rank very high. Going and looking at those key phrases, and then just going deeper into that topic. My edits are mostly minor rewrites, where I step back and just try to make things again, probably a year old or sometimes two, three, or four years old. I’m going to make that deeper, more comprehensive, and exhaustive by covering the topic from all possible angles.
At the very end, I use a tool like MarketMuse to make sure I’ve covered the closely related phrases, but I don’t go in and use it for content briefs, or I don’t look at the people and ask boxes and structure my content based on things I see in search results. I’m really just asking myself, as a digital marketer, what would I need from this piece? What should everything include?
The best marketers write for everybody and don't miss a chance to reach a greater audience. Share on XReporting, then finding striking distance key phrases and doing tons of deep content and semantic SEO, often including a video version of that article and embedding the video at the top. So now I’ve got two bites at the apple. The video might show up in Google search results or on YouTube. That one’s for sure.
Another one I would like to do is internal linking. I’ve got an AI prompt that does this. If you export Google Analytics data with all your title tags, you upload that to AI and tell it to infer topics from title tags; then it knows the topics of all your URLs. You can tell it to suggest internal linking opportunities, which is fun.
But those two things, I think, are semantic SEO based on striking distance phrases and topically related internal links, which give me the chance to use exact match or keyword-focused anchor text, albeit internal links. Obviously, you’ve done those things a million times.
Yeah. Do you have a video or training on how to do either of those things?
Sure. Yup. The AI approach may not be available yet, but yes. If you search for Semantic SEO, that page might even rank for that. If you search for Internal Linking Best Practices, that page might even rank for that. These are the things that I’ve done.
Okay, great. Thank you.
Of course, yeah. You mentioned earlier How to launch a LinkedIn Newsletter was one of those topics that you addressed. I know email newsletters are a thing, but a LinkedIn newsletter, is that a thing?
It’s a thing. I’ve had a newsletter on my website for 17 years or something like that. I clawed my way up to a subscriber list of maybe 16,000-18,000, which I’m proud of. It is what it is. I launched a LinkedIn newsletter and did a couple of things.
I gave it a really boring but descriptive name; it’s called Digital Marketing Tips. It turns out that if the name is taken, you can’t use that. I recommend using boring but descriptive names so the visitor who sees them decides whether that’s for me or not. The mechanism through which they subscribe is just one click because LinkedIn has their email address. It’s a brilliant strategy on LinkedIn’s part because now they’re emailing all these people who aren’t on their platform and inviting them to the platform using their members as a content factory.
This thing already has 215,000 subscribers, and it grows so much faster. I made it weekly. Every other week, I recycle something old; every other week, I make something new. The downsides are TBD: people might be linking to my LinkedIn newsletter, and it might accidentally hamper my own site’s authority.
Collaborating with someone by being a guest on their site or inviting them to contribute is zero-waste marketing.
The UX for subscribing is so simple. I highly recommend this to anybody with a B2B audience. Stephan, you would have an awesome LinkedIn newsletter. You can go daily, weekly, biweekly, or monthly now. I don’t fear duplicating content, so you can just copy and paste it basically directly from the native version on your site, which may have gone live the day before. Send it early in the morning.
You may make a post with it, and you have to make that post very engaging, of course. I have never read articles with contributor quotes, so I will mention everyone who contributed to the article. Speaking of, I’d love to get your contribution to this piece I’m working on. I’m working on one right now. That has been a super successful strategy that’s decoupled my fate from Google and made me improve our reach a lot without creating more content.
Was there a gold-standard LinkedIn newsletter that you like? I want to totally R&D this, rip it, and duplicate it. Was there somebody who was killing it with their strategy?
I was watching a few that had tens of thousands of subscribers, and that made me want to do it. I saw some that I didn’t like because I thought the names were too clever. For example, what’s in Brian’s brain? I think that so much of success comes down to the titles, headlines, subject lines, and other things.
It wasn’t that I was watching one in particular, but quite a few out there have gotten quite big. Now, I get emails from people asking me to advertise in my LinkedIn newsletter, which is strange. I have learned to advertise myself in there. If I have a webinar coming up, I will put it into a LinkedIn newsletter.
I have to remember to go back and remove it because the webinar will have passed like the blogger survey. For those six weeks, while I’m trying to scrape together a thousand responses, the very top of the LinkedIn newsletter has a, hey, if you haven’t taken it yet, we’d be super thrilled; when you feed it in mind, please take the survey. It becomes just one more platform for promoting the content, but also, in a way, if it gets big enough, you can promote other things there, too.
Yeah. Do you have an article or a video walking people through how to do that LinkedIn newsletter?
I do, yup. LinkedIn newsletter best practices.
Awesome. Where does Substack fit in do you think? Personally, I think it’s not good from an SEO standpoint. They are doing some SEO practices wrong, but some big influencers are using Substack and sharing some real brilliance there.
Do they care about search? Those might not be SEOs.
Maybe they don’t. I care about search, so I’m not that keen on using Substack personally. But I follow some interesting people on Substack and love what they’re posting. It’s worth it to me to Go out of my way to consume the content on Substack.
It’s similar to Medium, right?
Yeah.
I get it. One benefit is that it’s an absolutely zero-cost way to launch something. You and I could make a Substack or Medium newsletter by the end of the day. I mean, that’s a plus. Giving people access to these tools is a plus. The downside is it’s a missed opportunity. Whenever someone links to something you made, but it’s not on your domain, an angel sheds a tear. You know what I mean?
Yes, I do. That’s funny. Speaking of that situation, what are some of the biggest mistakes people are making from a link-building or anti-link-building standpoint that angels are shedding tears over?
Everyone thinks of repurposing as reformatting, which is smart.
The blogger survey shows it. Many bloggers never write for anyone but themselves. They act like prisoners in their own domain. Maybe they just don’t have enough bandwidth in their content strategy. Everything they produce has to go on their own site because their calendar demands it.
The best marketers are writing for everybody and don’t miss a chance to reach a greater audience. It’s really magical. As soon as you collaborate with someone, either by being a guest on their site or by inviting them to contribute to something you’re writing, I think of it as zero-waste marketing because as soon as you pitch to the editor or accept a contribution from an SME, you got value, you got a benefit. You grew your network; it created a relationship.
You can invite your prospects or your cold, gone-dark proposals, or go back to your CRM and see who in your pipeline is an expert on this topic and invite them to contribute. Guest blogging is partly networking. It made me a better writer years ago. For about two and a half years, two-thirds of my content was guest posts. That’s how I met Anne Handley, Joe Polizzi, and Mark Schaefer.
I just wrote for everybody I could: Copyblogger, Unbounce, and brands I’d admired. Working with those editors made me better, and the links from their sites helped us rank. Later, the content sometimes got repurposed again. I rewrote it years later for my own site, so I didn’t really lose anything. I think that’s a huge mistake that people fail to make.
Yeah. Are you still utilizing that strategy today, or have you moved on from it?
I do. It’s just a fraction of the level that once was. But I write, I pitch, and I offer content. More of my presence on other websites is through doing webinars and events for people, doing podcasts like we’re doing here, but I still love that outreach and networking component. There are still sites that are a career goal of mine to write for eventually. I’ll never stop doing that.
One thing we talked about years ago when you were on the show last time in 2020 was when you spoke about the evil twin as a strategy or an approach to writing a content piece pretty much just once and then revising it. Hence, it had a different title and enough spin on it that it could be submitted somewhere else. You have one version that goes on that third-party website, and that content marketing partner, and then the other version goes on your own blog. Is that something that you’re still employing? Has this evolved anyway?
Yeah. Everyone thinks of repurposing as reformatting, which is definitely smart. You made something that worked well in one format and published it in another. Repurposing by choosing the other angle is something that’s less popular. Yes, I mentioned one earlier. It’s like, these are Website Best Practices, and I’ve written that. I’ve written that many times, but these are things to remove from your website.
Instead of saying the best practice is to put your social media links in your footer, the evil twin version is don’t put your social media buttons in your header. That opposite angle is really interesting to people. I think it makes editors like it and makes for a good pitch. I still love that. Anytime you write what not to do, you’re using that strategy. Somehow, the psychology of it is really engaging to people.
Yeah, it draws you in. For example, with the title of the previous interview, episode 238 with you was Evolve your Content Marketing. We could evil twin that and say, Surefire Ways to Devolve your Content Marketing. That would draw people in like, okay, show me how to screw everything up in one fell swoop.
Yup. Just imagine six guaranteed fails or my biggest mistake, the confessions of a content marketer. As a headline, which we said is one of the key success factors, regardless of the format, it’s differentiated. It stands out. It feels like you’re going to get something different, more direct, or honest. If nothing else, it’s less common.
Somehow, the psychology of writing about what not to do engages people.
You mentioned webinars, and I’m curious to hear how they work for you these days. I hear other folks say, yeah, webinars don’t work anymore. Nobody shows up, the show-up rate is terrible, the conversion rate is not very good either, webinars have jumped the shark, etc. Is it still working? And to what degree?
Covid pushed tons of content strategies toward that format, so I think that viewers got a bit exhausted by it and that it got overdone. But if you think of it as its own thing and not as an alternate version of a live event, then you stop worrying about things like live attendees. I think that’s the wrong metric.
Here’s a headline: How do we not measure the performance of a webinar? Three mistakes because we all sign up for webinars without the intention of seeing them live. You want the video. You’re planning to watch it at one-and-a-quarter speed because you know it might be partly watered down. You want to get the recording because maybe there’s someone else on your team you want to vet first, or there’s only one part of it you want to check out.
The number of attendees live is really just one of the metrics. Over time, you should be looking at the total number of views for this thing. You should be repurposing this thing. I think that it’s misguided. The people who say it’s not effective anymore because it’s not generating leads immediately or because it’s not getting lots of live attendance, the average webinar, I think, has maybe 25%-30% live attendance, sometimes 20%. But it doesn’t bother me at all.
I think there are smart things you can do with that content afterward. Here’s another AI trick. If you take a transcript from a webinar or any video, upload that to ChatGPT and say, what is the most provocative soundbite? We should do it for this conversation. What is the most provocative soundbite? What is the most counter-narrative or unusual, unexpected insight? What is the one part of this transcript that people are most likely to be surprised by or most likely to share? That transcript stuff is great.
Also, here’s total speculation, but Stephan, I wonder what you think. The webinar has a transcript. The transcript will eventually get crawled by AI. You are then training the AI to mention your brand. You opened this by saying that Andy’s a co-founder of Orbit Media. It’s a web development agency that does web design and search optimization. That little chunk of language will eventually be one more place where the AI heard something about this brand. You need a million of those, so never miss the chance to say those words on a podcast that appears in a different place, video, or webinar.
AI is more about just consuming a huge body of work and then forming predictive correlations between the relationship between words.
I think we’ll all look back and ask why those brands always get mentioned when people ask the AI for a service provider. It’s like, well, because that Stephan guy was on one million things. He has a huge digital footprint. Webinars and podcasts are one way to do that.
Yeah, although that reminds me of that SEO joke or meme about an SEO walking into a bar, pub, or public house, the long list of synonyms for a bar. You can get a little carried away with that thing, right?
Yeah. That’s information retrieval in search and Google. It’s looking for those semantically related phrases. It has ontology, entity, and semantics. AI is more about just consuming a huge body of work and then forming predictive correlations between the relationship between words. That’s really interesting. I’m sure I’m going to end up thinking about that, though. To what extent will the use of adjacent phrases and semantic SEO-type strategies be relevant when we all have some interest in training the AI?
Yeah, but I would also hate to hear a lot of unnatural speech because essentially, the next version of keyword stuff is our speech so that we rank for whatever things or get the AI to answer with the brands that we want it to answer with.
I’m totally fascinated by this. Right now, getting mentioned and linked to from a high domain authority website, the site with many links to itself, isn’t relevant, but does AI care about that at all? Probably not. A podcast interview where you give a very detailed elevator pitch on the East Peoria News website might be worth just as much as one of the New York Times. They both get crawled once.
Is there any weighting in this? Apparently, Wikipedia is more heavily weighted, or it was when they first trained to ChatGPT 3.5, but maybe not. Maybe it doesn’t matter. Maybe it’s just simply lots of instances more of the way a local SEO works.
We shall see. You mentioned that you saw an AI show up in your referral sources for one of your websites for Orbit Media. Was it?
Yeah. Everyone thinks that if search volume drops, user behavior shifts, and people stop using browsers plus search engines to discover things and go to the app, that’s the death of the content marketer. I’m looking at my analytics and see AI sending people to my website. If you look for an answer on Perplexity, you’re in a browser, there’s a citation or a footnote or whatever attribution, and you tap that and do that on my website; this is happening right now in my analytics, I can see it.
We must show up with our best advice, take a stand, and share what we believe: that we are useful, helpful brands.
Fear not, dear marketer. Changes in user behavior are not the end of the show here. It’s still important that we show up with our best advice, are present in many places, and take a stand and share what we believe, that we are useful, helpful brands. Your only other choice is not to publish anything, be it on LinkedIn, YouTube, or your blog, whether it gets promoted through an AI, through a website, through Google, or something else; if you decide to just forego all of that, then you’re back to 1998 with a brochure website. I don’t think that’s going to help you.
Even though the outcomes and the traffic channels are all unclear, even the tactics, as we just said, are unclear. I think that the contest of generosity, which is what I call content marketing, to be the most helpful, useful brand or person on the internet will continue. There are still tons of chances and opportunities to do something different, cool, weird, and memorable. There will always be such a thing as top of mind. Those of us who choose to publish utility content will have a better chance at that than those who don’t.
Being utilitarian or helpful is really just one angle for getting links, mentions, being memorable, and so forth. You mentioned being weird as another one. What’s the weirdest thing you’ve ever done to get links, mentions, and press?
This is hardly at all in a content strategy. My blog literally says at the top, Practical Tips for Content Marketing, AI, GA4, and Web Design. That’s what I’m supposed to be doing. But one time, just for fun, this was during COVID-19; I reached out to a bunch of friends to ask them to send me pictures of their desks so we could see their cameras, microphones, and lights.
It made for this very engaging post called 27 Pictures of Marketers’ Desks. It had Drew Davis, YouTubers, podcasters, and virtual keynote speakers. It was such a fun piece. It checked a lot of boxes, even though it wasn’t really about digital marketing directly. It went straight to my network because I made new friends and reconnected with lots of old friends. It was a super visual piece.
It had some practical value because it showed how to set up the DSLR camera with a cam link and use that as your webcam. People still talk about that post, 27 pictures of marketers’ desks. I’ve never seen a post like it before since, and it was a surprising hit.
Cool. It spurs on another similar idea. Hey, podcaster, what are you wearing? Do you know how, especially during the pandemic, people would have nice, well-dressed upper bodies, and then they’d be in pajamas, fluffy slippers, or something else instead of being in the legit whole outfit? Yeah, I bet that still happens. In fact, I might be wearing pajama bottoms myself right now. Who knows?
As a universal topic, everybody’s wearing something. It answers an unanswered question if you wonder what your favorite podcaster has got on. I have a friend. I haven’t kept up with it lately, but Amy Landino used to be a YouTuber who talked about YouTube strategies. Looking at the data, she saw that her most successful content by far was about her morning routine, her lifestyle, and being positive and productive. She changed her content strategy, and it exploded on the internet. She’s a YouTube celebrity now. Follow the data, see, and there are some of those very mundane topics or just daily things. They have a much larger audience than GA4 for SEO, right?
Yup. Are any plans in the works for something even more weird?
Here are two weird ideas related to live events. One year at Content Marketing World was an orange-themed event, so I bought an orange spiral-bound notebook. On the front of it, I wrote Andy’s content marketing world yearbook. I walked up to 140 people during that event and asked them to sign my yearbook, and I met them. I met everybody.
I took their picture, and then I made it all a PDF where I put their face next to their signature and whatever they wrote. People wrote funny messages. People loved it, and it got talked about for a long time, not in a while.
The next year, I went and brought a microphone. I just used my laptop as the camera. I did this ask-it-forward thing where the first person I had asked a question, then I took it to the second person who answered that question but then asked their question, and the third person answered the second person’s question. It became this rolling pay-it-forward type thing, where everyone’s just asking and answering questions. That one, I got all the speakers, so I met everybody again.
Meet people, help people, connect with others, and spotlight them when they have something to say.
There are original, fun, creative ideas. Some of them, if you just think of networking as its own goal, are really fun. Very useful. It’s a simple way to say it in the end: it’s like search has paid an organic. Influencer marketing has also been paid organic, but organic influencer marketing is what we’re doing right now. At its highest form, it’s just really called friendship. Meet people, help people be helpful, connect with others, and spotlight them when they have something to say. It’s fun.
Have you inputted your best, craziest campaign ideas into an AI and then come up with some crazy, brilliant ideas like these?
I have done a little bit of that, and it actually is pretty effective. I’ve got to say on the creativity tests, AI does quite well. The prompts that have worked for me in that way are when you throw strange words into the prompt. It’s like coming up with some totally weird ideas that could still definitely work out. When the viewer sees them, they’ll think that’s rad. That’s the kind of prompt you write because it’s all about predictive relationships between types of words, so you want to throw a bunch of weird words in there.
Also, it’s subsequent prompts because AI is a chat tool. You’re interacting in a conversation. It’s not like Google with one query after another. AI is all contextual. If your second prompt after it gives you ideas, it is like, get even weirder, weirder, something totally off the wall. It’ll just keep coming up with different things. The things I’ve seen it’s mostly like run contests. I didn’t go far enough with it to find something unexpected yet practical enough for me to try.
Got you. What’s the most outlandish or original idea that you got from AI?
My main takeaway for AI users is to first give it a persona prompt, have it create a synthetic version of your audience, and then do everything you want to do. Either upload your ICPs or have it create a persona, and correct it to make sure it’s more accurate. I’ll miss it the first time or two.
Inside that prompt, inside that conversation, you want to start prompting it with what might this person click on. What do they hate about their job? What do they love? What messages are they likely to be attracted by? Who do they trust? How much would they pay? Who else is involved in the decision process?
You can run sales scripts against it. You can ask them what they search for, what they click on, what they subscribe to, what counterintuitive things might be successful if published for this audience, and what research they would need. That is such a difference.
I joke that AI stands for average information. If you don’t train it first on your audience, it just tastes like water. It just reads the internet, and it gives you back the internet. Really, I highly recommend beginning all your adventures into AI with a detailed persona, either created by using a prompt or by uploading it to it, and then jumping in the water and swimming around. You’re going to have a lot more fun.
What’s your favorite AI tool out of the big ones like ChatGPT, Claude, Gemini, and Llama? Which is your favorite?
Because I’m trying to write prescriptive utility content, I’m trying to use basic vanilla tools. I mostly ignore all the news about tools and do everything I’m doing in ChatGPT with a Plus account, ChatGPT Plus. It can draw charts for me; I can use DALL·E, upload files, and data from my Google Analytics.
I uploaded all my data from YouTube Studio and had to find correlations, which is really surprising. The content that attracts followers and subscribers is not the content that gets the most likes, shares, and engagement. Really interesting. I haven’t analyzed topics. I would find gaps in my content strategy. I’ve got it doing semantic distance analysis; just use ChatGPT Plus.
For personal use, I like perplexity.ai. I find it less creative but more of a hybrid between search and Google and ChatGPT. Because it’s more current, it can look things up. Perplexity is a really fun tool. It’s a bit of a glimpse, I think, into the future of internet use and content discovery.
Yeah, interesting times ahead for sure. All right, we’re at the top of the hour here. We need to wrap up. How does our listener or viewer work with you? How do they learn from you further? Is there a course or anything that you teach?
Yeah. Thanks for this opportunity. I love this. Orbitmedia.com is where I write an article, and that’s where I publish everything first. It’s once every two weeks. I’ve never used that email list for anything but publishing. I’ve never gated any content there. It’s all completely open and available. LinkedIn is my best network. Also, there’s an Orbit Media YouTube channel that I’ve invested a lot of time into. There are lots of 12-minute videos, but they are very deep into topics. You can find me there too.
Awesome. You’re an inspiration; you’re a light in the world. You give so much value, and you do it all for free. Thank you for all you’re doing for the industry and for the world.
I love this, Stephan. We touched on so many things that maybe no one’s talked about yet. We should do this again in another couple of years and see what happens.
Yeah, it’ll be a very different world then. Thank you. Thank you, Andy. Thank you, listener. Get out there, make it a great week, and reveal some light in the world. We’ll catch you on the next episode. I’m your host, Stephan Spencer, signing off.
Important Links
Connect with Andy Crestodina
Apps and Tools
Articles and Newsletters
Book
Businesses/Organizations
People
Previous Marketing Speak Episode
YouTube Videos
Your Checklist of Actions to Take
Focus on creating thought-leadership content with a strong stance or opinion rather than just informational content.
Use AI tools like ChatGPT, Claude AI, etc., to find provocative, counterintuitive, and surprising topics to which my audience will be drawn.
Optimize my content for semantic search and internal linking to boost my rankings for “striking distance” keywords.
Repurpose my content by flipping the narrative and writing an “evil twin” version with the opposite angle. Leverage LinkedIn newsletters as an easy way to grow an audience and promote my content.
Use webinars strategically and focus on total views over just live attendance. Remember, I can repurpose the content in various ways.
Collaborate with industry experts and influencers to grow my network and create linkable assets.
Analyze my Google Analytics and YouTube data to find content that drives followers and subscribers versus just engagement.
Experiment with fun, creative ideas like “yearbooks” and “pay it forward” Q&A at events to stand out with the audience.
Train AI models on my specific audience personas to get more relevant and actionable ideas for content and campaigns.
Connect with Andy Crestodina on LinkedIn to stay up-to-date on his latest content and insights. Visit orbitmedia.com to learn more about content strategy, SEO, analytics, and more.
About Andy Crestodina
Andy Crestodina is a co-founder and CMO of Orbit Media, an award-winning 50-person digital agency in Chicago.
Over the past 20 years, Andy has guided 1000+ businesses. He’s written hundreds of articles on content strategy, SEO and Analytics. He’s also the author of Content Chemistry: The Illustrated Handbook for Content Marketing.
Leave a Reply