In this episode of The Animalz Podcast, Alex takes us from the coffee shops of 2013 San Francisco to the cutting edge of AI content strategy. He discusses the "spray and pray" approach some marketers are taking with AI and reveals how he conducts "treasure hunts" with AirOps customers for overlooked organizational knowledge that becomes the foundation for distinctive content.
"We often get on calls with customers and they don't even realize that they have treasure troves of usage data." As zero-click queries and AI-generated answers reshape search behavior, he offers both high-level strategic thinking and tactical advice throughout the interview.
About Our Guest: Alex Halliday
Alex Halliday is the founder and CEO of AirOps, a platform that helps brands drive organic growth through AI-powered content strategies. Before starting his company about two and a half years ago, Alex worked at Teespring, Masterclass, and Bungalow, building product expertise that informed his approach to AI workflow design.
Alex's journey with AirOps began in the data space, helping teams pull information from different business areas. When early language models showed potential, the company pivoted to help creative teams get more value from AI than they could achieve out of the box.
Alex has maintained connections in the AI world, including with OpenAI's CEO Sam Altman, whom he got to know during his early days in San Francisco. This network has given him unique insights on where the real AI value is accruing: not just in the models, but also in the implementation layer where AirOps operates.
Insights and Quotes From This Episode
The conversation with Alex is filled with insights, stories, and practical tips, from 50-step AI workflows to insider connections with Silicon Valley's elite.
"We are kind of moving into the perfect information age... you can actually consume and synthesize content at scale." (00:04:00)
Alex explains how AI tools enable knowledge workers to process vastly more information. He acknowledges that summarization causes "fidelity loss," but explains how strategic use of AI research tools can help marketers become "wider readers" and more sophisticated content consumers.
"Where is the actual value in OpenAI? ... Right now, arguably it's more at the application layer." (00:14:00)
Alex argues that while many thought the value would accrue to the base models, the real opportunity is emerging at the application and services layers — where companies like AirOps help organizations implement AI in practical, business-oriented ways.
"We often get on calls with customers and they don't even realize that they have sort of treasure troves of usage data." (00:17:00)
Alex describes how his team conducts a "treasure hunt" with clients, looking for hidden pockets of unique knowledge that can be turned into distinctive content. These assets, whether usage data, support tickets, or expert insights, become the foundation for content that AI alone can't generate.
"There's a gold rush with people creating a lot of lower mid to low quality content with AI... What we'll be left with is a real emphasis on finding frontier opinion and unique data and telling unique stories." (00:18:00)
Alex predicts that as AI search features like "zero-click queries" become more prevalent, basic informational content will lose value. He argues that content success will depend on three types of content that are "impossible to copy:"
- Expert opinions.
- Unique metadata and proprietary data.
- Rich experiential content beyond basic text.
"We have people building workflows that are maybe 30 to 50 steps to create a single piece of content. But what you see out the other end... is so far away from what you'd get out of ChatGPT." (00:21:00)
Unlike simple prompting, high-performing content requires multiple models, sophisticated retrieval strategies, and systems for AI to check its own work. This sophistication produces results that bear little resemblance to basic ChatGPT outputs.
“We are now in the business of not only having to persuade a human being and the Google algorithm that your content is valuable, we now have this new middleman, which is the LLM-based experience.” (Approx. 00:29:00)
When discussing optimization for AI-driven search, Alex points out that it’s “very poorly understood” how to ensure your site or brand gets cited in large language model summaries. The key seems to be “rich content that’s not too fluffy.” In other words, factual, question-answer–style copy that’s easy for LLMs to parse and cite — rather than long-winded “romance copy” that’s purely decorative.
"I love our focus. I think some people misunderstand it, which is a huge advantage for us." (00:36:00)
Despite AI's potential, Alex explains why AirOps focuses on content creation rather than expanding into other areas. They've uncovered unexpected complexity in the seemingly simple problem of AI content generation, requiring eighteen months of work to take models from "can do something" to "creating on-brand, production-grade outputs."
About This Season of the Animalz Podcast: AI & Content
Hello... is there anybody out there creating real value with AI?
The AI conversation in content marketing has become deafening — skeptics shouting from one side, shallow tips from enthusiasts on the other. But somewhere in this noise, there must be pioneers who've actually figured something out, right?
We've gone on a search for the real pioneers — the ones who've ventured beyond the hype to succeed (or fail) spectacularly. Through their hard-won insights, we'll discover if there's actually something of value hiding in the noise, or if we're all just shouting into the void.
Check out other episodes in the season here
Links and Resources From the Episode
AirOps (00:08:00): Alex's platform that helps brands create advanced AI content workflows.
OpenAI Deep Research (00:02:00): A research tool Alex uses to discover content he "wouldn't have found otherwise."
Teespring, Masterclass, Bungalow, SocialGo (00:07:00): Companies where Alex worked, mentioned when discussing his background before founding AirOps.
ChatGPT (00:17:00): Referenced as the foundational AI model that more sophisticated workflows build upon.
Claude (00:20:00): An alternative AI model mentioned for multi-model workflows.
Schema.org (00:30:00): HTML schemas for information structures to improve AI readability.
HubSpot (00:24:00): Referenced when discussing content that strays from topical authority.
”Everybody Wants Thought Leadership Content” (00:43:00): Animalz article on "earned secrets" mentioned in the wrap-up discussion about finding unique organizational insights.
Follow Alex Halliday on LinkedIn or reach him at alex@airops.com to discuss AI content innovations and opportunities.
Full Episode Transcript
Alex Halliday: [00:00:00] You know, you think that a typical piece of content would maybe be five to 10 different steps or maybe even less, but really to get to excellence and get to performant content, you realize like it takes a lot more, it takes a lot of testing, it takes a lot of like mixture of models, it takes different.
Retrieval strategies. It takes a lot of reflection of the models, um, to make sure that they're checking their own work.
Ty Magnin: Welcome to The Animals Podcast. I'm Ty Magnan, the CEO at Animals. And I'm Tim s the Animals Director of Marketing and Innovation. This season on The Animals Podcast, we're focused entirely on AI content use cases, which means we're bringing you on a journey with us to meet with AI pioneers, those venturing beyond the hype to succeed.
Or fail spectacularly. Today's guest is Alex Halladay. He's the founder and CEO of Air Ops, which is a platform to help content teams make better use of tools like GPT and Claude. [00:01:00] Alex is a serial entrepreneur. He, from a very young age, was the founder of a successful startup called Social Go. Uh, he's led product at large companies such as Teespring Masterclass and Bungalow.
I'm excited for you to hear today's conversation with Alec. Did you know Animals Now offers a podcast service. We're taking over your audience's earbuds, reaching them during their commutes, their workouts, or when they're doing tours around the house. From show strategy to editing and distribution, animals can handle your podcast for you.
With that same originality and audience first approach that we bring to all of our content. Every podcast episode can become fuel for your broader content program. You can mine your podcast for ideas, for articles, social posts, and other kinds of content assets helping you create more high quality work in less time.
Ready to start a podcast worth listening to. Head over to animals.co. Book a call with us and we'll start talking about your podcasting goals. Well, Alex, thank you so much for hopping on the [00:02:00] Animals Podcast. Uh, we start all of our episodes with one critically important question, which is what content have you been consuming lately?
Alex Halliday: So I have, uh, spent a lot of time actually using, uh, the new opening ai deep research feature to find pieces of content that I never would've found otherwise. So I'm really interested in kind of frontier, SEO and interesting things that are, um, kind of emerging as trends that we should be aware of. And deep research does this amazing job of finding things that just wouldn't come across my desk normally.
So, um, I've been going deep in kind of Reddit threads that I wouldn't have found. I've found a couple of interesting, um, frontier projects that are really oriented around helping. Agents use websites. Um, and so there's a project called LMS txt, which has been, uh, getting some traction. I've also been reading, uh, a little bit more on the kind of business strategy side as well, especially as we think about, I.[00:03:00]
The next couple of years of the company and where we want to go. Um, I've been, uh, I've been really going deep on, um, some of the kind of like OG business books and things that, that, uh, that I think I must have read probably 10 years ago, but starting to kind of resurface. So, you know, it's, it's, it's, it's, we're living this amazing time where kind of information discovery and now synthesis is like really, really incredibly aided with different AI tools.
And so I'm becoming a much more, I think. Sort of wider reader and, and sophisticated consumer of content with, uh, with the a of these tools.
Ty Magnin: Look at that. So you are using AI to read about AI to inform your AI company?
Alex Halliday: Yeah, I think, um, you know, we, we are kind of moving into the per like a perfect information age, which is like, you know, you used to have like limited bandwidth to kind of synthesize content.
Um, just 'cause you can only read so much in a day. And now if you use these tools in the right way, you can. Actually with the right questions and and the right [00:04:00] framing, you can actually consume and synthesize content at scale. Now that does, that's not the whole picture, right? 'cause you lose a lot of fidelity when you summarize things.
But I do think it's really interesting to at least understand the pockets you can then double click on and kind of go, go deeper on. So I think the way that we consume and understand topics is gonna become. Um, supercharged with these experiences when used in the right way. And, and, and those that are kind of ahead of the curve are gonna have like a little bit of an unfair advantage, um, by adopting these tools early.
Ty Magnin: That's awesome. Tim, are your consumption habits similar?
Tim Metz: Yeah, I would say, well, I mean, even some of this interview was researched with deep research, so this is, it's kind of full circle, but I, I, I think, I think the perspective is interesting 'cause some people say we get, actually get lazier because we can just summarize everything.
But you're kind of saying it makes me smarter if I use it in the right way. Is it, is it, do you see where you use it differently from people who might actually make themselves lazier by using AI too much? To, to summarize everything,
Alex Halliday: there's definitely a, a sort of [00:05:00] fidelity loss, right? As you, as you summarize, you lose resolution, but you can consume a lot more.
So I think depending on your, I. Where you're at in your, like, understanding arc of a topic. It's kind of really helpful at the beginning. And then you probably want to go back to source materials towards the end. And so, you know, when I'm, when I'm trying to, when I'm trying to get smart on something, maybe working with like a new customer and I don't know anything about their space or I feel like there's a, there's a technology we should be considering.
I think it's a really great kind of, it's almost as if you gave like an MBA student, uh, a a day to just go research something and you get back. Kind of something that's like, you know, it's not a hundred percent right, but it certainly gives you a good picture, um, through the lens of your specific problem, which I think is really helpful.
And then, you know, it's up to you to then use, use your, use your sort of time to double click, go in and, and, and sort of go a few, a few layers deeper. Um, the laziness point is interesting. I separately think that AI obviously removes a lot of friction for a lot of tasks. I think it's a mixed bag. I think some things we are greatly aided.[00:06:00]
By its introduction and other things will be real losses actually for us. Um, some things that, you know, were, were hard, were incredibly rewarding because they were hard. Um, and I, I always think about learning to write as a, as a younger person, uh, that took me like five to 10 years at, at school, just writing nonstop.
Getting good at that. I don't know if I would've ended up as a good of a writer had I known that I could. Kind of run over to chat GBT at any point. So, you know, AI's a tool, um, and I think we all have to figure out where to apply it in our businesses, in our lives in, in a balance that kind of gives us net, net positive value.
But there are for sure like risk, risk modes as well.
Ty Magnin: Well, that principle sounds like it is at the core of what and how you're building Air Ops. Maybe just taking a step back for a second, Alex, can you give us your intro and tell us a little bit about. Your background and how you got to Air Ops, how you decided to, to focus here for this chapter.
Alex Halliday: [00:07:00] Yeah, absolutely. So my name's Alex Halladay. I am, uh, one of the founders of Air Ops. We help brands drive organic growth through content strategies and the creation of high quality content. We have been working on this platform for about two and a half years now. We originally started pre kind of LM wave.
Working a little bit more in the data space, so helping teams kind of pull data from different corners of their business and, and bring them into frontline teams. And so, um, that was the first problem we were building on. And we spent about six months there. And it was through applying early language models to that problem set that we had our kind of, you know, head explode emoji moment.
And we sort of realized that this technology just had. Such amazing potential even in those early days, pre chat GPT and uh, kind of just through a process of, as you do so often in early stage startups, you sort of pull a thread and you see where you land and you talk to customers and you learn. [00:08:00] We, we really realized that the business we wanted to build was to help really creative builders get more out of these models than they cut out the box.
And so through just a process of iteration and, and a lot of hard work we have. Um, kind of worked with probably a couple hundred companies now to take the raw knowledge and capability of something like GPT-4 oh or a. Or an oh one top style model. Take that raw, raw capability and turn it into a business process and business system that drives growth.
And so it's along the way from raw capability to business outcome. There's all these things that need to be solved. It's not like, you know, it's not necessarily how Sam Altman like say, where these models just, you know, able to kind of change our lives instantly. In many situations, they need a lot of.
Guardrails. They need a lot of thoughtful context and support. They need a lot of human oversight, and in some workflows that means they can help with 30% [00:09:00] of what needs to be done. In some cases, it's 80%, but almost universally in every deployment of AOPs. For every content creation or refresh workflow, there is a good amount of human oversight.
At multiple points in the funnel. And I think that's the model that we like. That's the model that we think works. And, um, case by case, customers with different comfort levels will figure out the right mix for them and the problems that they're trying to solve.
Ty Magnin: Awesome. Yeah. That's the model that would resonate with animals too, right?
One that allows. People to create content a little more efficiently, but is still using skilled content marketers to actually, you know, sort of usher this thing along a process that they're taking anyways, even without the assistance of ai.
Alex Halliday: And I would say what we're actually seeing is a trend is that the kind of language models really, and, and obviously things like AI search summaries that you guys are very familiar with as well, you know, they are really starting to eat a lot of the kind of what call top of funnel terms or, or kind of terms that [00:10:00] are, you know, easily.
The needs are easily met by, by sort of the default training of language models. And so where I think the emphasis starts to move more is um, into experiential content. So video and things that are visually richer, but also kind of frontier opinion, unique content. Things that like you as a business are uniquely.
Uh, position to, to sort of put out into the world. And so we worked with a company where they had, we realized they had all this amazing data on lawns across the us And so what we worked with them through to do is to create, uh, local pages, which gave a ton of really rich information around lawn care that was very data driven.
There's a lot of metrics on the page, uh, to help inform people on, on the seasonal lawn care and pH levels, all these things. But that is a great example of something that was AI assisted. Human oversight, but really leveraged their unique domain expertise and something that the models out the box wouldn't be able to answer.
So we love those kinds of like data stories [00:11:00] or opinion stories that um, really are just, just unique and, and authoritative and perform well.
Ty Magnin: Nice. Alex, I heard you have some kind of relationship to Sam Altman and the Altman Brothers. I think that's pretty cool. I think he kind of legitimizes. What you're doing to an extent even further, can you share a little bit of that just for us?
Fanboy here?
Alex Halliday: Yeah, a hundred percent. Um, I mean, for context, I, I moved to San Francisco when I was, uh, oh. In 2013. I won't say how old I was. Um, and, and, and I, I think the kind of, for those of people that were in San Francisco around the sort of 2013 to 2016, I. Era. It was a magical time. I mean, the, the, the, it felt like the tech world was there at the kind of blue bottle coffee and, you know, you'd see sort of Jack from Square in those days, uh, uh, walking around and, and Sam Orman and Peter Te, all these people were just, they were just part of the neighborhood.
And so, you know, I, I was. I was there, um, kind of right from the uk and [00:12:00] my mind was blown by this like magical tech wonderland I was a part of. Honestly, I, it, it, it almost gives me goosebumps thinking about what incredible time period that was to live in that city. I. And, and you know, through through living there, I, I got to know Sam and also his brother Jack, who I worked with, uh, at Teespring.
So Jack sat a couple of seats down. Jack almost sat a couple of seats down from me at Teespring and was the, honestly the heart and soul of that company. And I, I. Adored him working with him. He was awesome. Um, and, uh, and, and Sam at this point was in between, I think when I first met him, was in between Looped and yc, and then he went onto his yc, his YC base.
And, and you know, we, we, we talked a lot about technology and the future and YC and startups, and he's always been someone that's kind of. He's kind of the person, like, if you really can't solve a problem, you, you have to give Sam a call sometimes. And he, he's like, he's, he's probably the best tech and early stage company systems thinker [00:13:00] and just like, really, how, how do you solve like these really interesting strategic questions and often offers like.
Really incredible insight. So maintain that relationship over the years. And his brother Jack, um, was at the time running kind of the Altman, I think, family office and, and was one of our first investors, um, when we started AOPs. And so, uh, yeah, we've had a, we've had a few calls with, with them and got, got got insights.
And I'll say like, Sam doesn't really let me in too much ahead of, of the public in terms of what open AI is cooking. Uh, I have tried, uh, but, uh, but, but certainly his advice has been a bit invaluable. So,
Ty Magnin: Tim, look at that. We've made it however many episodes into this season focused on AI without even mentioning Sam Altman.
But here we are, you know, sitting just, you know, second connection away. Thanks for that anecdote.
Alex Halliday: I mean, one thing I will say is that if you just look at that business of OpenAI, it's actually pretty. Interesting where the value is accruing. I think, you know, there's, for a long time [00:14:00] everyone thought the LM models were where all the value would accrue, but that's actually, you know, the, the, the sort of value to K curve, these models is insane.
And so where is the actual value in open ai? And I'm sure they have models that are coming, they're incredibly impressive. But right now, arguably it's more at the application layer. Um, with chat GPT and why this is relevant to audience is because actually beyond the application layer is the services layer.
And I actually think what we're seeing right now is that this AI model shift the kind of investor sentiment and the, the OB just observing this play out. The, the kind of narrative has shifted so much from the models are gonna just gobble up all this value to actually its services and its thoughtful implementation.
And it's applying these to real world business challenges like marketing and growth, where there's a lot, there's a lot of present day value to be realized and a lot of really great businesses that are being built. And that's what we, I mean, that's, that's where we see ourselves and, and what gets us really, really excited.
Ty Magnin: Alex, you, you recently published a blog article that's called The New [00:15:00] Rules of AI Content. When by being impossible to copy. And you hit on this point already a little bit like kind of the philosophy of how you think AI and content marketers or marketers at large kind of play together to create a winning formula.
Mm-hmm. Can you kind of outline that philosophy, uh, similarly the way that you did in the blog article, or at least just the way it comes out today?
Alex Halliday: Yeah, and I think going back to the perfect information point, know consumers. Are gonna have increasingly amazing superpowers to synthesize available content and make it available to them.
And so, you know, your job as a, as a marketer, um, is to ultimately make sure that your customer's brand is appearing where your customer is making decisions. Be that in search, be that in sort of chat GPTs of the world or, you know, series about to go through its own revamp in, in a couple of versions. And so.
Where the content kind of [00:16:00] battle plays out and how people are successful is shifting. And so I think as we, as we see, you know, kind of informational queries or like, we think of like zero click queries are starting to eat a lot of the early kind of the, the, the kind of og kind of content opportunities.
And so where do we go now? And I think what we have seen again and again is that. There is certain types of content that are just continuing to become more impactful and more relevant. So along with authority, which is obviously incredibly important, we could do a whole podcast on the types of content that I think are really hard to copy and are, uh, continuing to be worthwhile.
Investments can fall in a couple of different buckets. I think the first one is. Really authoritative opinion. So experts inside of your company or experts that you find just really understanding kind of their point of view on a topic and making sure that LMS in particular can understand the origin of that content and why that particular [00:17:00] author is authoritative is something that's really, really important.
So, so for example, you know, we are. Um, working with customers now that, um, will spend a lot of time creating questions using AI that then their team answers to then feed back into content in order to make sure they have that, that content grounded in original opinion from experts internally. That's just like one example.
The second category is unique kind of metadata, and so that's instances where. Um, you know, you have unique data sets internally and the thing that always surprises me is that people just don't know what they have. So, you know, we often get on calls with customers and they don't even realize that they have sort of treasure troves of usage data.
Or maybe it's, um, like deep, maybe it's PDFs or, or, or kind of internal docs for. For, for like, wine is one of our, one of the recent ones I worked with where they had all of this information that they could leverage, um, and, and were able to turn that into kind of unique pages with unique metadata that was, that was telling a unique story, adding value to that customer's search [00:18:00] journey.
Um, and then, then, and then the third one is, is really kind of experiential content. Things that take time to produce that're not just text-based and that are really able to. Um, so captivate the user and meet their need and something that, that, that really is like a kind of cons, um, destination content, as I've heard it called.
And so I think you could imagine that like there, there's, there's kind of a gold rush a little bit with like people creating a lot of like lower mid to low quality content with ai. I think that's gonna shift a little bit. As zero click queries kind of emerge and, and, and continue to, to erode into top of funnel.
And I think what we'll be left with is, is a real emphasis on finding frontier opinion and unique data and telling unique stories and delivering great experiences through content. And so that's your point about, you know. Humans being an important part of the content production process. I actually think that continues to be the case.
And, and AI is an important tool, but really, uh, an emphasis on quality, originality and authority is gonna just become ever so [00:19:00] more important. The other thing I will say is I think depth. Depth of content in the, on the basis that people can synthesize instantly using ai, actually making sure that there is really great depth in the content that's created and it's, and it, and actually making potentially longer pieces of content is, is something that I could see, uh, increasing as well.
Um, just in, in this new era.
Tim Metz: I, I, I think we should, because I, I realize 'cause we're talking a lot about Arabs, which makes sense, but I think we three are all familiar with what it is. But if as a listener you might not completely understand. What the platform looks like or what it does. So maybe you can walk us, like, kind of paint what one of these workflows would look like, maybe with the wine example or another example.
Just like, can I pitch Air Ops the way that I see it and see how, please see, see how my messaging resonates
Ty Magnin: with you, Alex. Yeah. The way I have described Air Ops to friends and colleagues is that it's like a canvas, uh, for you to build workflows, to create content. And in that canvas you can [00:20:00] have all of your prompts laid out and you can query, or you can like run those prompts against different LLMs, you know, depending on what's the right fit for that exact query.
So you start with perplexity to do the research and then you could, you know, use some, you know, GPT and then you can actually use Clot if you like, the way the Claude writes, that kind of thing. Yep. It's a lot more sophisticated than that, but again, like a canvas to build these like. AI workflows.
Alex Halliday: Yeah, I love it.
I think, uh, it's, it, we, we see the best results from instances where people have big ideas that they wanna build. And so, you know, I think AIOps is an incredibly flexible platform. We have a workflow builder and a canvas that lets you build with a mixture of AI models, uh, internal retrieval. So pulling insights and nuggets from your own data and content and knowledge.
And then also external data sources as well. So we, we basically allow you to compose with all of these pieces. And, you know, you'd think that a typical piece of content would maybe be [00:21:00] five to 10 different steps or maybe even less. But really to get to excellence and get to performant content, I. You realize like it takes a lot more, it takes a lot of testing, it takes a lot of like mixture of models, it takes different retrieval strategies.
It takes a lot of reflection of the models, um, to make sure that they're checking their own work. And so, uh, we have people building workflows that are maybe 30 to 50 steps to create a single piece of content. But what you see out the other end when you first look at it is so far away from what you'd get out of a chat GPT that suddenly it clicks.
It's like, oh, actually. You know, I, I need to transpose my knowledge of my business and my customer onto this system and have AI do as much of it as I'm comfortable with it doing. And then I, as a human, uh, partner need to step in and perform tasks along the way to course correct. Provide feedback and make critical choices.
And so we really see ourselves as the platform where big ideas kind of get built out [00:22:00] and tested and, and also we give. We put all of these amazing raw capabilities in the hands of creative, uh, marketers to then, then bring these big ideas to life. And that's really it. When, when people have these big wins or they make something amazing and it, they get a great outcome, that is the best moment at AOPs in the week that, that goes in our wins channel.
The team comes to life, makes me really happy. It's why I do this. So, uh, yeah, that's what it's all in service of.
Ty Magnin: Nice. Same. Yeah. I mean, we kind of have this, yeah. We love making customers, you know, grow and win with content too. Yeah. So, uh, we, we get off on the same, uh, same thing. What would your advice be to folks that are doing what we call the spray and pray SEO technique?
You know, say they're doing a hundred articles a month, they're lightly editing it. Right. I made quote fingers for those listening, uh, which means like, maybe not editing, whatever. Someone might look at it quick. What would you tell 'em?
Alex Halliday: I think a hundred article. So [00:23:00] great question. So for people creating larger volumes of content, um, I think that there's definitely a relative element to this.
If you have millions of pages, a hundred articles a month will be kind of a, it could be a totally reasonable amount, I think for there is a window of opportunity. Where I would say kind of in general, informational content can still perform well and there's still an appetite for it. If you, if you have authority in that area and you can rank, there's still a very valid play to create quality.
I. Authoritative content to get at those top of funnel queries for a period of time. Bonus points if you are formatting the content in a way that is, uh, highly optimized for answer engines, which looks like more question, answer style content. I think there's, there's definitely plays there that are valid.
I think the, the more you stray away from your topical authority, the more risk you introduce. It's like the HubSpot example that is [00:24:00] very well publicized is, you know, they created, they created all this content that's like, you know, just completely unrelated to HubSpot. They created a scale. How do you tie your shoes
Ty Magnin: in?
Chew gum. Yeah. All. How do tie your shoes crap.
Alex Halliday: Yeah. I mean, to be honest, I've been in companies where we have created totally unrelated pages that have done very well. So I know how this happens, but it is, it sort of. It's probably no skin off their nose, given that I don't think that was converting into a $40,000 a year HubSpot contract, uh, that piece of content.
But, but I think, I think there is, there is a window of time where general informational can work. I think it's changing a little bit, but in terms of the kind of the real gold that will be. That will rank be cited and will be persuasive to both agents and end users, right? Because we have a new middleman to convince now.
Uh, I think just really thinking about how am I adding to the conversation? How am I meeting this search intent better than anywhere else? And, and, and what unique. Kind of insights, opinion, or data can I bring to the conversation? And I think if you can, if you can kind of balance the, that, those [00:25:00] two realities, you'll, you'll be set up for success.
But as with all things, and I'm sure you guys know this too, you do this with your, with your customers, there is a need to be, be keeping a very close eye on what's working, test and learn. Don't. Dump 10,000 pages overnight. You know, there's, there's kind of responsible ways to do this, um, yeah. In order to make sure the health of the system continues.
Ty Magnin: Yeah. We emphatically agree. Right. You know, I think about like, if you don't have the time to read and edit this thing, like why should your audience, you know, what kind of relationship are you building with them if you're just like publishing crap? And so like, yeah, there might be short term arbitrage here, but at animals, you know, we like to think about long term solutions more often than short.
Alex Halliday: Yeah. We've been thinking about. This idea, like where does all the frontier opinion live for your, in your business? Like if you're talking to, you know, if we're talking to a, a wine, like a wine merchant for example. Like, okay, so there's probably locked inside that organization. Some amazing nuggets, whether they're in support tickets or they're in tasting notes, or they're in interviews they've done [00:26:00] with potential vineyards.
And so I love this like treasure hunt a little bit with customers where it's like, okay, like how can we like just start telling me about your business and let's see if we can find some unique little gold, sort of like gold veins that we can then go mine because. I think that that's, that there's, there's sometimes some really great opportunities there that people don't even realize.
And do, do
Tim Metz: you do that treasure hunt with a traditional content team? Like let's say with the wine business who comes to you? Is it, is it still content teams or is it totally different kind of roles and skill sets that work with you? I.
Alex Halliday: I mean, the amazing thing is just the spectrum of resourcing that companies have.
Like I, we talk to some companies and they're super organized and they come to us with, you know, I'm sure you have this too, right? There's some companies come to us and they're like, they have everything cataloged and they've got their Google Drive folder with everything in, and they, they're like very buttoned up and other folks just have no.
No resources at all. 'cause maybe they've, they've never built that muscle internally. Um, so it does vary a lot. I think we, we love, we love, we love to find a content team internally. We love to work [00:27:00] with an agency that a, a company's already engaging with. Um, I mean, typically like, we really like to find that champion that gets it, that knows where all these things are, are located, and also really understands the business because there is that kind of, you want that kind of customer.
Um, you want that kind of customer audience persona ingrained in the head of your counterpart in these companies. It's 'cause they know what their audience wants. They know how they like consume content. They, they know what they're looking for and they can also just like, they can just, they can just tell us something passes the sniff test or not.
You know,
Ty Magnin: Alex, you were starting to make mention of how to optimize content for. AI overviews or you know mm-hmm. How to do generative engine optimization, whatever the hell we call it today. Yeah. You know, someone needs to decide other week. Do you have any perspective on tactics or strategies that you might advise teams to, to, to use, to optimize for generative search?
Alex Halliday: Yeah. I, I, I sort of love this stuff to be honest. Um, in part because [00:28:00] we got product market fit with content marketing. Pretty early on in the business, and I felt like on that front I've been learning a lot. And on this front I feel like nobody knows anything yet. So it's really awesome. Um, so, okay, so, so I think, you know, there's a lot of people claiming to have certainty in this when there really isn't.
Um, and that, that sounds good, uh, for a sales pitch, but the reality is these models are shifting, the retrieval mechanisms shifting, you know, it's, it's all, it's going through a lot of change. And so. One thing's for sure. It's like test and learn. We would say this, it's just like you've gotta, you've gotta kind of figure out things for yourself and you've gotta also, um, make sure you are just staying up to date with, with what's happening because, um, there's a lot of change that's happened recently and is coming.
Things that I. Still matter. Obviously, uh, SERP authority really matters for your topic. SERP is still used as the primary ranking mechanism for retrieval in AI based experiences. So chat, GPT will make queries, uh, uh, to bing [00:29:00] search in their case, will retrieve some pages, synthesize them, and return to users.
So you know, all the things you do to get authority in your, in your relevant topics, you should continue to do that really matters. I. The, the next thing is the actual content you are creating, um, on the pages themselves. So there's a world in which you rank for the, the, the, the, um, the query, but you don't get cited.
And so, um, you know, the, the, you're now in the business of not only having to persuade a human being, uh, and the Google algorithm that your content is valuable. You now have this new middleman, which is the, the agent or the, the kind of LLM based experience. And so. Making sure your content is cited is, is again, something that's like very poorly understood, to be honest.
However, what seems to be working is kind of rich, uh, content that is not too fluffy. So for product descriptions, that means, you know, being very clear, uh, on the, on the, on the sort of facts. In the product description and, and moving away from romance copy as they [00:30:00] call it in e-commerce. Um, and then potentially breaking down certain chunks of content into question answer formats, that, that more closely match the things that customers are gonna gonna want.
So that, that, those are, those are sort like, that's fairly. Understandable. I think like just from a practicality standpoint, it's gonna gravitate towards chunks of content that are not too long that meet the question intent that the user is expressing. I think beyond that, I would say that like only Google Bot can render JavaScript, so making sure that the content you're publishing is in like the underlying HTML.
So don't use any of these like Js SEO tools 'cause they're not gonna help you. Then also there's sort of increasing evidence that the, um, that the schema, so schema.org has all these different, uh, html schemas that for information structures. So having your content structured in that way helps with readability for certain types of content.
I think this is all gonna change. So I, I, I have a few ideas about where it might go, but I think that for now, those are some of the things that we're [00:31:00] observing and recommending.
Ty Magnin: Funny, I've, I've also heard contradictions to what you just said around like schema, particularly being something that is more ignored, uh, by LLMs.
'cause they're just reading the straight H-G-M-L-I describe it as like it's a really opaque black box. Like we always say SEO is a black box, but like, is it really like we've reversed engineered this thing pretty well over the last decade or so. Yeah. But now it's like we have a new black box and we haven't fully reverse engineered this thing yet.
Alex Halliday: I think there's a, there's a, there's also a sort of new, new crop of tools that are claiming to kind of help with understanding your, your sort of citation rate and relative ranking. I think I've, I've looked at a bunch of them. I think they're, they're still so early that, you know, it's, it, it's just much like search was at the beginning.
I think it's just gonna take a little bit of time for the rules of the road to really emerge and the, and the tooling that we, the tooling stack there to emerge as well.
Ty Magnin: I find it interesting that. You find this topic interesting, right? I mean, as the CEO at Air Ops, like why do you care so much about the [00:32:00] latest tactics around, you know, GEOI
Alex Halliday: care about this because ultimately we're trying to help customers put high quality information in front of customers that are making, buying decisions and, and seeking to learn.
You know, I think that, that, that kind of. There is a very real world in which the majority of content consumption happens via, uh, an agent based experience in the next 36 months, and it just, Chrome starts to incorporate this at the browser level. We obviously Google search summaries expand and, and chat GBT and perplexity continu to grow.
So I'm interested because it's, it's a, it's a fun problem. Um, it's an exciting. New paradigm. I think for things like comparison shopping, it just changes the game. Uh, and, and so that's great for customers and, uh, and, and, and so people that wanna really go one click deeper and it presents a new set of challenges for our customer and our partners.
And so I. To the extent that we can be educating and building product to [00:33:00] support and, and sort of commissioning research where appropriate. I think that's a really important role for AOPs to play because we never intended to be just an SEO content. Optimization tool or, or creation tool, right? We wanna, we wanna create, uh, engines that drive growth through the creation of content and meet customers wherever they happen to be.
And so that's where, that's where this emerging like landscape really becomes exciting for us because we get to, we get to lead and, and, and do some really exciting frontier work. And on the product side, like I think that's where we're really excited as well to start to ship things soon that will support our customers in, in this direction as well.
Ty Magnin: So, Alex, I've. I'll confess, I've been a little skeptical that Air Ops would stay in the niche of content marketing in perpetuity. Do you really plan to stay here and continue to, to focus on this market? Or do you, is this the beachhead and you really do hope to expand? Um, 'cause it feels, I want you to know as a content marketing agency owner, like [00:34:00] I feel the commitment and I think it's wonderful.
But I'm always like, you know, the name is AOPs. Right? And like, there's a ton of other things you could go automate. Help us understand that, that vision and and broader go-to market expansion.
Alex Halliday: It's a great question. I think we have a, we have a major focus and a minor focus. You know, we spend probably 70% or 88% of our time.
On the core sets of use cases that are similar to the, to yours, right? So we, we really care about research brief creation content and, and increasingly things like content refresh and health and, and metrics. I think that's like a, a little bit of an illusion of a problem set, meaning I. When you first look at it, you're like, oh, I get this.
You like slap an LM and you create some content and you're like, that's a should be a week of product development work. And then you have something. And the answer is like, absolutely not. Like the more we peel back the onion. Yeah. And the more we work with brands that have really high standards or have regulatory concerns, like the more product there is to build, the more things we have to solve.
And so [00:35:00] I've loved just being like. Really surprised and in some places humbled at what it actually takes to get these amazing models, AI models to be actually useful in real businesses like crossing the chasm from like AI can do something to AI is actually creating on brand production. Grade outputs has been like a full team's effort for the last 18 months and we're still only a fraction of the way.
Of, of the way there, honestly. So I think that, that, it's always fun to tell a big story in a startup of like, we let you build growth systems, right? Which is was our positioning for a while. But the reality is like that is very hard for someone like you to, to assess like the fit of a platform that makes such a big promise to your real business.
And so we've enjoyed going really deep. I think we've enjoyed building relationships with, with, um. Content marketers and content teams and and organizations where content is a core part of that, what they put out into the world and, [00:36:00] and, and what they, what value they want to deliver to their customers. I think for us, it makes way more sense for us to continue down that path and then sort of think about the next adjacency that we want to add on, rather than going to something like sales, for example.
Like sales would be a total, like, would be a totally different conversation. Different buyer, different customer, different problems there. And so. I love our focus. I think some people. Misunderstand it, which I think is a huge advantage for us. Like we kind of hide in plain sight to be honest, versus like I SDRs or something, which is like 300 startups pursuing and, and when you meet and, and sometimes like you meet customers and they don't get it, but more often than not, the people that really care about their stuff and they, they love it and they care about.
The conversation that's, that they're having with their customers at scale. We, we like, we have a really great partnership and, and they end up being like really great customers for us. So no regrets. And I think we, we will expand, but it probably won't be like, into a totally different space. It'll be continuing to [00:37:00] build on, on the beachhead and, and hopefully provide more things that, that folks like you would find valuable.
Ty Magnin: I think it's a great community, you know, to serve, uh, a lot of interesting intellectual people that, that, you know, have a creative spark.
Alex Halliday: Yeah, we call, we call them moonshots. So we call, we call, we call folks like you moonshot marketers, where you, like, you basically have these ideas that would've been impossible two years ago.
And we, and, and between us, we help you bring them to life. And, and it's like, and that, that makes me really happy that we're able to kind of give. Give people the rails and the tools and the, and the, the Lego blocks to go build things that get them really pumped to get their customers great outcomes. And it's just like, for us to be able to facilitate that is like absolutely awesome.
I love it. Yeah. It's very, very motivating for, for the team and for me.
Tim Metz: I think that's a, that's a perfect bridge to one of the last questions we had on the list. What is, what is, what is a moonshot project or idea that you'd like to see realized with AI in the coming
Alex Halliday: one to two years? Moonshot idea for ai.
So I do a lot of [00:38:00] dictation to, to models, and I, I like to walk and talk and then, uh, ask it to like synthesize it, whatever, and come back in the morning. So this is how I clear my head at the end of the day, right? So this is like pro tip to anyone. If you feel overwhelmed at work, go for a walk, just. Just talk for 45 minutes and let AI organize your thoughts and then go and take a nap, come back to it.
Life will be materially better after that. Um, but what, what I think is helpful is kind of this AI as a thought partner and, and, and I, so I'm using it on the individual level, but I actually, I. Would love to have a better kind of context rich assistant that could actually participate in meetings and would, would be able to kind of provide like real time feedback with a good understanding of the bi of a business for like quite hard problems that we're dealing with.
I think I. You know, you sometimes in a, in a small team reach a kind of local maximum of where you're able to take the conversation, and I think having an AI participant added to like a debate or a, or a brainstorming session could be really helpful if it, if it has the right context. I'd like someone to build that, please.[00:39:00]
Tim Metz: Nice. Yeah, I know. I got it. Yeah, I thought, I've been thinking about something similar for these podcasts, for example, where you could kind of have this head up desk display and it will say, this is the best next question to ask. It was kind of listening in and you had that for a. All kinds of meetings, right.
Where it just, uh, shows you like, you can go here, you can go there and like Yeah. I think that'd be awesome. Yeah.
Alex Halliday: Or you can create one which is like, Hey, you want one to be a customer and one to be an investor, and you want one to be a, a competitor and you can have like these different actors, right. We actually doing, we're doing more kind of micro like synthetic audience.
LM steps where we ask them to kind of like read things and pick their favorites. And that's a super interesting trend as well, where you can kind of pretest various pieces of content. And I lo I love that stuff. So good, good fun.
Ty Magnin: Yeah, yeah, yeah. Share the wealth there. I mean, I love to put my hands on a product like that.
That's really interesting. Well, it's, it's been. A real privilege to have you. I learned a ton. Thanks for spending some time. Where can folks follow you and, and, and sort of, uh, continue to [00:40:00] learn from, from you and the Air Ops story?
Alex Halliday: Yeah, so we're a, we're a team that loves talking to people, building in the space with, with big ideas.
Uh, so you can find us at, uh, at AOPs HQ on Twitter. Or X and then our, our website's, aops.com. Uh, if you wanna email me personally, always open to chat alex@aops.com. It's an easy one. Um, I think, you know, particularly if you have a big idea that you are, you wanna bring to life, uh, and, and you're not sure if it's possible or, you know, you've, maybe you're working on something really cool and wanna integrate with us.
There's lots of, lots of things we are. We're open to discussing. Um, but I think you know, this, this entire space that we're in is gonna go through a lot of shifts and changes. And so we, we all kind of have to knowledge share and I. Uh, and figure it out together. Um, and, uh, would love to connect with folks who are working on kind of adjacent problems or, or, or kind of curious about how this is all gonna unfold.
Ty Magnin: He also has a great LinkedIn [00:41:00] feed going. You didn't mention. Oh yeah. So Alex Halliday, H-A-L-L-I-D-A-Y. Search him on LinkedIn. Um, I just started following you and have been digging the latest stuff you've been putting out. Guys, huge pleasure. Thank you for having me. This is great. I honestly think, Alex, and generally what I've heard from AOPs is the most philosophically aligned tool partner that I've met.
Actually, I would give a shout out to Kyle Coleman too and like copy ai, but I think the air, like what Alex was saying today was like really aligned to the way that. I have experienced using AI to create content, you know, for side projects, for myself, whatever. Uh, and I've seen now animals also, you know, land on as we've experimented over the years, and the, the nucleus is still like, content workflows are very complicated.
So like, you need a lot of prompting, not just like, write me a blog post on X, Y, Z. You need a person there to check, to check, to check, to edit, you know, to give it feedback, to run it again. 'cause the output was kind of weird. 'cause that's just the age of [00:42:00] AI that we're in right now. My confidence increased throughout that conversation.
Tim Metz: Yeah. Mine also, but in a different way maybe, but. I, I felt it's still early days, right. There's still so much to figure out. It's something I heard echoed also in other places, like even people working on these LM models like at OpenAI and in who are saying like, we also don't really know what's gonna come out.
Right. And we discover new things every day. That's it's, it's not only a black box at like the SEO content level, it's like a black box, like all the way down. Yeah. I think he echoes that sentiment also of like, there's just so much to discover and it's so early, which is, which is really exciting I think.
Ty Magnin: Yeah. I also find when we chat with. Early adopter types, like tinkerers. If we can call you that, if we can call Alex that like, uh, I get also inspired to like tinker more. Right. And like a lot of the answer, I mean, if you take this combination of like it's a black box, people don't really know and like, I like to tinker.
Like that's kind of how you end up having breakthroughs, right? Being really ahead of the game.
Tim Metz: The other thing [00:43:00] I noted was like, I had to think about our, our thought leadership piece on the blog, uh, where we talk about earned secrets. He, he phrased it differently, but I, it was like, you have to look for these pockets of, or or nuggets in an organization.
Like where do you have this goal that can kind of feed into the model that will make great output? It's like, yeah, that's the earned secret, right? Like you kind of have to search for. They earned secrets and they show up in unexpected places and trying to find those. Um, I think that's, that's also a good kind of takeaway.
Like we knew that a little bit, but it's the obvious ones of like, oh yeah, your sales schools or whatever. But like, indeed what he said, like, yeah, the list of, what was it? Like wine tasting notes or, yeah. There's so many things that could potentially go into these models and create unique content. I, I think that's awesome.
That was also good. Reminder. Totally,
Ty Magnin: yes. Like expert led or however you wanna put it. At the end of the day, like it's so funny, all of the kind of like innovations or, or changes in SEO, like it kind of all continues to come back to something that's pretty core to animals, which is like right. Content that people will [00:44:00] find interesting and enjoy, you know, and like, and it's valuable, right?
Yeah. And then it will perform. And so like that principle has gone unchanged through so much change.
Tim Metz: Yeah. And it sounds like that's also where he's also still sees a lot of value or potential for the future. That kind, what he call it, exper experiential content.
Ty Magnin: Okay. And one more piece that I'm still, like, again, I said this very blurry for me, but like AI agents and the way that changes, kind of like the way business is done, the way content is consumed, the way, then content gets created, like, whew.
It's uh, it's fun to kind of imagine into it, but like, man, it. People are really sort of grasping at figuring that out next. And that to me is like the forefront of the forefront. Or like, you know, that's like two steps
Tim Metz: ahead. It's, it takes me five and 20 minutes to come back to you. Right? So what you see, what I do is like, okay, research this and this and this for me, and then, and then it's doing it kind of in the background, right?
That's a glimpse I think, of what's gonna happen. You're [00:45:00] gonna like, almost like you, you assign a task to a coworker or something. It's like, okay, go do doing this for me. And then that's running in the background. So I could actually imagine you have all these kind of things running in the background while you are doing something else, either working or lying on the beach.
But
Ty Magnin: I mean, responding back to other bots, right? It's like, yeah, exactly.
Tim Metz: Yeah. Then coordinate the next one. Yeah.
Ty Magnin: Yeah.
