Thenuka Karunaratne (daydream) on Building SEO Moats with AI

“When you think about what’s happening in search right now,” Thenuka says, “there are two big changes. The first is that more and more search volume is going to AI answer engines. And the second is that the way you create content is becoming much more programmatized.”

Instead of viewing AI as just a content creation tool, Thenuka shares a broader vision: AI can transform strategy development, performance monitoring, and real-time opportunity spotting. His perspective, grounded in years of programmatic SEO experience, offers content teams a roadmap for a future where “computers talking to computers” becomes the norm.

Despite predicting such a future, Thenuka cautions, “The idea of SEO on autopilot, where you have this super lazy approach to using AI to churn out mass amounts of things, that doesn’t work.” But for teams willing to integrate AI into their workflows, the results can be remarkable — demonstrated by his work with Twingate, where his team’s programmatically generated content quickly reached 50,000 monthly visits.

About Our Guest: Thenuka Karunaratne

Thenuka Karunaratne is the co-founder and CEO of daydream, a platform that automates programmatic SEO. Founded in September 2023, daydream has raised $3.8M in seed funding from First Round Capital, Basis Set Ventures, and other investors. Thenuka also serves as an SEO Expert in Residence for several venture capital firms, advising their portfolio companies on organic growth strategies.

His journey in search began in high school building affiliate websites. His first successful venture helped Canadians access American Netflix content. Before Daydream, he founded Flixed, a lead generation business in the streaming industry that sent over 100,000 subscribers to services like HBO, Disney, and Hulu through programmatic SEO.

Flixed provided Thenuka with a crucial insight: in industries with concentrated long-tail search volume, manually targeting each query becomes impossible. This challenge — and the lack of adequate tools to address it — led him to create daydream, using AI to make programmatic SEO accessible to more businesses.

Insights and Quotes From This Episode

Thenuka speaks the language of someone who’s built the systems, seen the traffic results, and measured the conversion rates. His observations challenge the “SEO is dead” fearmongering and the “let AI handle everything” oversimplifications dominating today’s AI and content conversations.

“The idea of SEO on autopilot is very good marketing language, but that doesn’t actually work.” (00:12:00)

Thenuka emphasizes that despite the hype, you can’t simply set AI to automatically generate content and expect good results. If you generate content without customization or human review, “you’re bound to fail” because you’ll produce articles identical to thousands of others. Instead, you need to use AI thoughtfully within a structured process.

“The way you should think about it is if you use AI tools well, can you create a piece of content that would match or exceed the quality of what you would expect someone on your team to do?” (00:13:00)

Thenuka suggests we need to focus on using AI to achieve high-quality results comparable to human work. This requires understanding what data points and structures lead to effective content.

“People talk a lot about AI and content creation — very little about how AI can automate the strategic aspects of growth marketing. That’s a massive area. Because if you think about how a human creates an SEO strategy, it’s a lot of repetitive, mechanical tasks.” (00:16:30)

Thenuka challenges marketers to think beyond content creation when considering AI’s potential. He points out that strategy development — keyword research, competitive analysis, and auditing — involves many repetitive tasks suitable for automation.

“The Twin Gate story… 20 percent of the traffic going to their demo page is all from daydream content.” (00:25:00)

Thenuka shares how his company helped Twingate, a zero-trust cybersecurity product, by creating ~1,000 programmatically generated articles about data breaches and vulnerabilities. This content now drives 50,000 monthly visits, with 20% of their demo page traffic from this programmatic content.

“If AI makes it easy for everyone to generate content, then you’ve got to find deeper ways to stand out — often via data that no one else has.” (00:26:50)

As AI streamlines content production, the competitive advantage shifts toward unique assets that are hard to replicate. Proprietary data, exclusive insights, and distinctive perspectives will become the true “moats” protecting content strategies from competitors using the same AI tools.

“I think for most of SEO, the right background for those people should over time skew more and more technical.” (00:27:30)

Thenuka believes successful SEO professionals will need increasingly technical backgrounds. While certain types of content — like founder interviews and thought leadership — remain unaffected by automation, more technical skills will be required for managing the automated aspects of content strategy.

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

Unbroken (00:01:40): The biography of Olympian Louis Zamperini that Thenuka mentions reading at the beginning of the episode.

Flixed (00:03:00): Thenuka’s previous lead generation business in the streaming industry that sent over 100,000 subscribers to services like HBO, Disney, and Hulu through programmatic SEO.

ChatGPT and Perplexity (00:07:30): AI answer engines frequently referenced in the discussion about changing search behavior and their impact on SEO visibility.

Zapier (00:09:10): Mentioned as a tool that an AI answer engine might suggest for connecting SaaS products.

Daydream Library (00:12:00): Thenuka’s company blog recently announced their AI visibility tool.

Twingate (00:21:00): The zero-trust cybersecurity company in Thenuka’s case study, where programmatic SEO content drove ~50,000 monthly visits.

Coinbase, Airbnb, and Pinterest (00:26:30): Companies Thenuka mentions as examples where SEO has become a product function rather than a marketing function.

Nexus (00:29:00): A book by Yuval Noah Harari mentioned in the post-interview discussion between Ty and Tim about finance and AI.

Follow Thenuka and daydream: Learn more about daydream at withdaydream.com or follow Thenuka on LinkedIn or X.

Full Episode Transcript

Thenuka Karunaratne: [00:00:00] From a practical standpoint, I think what most people want to know now is where am I even surfacing? Like, what is my visibility in the answers? And most people don’t even have that. Welcome to the

Ty Magnin: Animals Podcast. I’m Ty Magnin, the CEO at Animals. And I’m Tim Metz, the Director of Marketing and Innovation at Animals.

This season on the Animals Podcast. We’re focused entirely on AI content use cases. We’re bringing you on a search to meet the AI pioneers. Those venturing beyond the hype to succeed or fail spectacularly. We’re here today with Tanuka from Daydream. I’m really excited to share this chat with you because Tanuka’s been doing some really innovative work with AI and programmatic SEO.

And he has some pretty futuristic ideas about how content strategy and SEO might look a couple years down the road. But first, a word about Animals. Did you know Animals now offers a podcast service? We’re taking over your audience’s earbuds, reaching [00:01:00] 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. Tanuka, we always start our episodes. With a vital question of what kind of content are you consuming these days?

Thenuka Karunaratne: Uh, I would say outside of what I do at work, I’m reading, uh, one book called Unbroken.

Right now, I don’t know if you’ve heard of it. There’s a, there’s a Netflix movie by Angelina Jolie about it. It’s like Olympic runner that he’s after the World War II and then his plane crashed and he was at sea for like, I think 40 some days [00:02:00] got rescued but in a prison of war camp was there for two years and then came out But then eventually ended up really a really great life and using his life to serve a lot of people Uh, but it’s like his biography.

So that’s what i’ve been reading but very non seo related for sure. Yeah, that’s an

Ty Magnin: epic story It really is. Yeah I’m, really excited to have you here today because you kind of blew my mind when we chatted a few weeks back as you know animals is Always interested in staying at the forefront of what’s happening in the SEO world.

Um, that includes exploring tools and strategies and tactics. You have quite a history in search. You built a few businesses here. Would you mind giving a recap of some of the experience you’ve had that led you to Daydream? Yeah, I would say the most pivotal. So

Thenuka Karunaratne: I think one thing was I was starting in high school.

I was building a lot of. Affiliate type websites that drove most of their traffic with SEO. Uh, so for example, the first one was, uh, a website that was called netflixdnscodes. com, where we would just [00:03:00] summarize what are the working American DNS codes you could connect to, to hop into the American Netflix in Canada, because I’m Canadian, you know, figure out how to do SEO on that for like how to watch American Netflix on whatever device.

American DNS codes, all these search terms. And then eventually started recommending VPN products because they were a better paid option to access the content catalogs in different countries. And then that was my first bet. That’s how I paid for a lot of my college. You know, right before I started Daydream, I was running a lead gen business in the streaming industry.

Another website in the streaming industry called Flix, where we collect the data on what was available across different streaming services, and then use that to answer all the questions of how to watch X. TV show, movie channel, whatever, uh, programmatically. And so the business model there was again, capturing the traffic, giving people the right streaming option, and then collecting an affiliate commission when we would refer those subscribers to HBO, Disney, Hulu, those sorts of companies.

So I think we drove somewhere around like a hundred thousand plus leads to those companies. Uh, from [00:04:00] programmatic SEO.

Ty Magnin: That’s actually a really helpful service. I can’t tell you how many times a friend recommends a show and I’m like, What the hell streaming service is that thing on, right? Like, you go to Google.

Like, everybody’s done that.

Tim Metz: And then how did that lead to Daydream?

Thenuka Karunaratne: Yeah, so with streaming, it’s a very interesting SEO problem because all the search volume is in the long tail. You might have some search volume for terms like best streaming service. That’s live TV streaming service, but the majority of the volume is actually concentrated in the long tail, because you have these queries, like how to watch blank or how to watch the blank on whatever device.

So in that industry, if you actually want to capture the SEO traffic, you, there’s no way for you to write articles one by one to target every how to watch term, that would be crazy. We built a whole programmatic SEO stack where we would collect the data, then generate pages off it. And yeah, basing more data meant more pages that you could use to target different search terms.

And it really worked in that industry because there’s no other way. There was no alternative to [00:05:00] programmatic to actually capture the SEO opportunity in streaming. But then as I was doing that, I started to notice a couple of things. A lot of people would come and ask me how to do what I was doing at Flix for their company, some type of programmatic SEO playbook.

And while I was getting asked that question, I’d often get asked, would you consult with us? But I wasn’t interested. I was doing my own thing. And then when I said no, they would ask me, is there somewhere you can refer us to? And I also didn’t know where I would actually send these customers. The other thing that really started to hit me was a lot of the complicated parts of programmatic SEO, because you first.

identify what is the programmatic SEO playbook, which is an SEO strategy game. And then you have to figure out, okay, how do I get all the data to create all of the programmatic content at scale? What should the structure of those pages be? What types of data points should you include? So to do that usually would require a bunch of skills, like design, engineering, some SEO skills, some content skills.

But my sense was that a lot of companies had a great use case for programmatic [00:06:00] SEO, but the difficulty of actually executing that motion was really difficult. And there wasn’t a great agency that we could think of, and there wasn’t a good tool necessarily that we thought solved the problem adequately.

But when I saw how quickly LMs were getting better, I felt like this was actually the perfect problem to solve because. You know, identifying keywords, long tail search, uh, patterns of scale. That’s a, that’s something that can be automated. Then there’s actually creating all the content scale. That’s also a good use case.

Uh, and then even automating things like the reporting so that that’s not a manual endeavor. That also felt very promising. So I sort of did him on the premise that you could now automate programmatic SEO from end to end. And also that more and more content could be programmatically created where it couldn’t be before.

Ty Magnin: That’s

Tim Metz: awesome.

Ty Magnin: Yeah. I think what’s interesting there is. How you’re bringing together these different disciplines that add up to programmatic SEO because it makes sense like even just looking at editorial efforts. You still need some technical know how you still need some content strategy chops. You need [00:07:00] some editorial chops and you’re bringing that to kind of same thinking to a different search strategy.

So I’m also curious, like the idea that many SaaS companies have programmatic opportunities. Can you share a little bit more in your thinking there?

Thenuka Karunaratne: Yeah. So in general, you would be right to say that programmatic SEO will generally work better for B2C SaaS companies because B2C startups basically by definition target a consumer.

Consumers will tend to do their research through Google, but you can’t say the same thing for a enterprise buyer. So if an enterprise buyer. As a question about what to purchase, they will likely not go to Google to like research that they probably have other places they go to for that information. So SAS, yes, but B2B SAS is an interesting one there.

Tim Metz: How do you see the role of AI also answering some of those queries? Because I find that also interesting. It sounds to me like they’re also the perfect queries where you might actually start going to chat GPT or [00:08:00] perplexity. And then how does that tie into the kind of content that you put out? And do you see like, is it also optimized for those kinds of engines?

Thenuka Karunaratne: Yeah, so when you think about what’s happening in search right now, there are two big changes. So one of them is what you just mentioned, which is more and more search volume is going to AI answer engines. And the second one is that the way you create content is now becoming much more programmatized. So you can basically sum up the two biggest changes for SEO in general and reading those two.

The AI answer engines. Um, we’ll take what Google’s featured snippet did, basically blow that up a lot. A lot of the more factual informational type queries, I think at the top of the funnel will get eaten. But those queries were never where the majority of the value was created in the first place. The most valuable queries from a commercial standpoint were always the middle and bottom of the funnel.

In any query where you actually need to do something, so let’s say you need to integrate SAS X and Y, even if perplexity Or Chachabiki gives you the answer, you still need a [00:09:00] tool to go and do that. So if they’re citing Zapier, you’ll still go to Zapier. But a lot of the queries where you just want an answer to a question, where you don’t need to do something, I think those will get eaten away by AI answer engines, uh, quite significantly.

But the middle and bottom of funnel terms, I think it’s less likely that it will get eaten, but there are some exceptions to that that we can talk about too. And then in terms of how you optimize for AI answer engines, I think that’s a relatively hard problem. But, you know, for example, ChachiBT, uh, uses Bing’s search index.

From a practical standpoint, I think what most people want to know now is where am I even surfacing? Like, what is my visibility in the answers? And most people don’t even have that. So I think the first set of startups in this space, including us, because we just announced this, offering this, uh, last week, will essentially allow you to understand for the types of queries.

In your industry, let’s say it’s something like credit cards, right? It will show you which percentage, like what percentage of queries going through AI answer engines [00:10:00] you surface for. Okay. So now, you know, like what’s queries you’re surfacing for, but then the question of like, how do you optimize that, right?

That should be their O in SEO or AIO. Uh, that is very unclear. Like, I think a lot of the same best principles from the Google era actually still apply. So you want to create at least. I did, like, like, uh, conceptually the, the vision for Google was you want to surface the most relevant, helpful results that are written with the highest quality of information first.

So I think if you’re creating content that hasn’t really changed, if anything, what stayed the same is even now, once you know that, okay, I surfaced here and here, but not here, what are you going to do about it? You will most likely then decide that for the areas you don’t have coverage in, you’re going to create content.

And now you need to create content probably at scale to hit all the different interest points. If you want to have a chance at surfacing it. So there’s a lot that’s similar, a surprising amount, because it ends up in the same place. And then there’s some that’s different, in the sense that understanding how to optimize for AI Answers is probably harder than Google.[00:11:00]

Ty Magnin: It’s like a new black box that we haven’t reverse engineered yet.

Thenuka Karunaratne: Yeah, I think there’s a few teams, including ourselves, looking at that problem, but I think it’ll take some more time to have a better understanding of how you do that. But I think for now, our advice has been, most companies that are doing more in Google will probably do well.

In the AI, uh, chatbots, and vice versa.

Ty Magnin: Yeah. Can you tell us a little bit about how this tool that you just released works? How are you able to get some insights into what’s ranking and showing up in AI overviews? So we’ve only launched it as a pirate

Thenuka Karunaratne: offering for some early customers, like an early beta.

But the way we would do it is you can essentially run thousands of queries through these AI answer engines, see what sources are cited, and then summarize that research. So you could look at the top, whatever number, thousand number of queries in a particular industry. And then based on that, you would be able to say, okay, you are surfacing at an average position of here across a spectrum of queries this big.

And then that gives you a sense of, okay, what is my general [00:12:00] visibility compared to other companies that might be competing for the same terms?

Ty Magnin: Makes sense. Nice. Is there a place people can go check that out or sign up for the pilot or? Where can they find that?

Thenuka Karunaratne: We announced it on our blog. So the Daydream library, if you go to company news, it’ll be there, but we’re going to send up a custom landing page and all that.

Tim Metz: Nice. Cool. I guess maybe look at the broader industry a little bit. Like where do you see AI’s capabilities today versus what most people in content marketing think is possible?

Thenuka Karunaratne: Yeah. So the idea of SD on autopilot. Does not, that’s not a thing that is very good marketing language that people say, but if someone says that, you know, they’re probably not thinking about the right way, because if you put it on autopilot and just, you know, use AI tools, just churn out stuff without any type of customization or scrutinizing from a human standpoint, you’re bound to fail.

So this is where, like, why you shouldn’t just generate an entire blog post of a title, because if you can do that, you are basically generating an article of which there are a thousand copies. And then if Google is [00:13:00] looking at this piece of content and it’s virtually identical to a thousand copies, why would it surface yours?

Totally. So the idea of SEO on autopilot, where you have this super lazy approach to using AI to churn out mass amounts of things, that doesn’t work. The way you should think about it is If you use AI tools well, can you create a piece of content that would match or exceed the quality of what you would expect someone on your team to do?

And then, but doing that is a much more involved exercise. It’s not as simple as, let’s just put in a title. You have to think about what types of data points do you want? Do you have the data? Blah, blah, blah, blah, blah. And then once you’ve actually, from a human standpoint, thought about the user experience, then you can scale it within like a certain sandbox or set of rules.

But you can’t just let things run wild and then expect performance to go well in the earlier days when chat GPT and GPT 3. 5 were announced, there’s some of those cases of companies that scale traffic really quickly and then completely crash for all the hype that exists about AI SEO. There’s very few teams, especially in house teams that have learned how to [00:14:00] use these tools at all.

Like if you go to a lot of startups and things, even, even later stage startups. You find that the process is surprisingly traditional. So I think the uptake on a lot of these tools is slow because most people will try them out, realize that the initial thing that it produces is bad quality, and then they don’t know where to go.

You can invest much more time to actually figure out the nuances of how do you alter all the different factors you need to get it to a high quality and then scale it.

Ty Magnin: Well, how is. AI now leading to the next innovation in programmatic, right? I mean, I think I know the answer here, but if you can say that explicitly, like, where is it filling in the gap or helping advance this kind of deployment?

Thenuka Karunaratne: One is on the content creation side. There’s many types of content that previously you could have only written editorially. But now you can write that content using AI. That wouldn’t have been possible in an era where you just had to template strings. So that’s like version one of programmatic SEO. I think the other parts that are actually really interesting is actually on the strategy piece.

So we’re doing a lot of R and D internally on this right now. But if you think about a lot of the problems on the strategy side of [00:15:00] SEO, it’s figuring out, okay, given that I have this business, this customer, these competitors, what should the playbook be? How do you do that whole strategy piece? Um, much faster and at a higher level of quality with more reliability and consistency.

So there’s a lot of attention is paid to the content creation side. Very little is paid to the strategy side because historically I see people have been very proud that that’s a unique skill set. But I don’t think that would be the case for that much longer. Yeah. Those are the two areas that I would point out.

Ty Magnin: That’s my edge right there. You know, I’m like, okay, I’ve learned to accept, you know, AI’s role in content production. It’s got to be there, right? Like it has to help you accelerate the process. Now bringing in into the strategy part. That’s where I’m like, no, but wait, like, can’t us humans be better at that?

But I think you’re probably right, like it’s gonna be able to impact the research elements of that. I mean, we do a lot of auditing as part of the SEO strategies that we create, right? So we’re sifting through a ton of GA data and [00:16:00] conversion data and search console, etc. Why can’t a machine help analyze some of that data faster than a human can?

I mean, it can, right? It’s just about how do you use it? How do you prompt it?

Thenuka Karunaratne: Yeah. So, I mean, you think about like, how does a human being come up with an SEO strategy? What you realize is SEO is actually an industry where there’s a lot, there’s a huge long tail work agencies. They all have some variation of the keyword research process or strategy process.

Many of them are really suboptimal. And many of them are very time consuming, but what are you doing right? When you’re doing keyword research or coming up with an SEO strategy, you’re essentially looking at, okay, what are the keywords or search patterns based on competitor research? What are the keywords or search patterns based on just more general keyword research?

You have to factor in some of the input on the product roadmap, anything the buyer, you know, the customer has said. And then based on that, you come up with, okay, here are the terms. And then you try and quantify the volume and then you arrange it nicely in a report. That is the perfect type of problem for AI to solve.

So my sense is the strategy side has not been touched yet [00:17:00] for a long time, but that will be the next phase. Cause I think on content creation is very obvious. Now that’s like, everyone knows that I’m really interested in how AI will help automate a lot of the strategic aspects of growth, marketing jobs, even beyond SEO, paid search.

social, other types of channels. How much of this work is actually better done by a machine compared to a person? I think that’s a very interesting question.

Ty Magnin: Sounds core to the daydream thesis.

Thenuka Karunaratne: It is. Yeah. We started out with a core piece of our product, which is very focused on content generation. We used it to essentially serve customers full service.

And we noticed our results are 10, 20 times faster than what an agency can provide in a lot of cases. So that’s great. But the reason to do that is not to scale the world’s biggest agency. It’s to really understand all the Pete, like tiny, tiny, tiny, tiny details. Like how do you think about strategy? How do you think about taking that strategy and then creating the content out of it?

How do you think about then pushing that content, reading Google’s response, [00:18:00] knowing which pages to, uh, no index and save from Google’s call budget. Which ones do you allow to be indexed? Like those details are really subtle. So that’s why I don’t think if you just produce some random SEO, AI SAS tool, and you just throw it out, sure.

You may get some adoption. But until you can understand how you can do something better than what the best people in that industry are doing, I don’t think you’re building anything really disruptive,

Ty Magnin: super interesting. And it sounds like it takes a master to be able to really program AI, to be able to do these things, which is reassuring for us.

Folk that might consider ourselves pursuing some mastery over some kind of content production. There’s still a role for us in the future

Tim Metz: because I’m, I’m very curious about it. Like how do you, who is ideally eventually the person who’s operating that? Like which, which kind of humans do you in the future still see involved?

Let’s say, let’s say you’re two, three years ahead, right? With your vision. Like what, what does it look like? Like what, what are the teams look like? Who’s using this? Like what, where? Where are [00:19:00] humans still in the process? Which parts are done by AI? Like, I’m really curious about that. Like how you see that evolving.

Thenuka Karunaratne: Yeah, I think enough of the complexity of SEO will be abstracted out and taking care of to a point where the person that runs the SEO team is more ahead of growth, but not necessarily ahead of SEO. I think a lot of that functionality will be consolidated into the agent. If the agent is basically doing the majority of the strategy work, the majority of the writing, the majority of the performance monitoring and reporting, the places that you’ll still have, uh, rooms for people on is, especially on content differentiation, that’s a big, big area.

Like, yeah, AI might be good at giving you a pretty good outline or the structure of an article, but if you just only use creation model capabilities to write that piece of content, it’s not going to be differentiated. So you still need a person to think about, okay, if you want to go beyond the baseline of what a model can create, where do we get interesting data sources?

How do we inject expert opinion? I think those things will matter a lot. But I think that consolidates the team a [00:20:00] lot where the majority of the writing would probably be done by AI or the majority of the strategy will be done by AI, the majority of the reporting is done by AI. But you’ll still want.

Human intervention, literally because if that becomes a level playing field, then you still need dimensions to compete against each other on. And I think for Google’s purposes, that’s like net new information added to the internet. So there’s still some human component required to augment what the AI is going to produce.

Otherwise, you get everyone sitting at the same level, right, for the same piece of content. So I think over time, many of the pieces of the process which are just like table stakes will get automated away, but where there’s strategic thinking on how to differentiate against another human that’s also using an agent like that, that’s where like actually the battle will be fought, right?

I think that’s more of what it looks like in the future.

Tim Metz: There are a bunch of humans with these huge Machines fighting against each other, like a few operators on top of, almost thinking of transformers or something with a, with a, with a person on top of it. Yeah. Yeah. Okay. Yeah. That’s super interesting.

Transformers of AI. Yeah.

Ty Magnin: Tanuki, can you talk us through a case [00:21:00] study of yours with Daydream? Ideally, it’s a B2B SaaS customer and that’s who Animals focuses on. That’s who our listeners are. In some kind of like the deployment, sort of end to end, you know, what were the pages you built? How did that perform?

Thenuka Karunaratne: For sure. So one of the customers that we worked with, um, was a company called Twingate. So they’re a zero traffic cybersecurity product. Uh, they came to us and they were interested in SEO, but they didn’t understand what the strategy was. So. You’ve had a lot of companies with that general view, right? I’m interested in SEO.

I know it’s a big channel, but like, what do I do? So the first piece is strategy. Okay. So if their buyer is a head of security or head of it, then we start looking at, okay, what are the long tail search patterns that could attract someone like that? One of the strategies that came up was data breaches. So there’s like thousands of data breaches every year.

Uh, and there’s also vulnerabilities in software. The next is how much search volume is there for all the data breach related terms. And then within those, what are the specific searches, like the top 100, [00:22:00] 200, 300, to have the most volume. And then it was looking at, okay, now we know what the ICP is. We have a thesis on what they could be searching for.

We know it’s in a long tail search pattern that’s compatible with programmatic SEO. And we also know, okay, here are the top 300 maybe that we might write about in each category. Then the process is, okay, well, what should the structure of that content be? Well, if you. Are looking at a data breach. You probably want to understand what was the impact of that data breach.

When did it happen? How did the company react if you were one of the users that could have gotten affected by that? What should you do? So there’s a little bit of thinking on working with that team in terms of looking at the sections of what the structure of that information should be. But now you actually need the data on what actually happened for every data breach.

So you want a vetted list of sources, government sources, reputable press articles that cover that. Then, you know, we designed the template inside Daydream. We added all the data that we needed into the platform as well. And then we were able to generate all these [00:23:00] articles to get 50, 000 visits a month off of that content.

How many articles? If I’m not mistaken, it has been somewhere around a thousand. Forget what the exact number is.

Ty Magnin: Wow. So a thousand different data breaches you cover, getting 50, 000 views. You said a month?

Thenuka Karunaratne: Yeah,

Ty Magnin: the

Thenuka Karunaratne: data breaches and

Ty Magnin: vulnerabilities that’s combined. Okay, got it. Can the machine keep up with the new data breaches or vulnerabilities that it finds and produce new articles like in real time?

What does that part of the program look like?

Thenuka Karunaratne: Yeah, so the identification of, okay, what are these new data breaches? Because as you know, search volume data lags, right? Like lags by about a month. What we decided was that part, we just were vigilant with like monitoring the internet basically to figure out what the new data breaches were a better approach in the future, though, and something that we would hope to build is you can see, okay, I deployed the strategy, but actually 100 new data breaches came out and oh, you may want to create an article about this.

Do you want to create and scale 1 using the same template that would be an optimized workflow, right? That’s not quite there [00:24:00] yet. So we had some human intervention, but that’s how we thought about it. But these are the things where I think also where software is very good is reaction time. So in the Flix example, if a piece of content removes streaming services, a human will never catch that fast enough.

In the same way, a human would not be able to catch all these data breaches fast enough, but software can. So I think this is one area where people think a lot about AI and using it in content on cost and speed, but not on like quality and reaction time necessarily. So this is like one area where I think there’s a lot of promise.

If you could know all the new terms that came out and have surfaced, would I like to write about them? Sure. Right. That’s the type of thing that we’re interested in

Ty Magnin: that twin gate story. I just got to give you kudos. That’s a really awesome use case for programmatic thinking about, yeah, how it aligns with the buyer’s journey, uh, how you’re able to create a high volume of this stuff and seeing that it’s getting traffic.

Do you have any anecdotes just to validate too, that It’s the right traffic, like these people are [00:25:00] converting at some level. Um, you know, there’s been revenue driven through that.

Thenuka Karunaratne: There’s a few things that are a little bit harder just because of the way they attribute conversions, but the metric that they gave us that was the most reliable and attributable to Daydream is that 20 percent of the traffic going to their demo page is all from Daydream content, which should be a pretty strong signal if they’re getting to

Tim Metz: that page, right?

I keep thinking already for a while now, and especially now you’re describing this about like, Is SEO becoming what like data rating has become, where it’s like 90 percent is just kind of computers interacting with each other. And it’s like, you know, jumping in on these kinds of things, like things in milliseconds, and maybe everybody needs to move their servers closer to where Google servers are or whatever.

Like, it’s like, it sounds a bit similar, right? Where it’s like, where, where, where, yeah, where you jump in on timely events, you monitor the internet and then automatically things will happen. Because of that, and it’s just, you know, again, like 90 percent of it is computers operating it and

Ty Magnin: doing things.

Didn’t Gary Vee just put a book out recently that’s like day trading attention?

Thenuka Karunaratne: I [00:26:00] thought that’s funny. I think in this case, the interesting dynamic is that as these tools come out and different people get access to them, it just redefines. The dimensions that you have to compete on. It’s very easy to create content.

If it’s very easy to identify strategies and it’s very easy to spot opportunities. Then a lot of the difference is made in, for example, who has the most useful and interesting data set to enrich those pages that gives them defensibility. So I think you have to start looking at much more, like much deeper ways to get defensibility.

Largely in the form of data, uh, of some sort or assets.

Ty Magnin: Do you think then that product managers get more involved in these kinds of SEO programs over time? Yes. I think you can actually see that change

Thenuka Karunaratne: already. Like if you look at Coinbase. Or, uh, there’s probably a few other examples I can think of coinbase, for example, like SEO is a product function and there’s a big debate right now about should SEO be a marketing or a product function in a lot of more product focused type companies, it’s a product function in Airbnb.

I would assume that it’s also a [00:27:00] product function and Pinterest out of seeing the same thing, mostly because the SEO growth engine is so intertwined with a core product that it really becomes more of a PM job at that point. But then I think SEO as a marketing skill set is more focused on like creating content or blog posts and things like that.

Tim Metz: And do you think the people who currently run SEO programs, like the old school way, you know, a lot by hand. Do you think they have the right skill set to run this or does it require a really different kind of person who thinks in workflows and processes and that there’s a mismatch there? And that’s also why you might see some friction and in adoption.

Thenuka Karunaratne: I can see that basically for most of SEO, the right background for those people should over time skew more and more technical. There’s a certain type of content which is more less SEO focused and more like brand thought leadership focused. Where that, I think, has quite a bit of defensibility, like interviewing a founder and getting his thoughts and publishing it.

That more traditional skill set, I think, has a long, uh, and bright future because there’s no way to generate what’s [00:28:00] inside someone’s head using AI, right? But if it’s mostly just summarizing it, adding some things like that becomes much more automate able. So I think the people that should run those channels should be much more technical.

Tim Metz: It’s like triaging existing information and then infusing your own proprietary things into it to make it unique and then run it at scale. Makes sense.

Ty Magnin: Tanuka, thank you for participating and giving us some of your thoughts, sharing a little bit about the Daydream. Journey with us. Um, really fascinating to see how you’re thinking about AI and leveraging it for customers, you know, in technology.

Uh, today and in the future too. Amazing. Thank you guys so much. Damn. So Tim, this was your first time talking with Tanuka. I had the advantage of spending a little time with him a few weeks ago.

Tim Metz: Yeah.

Ty Magnin: How’d it go? No, it’s cool.

Tim Metz: I mean, he’s, he’s living in a different universe a little bit. Uh, like in both ways, like, like 100 percent SEO.

I’m a bit, I’m a little bit less of an SEO person. Dare I say, he’s living in a daydream. [00:29:00] No, no, I mean, I think, I think a lot of what he says actually makes a lot of sense when I kind of mentioned interviews, like I keep thinking about this finance. I just read that in a book nexus by you. I’m going to mess up his name and he also read sapiens.

He read a really cool book about information. Uh, and mostly about AI actually. Um, and he, he, he cited that example, like 90 percent of foreign exchange is purely computers talking to computers. Like there’s no humans involved and the humans don’t even understand anymore what’s going on. So like, as he was talking, I just kept thinking about that because how he’s describing SEO basically sounds a bit like that.

It’s like computers. Making content for an hour for Google, which is also an algorithm. Right. And so then it’s like, okay,

Ty Magnin: where are the humans? Totally. Right. That’s one thing. One thing I started thinking about is like, it’s helping clients publish high volume of articles. Those are getting seen. However, like as everybody increases the volume of assets that they’re producing, do they all still get seen?

Like who has time, you know, to consume all this stuff?

Tim Metz: I feel a little bit that everybody talking [00:30:00] about SEO, he also assumes that Google will stay kind of static in the way that it finds content. But I was also thinking, what if Google starts to apply Gemini to actually the Google algorithm, right? Like maybe Google will just be able to say, Hey, this article that everybody has missed is actually the best article on this topic, right?

Like What if the way Google, what is it, triages or, or services information completely changes, then also the game completely changes. And I feel people are maybe not factoring that in, that that is also something that might become completely different. It’s hard to

Ty Magnin: factor that in when you really can’t control it, right?

Tim Metz: Yeah.

Ty Magnin: I thought one takeaway for me that was interesting was He basically said what you would do to get more presence in a AI overview or in a LLM’s result. It’s kind of the same thing you would do for search. Right? So it’s sort of like those best practices still apply. Which is nice, you know, they’re sort of quelmed a little bit of my anxiety.

Another takeaway I had today was he [00:31:00] talked about how AI can be used to help produce strategies, especially SEO strategies.

Tim Metz: Yeah.

Ty Magnin: And that’s kind of opened up my aperture a little bit to start thinking about other places that, that yeah, I can serve and maybe the strategic part isn’t as defensible two years from now as it once was, you know, like I said, I was ready to let go of the production part a little bit, uh, and, you know, bring GPT or AI models along for the ride wasn’t as ready to do that with strategy.

And, and yet it sounds like they’re finding some, um, Opportunity there and I don’t see why we couldn’t to and he

Tim Metz: also kind of said that right you still need humans who provide inside and unique things but it’s just. The starting point that you have in terms of information and data and things that you start to build your strategy with is just completely different because it’s not any more human who goes through a CSV file or whatever and does some research.

It’s like the AI is going to give you so much more about like, you know, here about topics, about opportunities, [00:32:00] about what others are doing and whatever, but then based on that, you still need to make a strategy. You can just make a much more. Sophisticated and comprehensive strategy. And you’ll be kind of battling against people who are doing the same.

So it’s going to be a much more like, I don’t know, maybe it’s the difference between like how they did warfare a hundred years ago and how to do it now, right? There’s still generals trying to figure out the strategy, but the information they work with and the tools are just like, you know, on a completely different scale.

And if you keep using the old tools, then yeah, for sure you’re going to lose. But if you figure them out, then you’re just doing a much more sophisticated. Let’s call it competition, not war or battle, but yeah, yeah, no, it

Ty Magnin: is a good metaphor and maybe a

Tim Metz: comparable industry in some way. Well, I was surprised that he in the end still said that he thinks where the humans are needed.

Yeah, I think it’s going to be really important how you structure your processes and how you. How you innovate those and what you ingest in your processes and be really thoughtful about that. And I think that’s also where some of the tension is and what kind of skills and mindset you need to build [00:33:00] workflows.

It’s not in a way it’s also creative, but it’s not like what people normally consider creative work. So I can understand why, why some people who, yeah, are not, are not like happy to jump into that, whereas others might really love it. Right. But it’s a different kind of

Ty Magnin: work.

Tim Metz: It’s

Ty Magnin: sort of like training the team, you know, on your methodology to doing your marketing a certain way.

Yeah. It’s not that different, right?

Tim Metz: No, but I mean, tinkering with a workflow is different from like spending a few hours researching and writing a blog post, right? So if that’s the part you love, but then again, I mean, he was actually saying that he thinks some of the editorial part and some of the insights and creativities, or he didn’t say creativity, but it’s actually still gonna be important.

So yeah, I think that’s cool as well.

Ty Magnin: Yeah, I’m inspired. I’m also a little bit, I’m just curious about what the next couple of conversations are going to be like.

Tim Metz: Last thing that I picked up when he said his view is that most content teams are not there at all, right? So there’s a lot, I think there’s still

Ty Magnin: a lot of opportunity.

Totally. Keep listening. We’ll help you get there. We’re trying to figure it out ourselves too. See you next [00:34:00] time.