SEO FOR PLG MASTERCLASS | Luis Rodriguez, ex Head of Organic Growth @ Miro | #18

00:00:00: Myro is one of the most successful product-led growth stories in SARS history.

00:00:05: Founded in two thousand eleven, they now serve more than a hundred million users across the globe.

00:00:11: To talk about all things PLG, SEO and organic growth, I invited a special guest to the podcast today, Luis Rodriguez.

00:00:20: Luis led organic growth at Myro with over ten years of international SEO experience.

00:00:26: He owned the global SEO and AEO strategy.

00:00:29: created an in-house answer engine optimization framework, built scalable, generative AI workflows for programmatic content creation, and led a cross-functional team spanning content, tech, SEO, and data.

00:00:41: Before Myro, he led growth and SEO at Uber or compared the market, where he built industry-leading SEO tech and became the most visible platform in the UK.

00:00:51: So welcome to the podcast, Lewis.

00:00:54: Hey, Nick, how are you?

00:00:55: Very good to be here with you.

00:00:56: I'm so stoked to speak to you because I think you have a few very interesting POVs on the whole industry.

00:01:03: So before we dive into your whole journey and everything, let's maybe start at the beginning.

00:01:09: So how did you even end up in SEO?

00:01:13: Yeah, that's a really good question.

00:01:14: And I feel that there's not a single path on how you can end up in SEO.

00:01:19: It's at least something that I've seen throughout my career.

00:01:21: But my path started starting selling rocks.

00:01:25: So I was working for a stone company in India called Arbicon International, one of the biggest exporters of granite, marble, these kinds of things.

00:01:34: And the job was actually called online marketing.

00:01:36: But most of the business development that I was doing or online marketing that I was doing need to be done throughout the night, given the time zone difference between the markets that I was handling, which was like Mexico, Colombia.

00:01:50: Spain and I decided not to wake up at two in the morning to work.

00:01:55: So my proposal was to shift into the leads that were coming to the website.

00:02:01: This was still the pre-massive approach to SEO or the awareness of it that will be part of any kind of company.

00:02:08: So it was very much to learn the Google guidelines, webmaster guidelines, reading through forums, making sure to understand how to generate.

00:02:16: money and leads to a company through the internet.

00:02:21: And that's how it started.

00:02:22: I later on moved on to a more professional approach to organic growth in rocket internet and some of the companies that you already mentioned.

00:02:32: But it's been quite nice to be honest.

00:02:35: A lot of the things that got me through SEO was I studied international relations.

00:02:40: And to me, having a little window into what people are searching and what do they want and how do they try to find things was quite a handy thing, particularly when you see it across languages, across borders.

00:02:52: It's something that got me.

00:02:53: So yeah, that was for the beginning of my journey.

00:02:56: So did I get that right?

00:02:57: You started at a stone company in India.

00:03:01: Yes, correct.

00:03:02: How did you

00:03:02: get that job?

00:03:04: Through a company called, it's not a company, it's an organization called ISSEC.

00:03:08: It's a student university type of thing.

00:03:11: And I was doing an internship.

00:03:13: They called it online marketing, but in reality, it was business development.

00:03:17: And it was quite a good strategy where you will get people from the countries, bringing them to India to understand the actual dynamics of the stones and the mines and all this kind of stuff.

00:03:29: And to promote and sell their products.

00:03:32: So it was quite a crash course into a lot of things, business development, online marketing, cold calling even, together with international shipping.

00:03:42: It was a really cool spectrum role to be honest.

00:03:46: And then you found your passion basically around SEO organic growth and you like specialized or niche.

00:03:53: down over time?

00:03:55: I understood that generating money online was a thing and that you could actually generate a lot of money because these deals were millions of euros or it will last for two years type of deals.

00:04:08: So even when I understood that someone was able to spend two million dollars buying rocks online, it was an eye opening for me.

00:04:17: And my passion really came through the international part.

00:04:20: So I've been really lucky being able to work in different countries, different companies and industries.

00:04:26: And it always had that need.

00:04:28: regional or at least international approach where I could leverage my knowledge about how different cultures are and really enjoying learning more than anything through the search lens.

00:04:40: So my passion of SEO organic in general comes through getting to know other people in other places.

00:04:47: The growth part, it really comes from, I guess, that a natural inclination to move forward.

00:04:56: I really enjoy seeing things becoming zero one five ten and it's quite an experience to see that your work is actually having an impact into the bottom line of a company.

00:05:10: My experience took me to companies where thirty percent, forty percent of the revenue was being generated through organics and to me it made a lot of sense.

00:05:20: Yeah, it's an impactful industry, I would say.

00:05:22: And now that we are introducing different methods of search, I think that the awareness, the impact that we can make through understanding organic platforms, it's something that I'm finding rewarding now.

00:05:39: So definitely a good investment back then.

00:05:42: My obvious next question is now What are the key learnings that help you from the stone company in India for your role at Uber?

00:05:52: That's a really good question.

00:05:53: You know, scrappiness.

00:05:55: I think that one of the biggest things is to do the best with the tools that you have.

00:06:00: SEO, particularly for a company like Uber, was just not needed.

00:06:04: The company for the very first five, six years, it grew through international expansion, opening new markets, and ride sharing was not a thing.

00:06:12: It was not something that people would be searching for.

00:06:14: When more businesses open like Uber Eats, that's when we started going into the more massive markets, really speaking to the daily people and hitting those really high search volumes because everybody wants a burger delivered across the world in all languages, in all cities pretty much.

00:06:30: So when I joined Uber, the SEO or at least the organic development was not necessarily at the same size as other more sophisticated parts of the company.

00:06:41: So we need to invest a lot and do and be handy.

00:06:44: We need to pull the different strings that we had on our hands.

00:06:48: So I think that being scrappy was one of the things.

00:06:51: The second one, being independent, I feel that as an SEO sometimes we get particularly an enterprise level SEO when you're working large companies.

00:07:01: We tend to overcomplicate things where in reality technology, particularly today, allows us to fast-track a lot of the analysis, fast-track a lot of the research phases.

00:07:11: and prove really the points about the channel itself.

00:07:18: So those are the two things.

00:07:20: Probably the last one, I lost the fear to call calling.

00:07:23: So at least to me, it was something like, I put myself into the customer shoes quite often.

00:07:30: And in the Stone Company, it's like some random guy from Mexico is calling me to sell me rocks from India.

00:07:37: it's not necessarily a call that you would say, oh, I really trust it.

00:07:41: So, you know, losing fear and just going for it, asking for the things that you want, it's something that in my career I do were really helping out.

00:07:48: The company is a fantastic place to develop, to learn and to try new things.

00:07:54: And being able to raise your voice, to raise your hand and say, I want it, I want to try something new was something extremely valuable in the company.

00:08:03: So the first thing I'm noting down now for me is if you're hiring SEO people, look for people that have at least done cold calling once.

00:08:12: So they're comfortable embracing the unknown and like growing beyond your comfort zone.

00:08:18: Definitely.

00:08:19: And I think that having also a background of.

00:08:22: something client facing, something where you actually had to sit down with a client or a customer to do that.

00:08:28: In my case, I worked at a front desk in a hotel when I was in the university.

00:08:33: And that really helps you out to understand that, yes, they are searching, but it's also a person, right?

00:08:40: And that person has a persona, has biases, has a profile, and understanding that there's someone behind all of these searches can help you out to make the connection to perhaps better as your strategies, where you move away from what people are searching for and just specify the keyword to really understanding what they mean.

00:09:00: What do they actually want?

00:09:01: What is what is motivating this kind of search?

00:09:04: So if I was to hire, when I hire, I definitely appreciate people that have client-facing experience in any kind of level, customer success, even sales is something that I feel it helps other a lot.

00:09:20: Now, that doesn't discard anyone because one of the best SEOs in the industry that I've had the chance to work with, he was a former policeman and he was a police for like, twenty-five years.

00:09:32: High-fives to Mike down there in Australia.

00:09:35: So yeah, I feel that even thinking about it, a policeman probably has a lot of contact with people and a lot of client-facing situations.

00:09:45: So being with people helps.

00:09:48: Awesome.

00:09:48: I think there's also another uh seo from the uk that has been a policeman.

00:09:54: i

00:09:55: can't remember his name right now but um he occasionally writes stuff on linkedin.

00:10:01: so uh now your last ten year was obviously at myro and i also said in the introduction that myro is like this crazy plg.

00:10:13: uh sorry so for everybody that's not familiar with the acronym like product that growth.

00:10:18: um I think there are a lot of cool SEO success stories around there, and some have the PLG approach, some have a more classic demo-lat BtoB SaaS approach, and BtoB SaaS has obviously also been a big, or they used SEO as a key growth driver from times.

00:10:39: From your experience and also from just your point of view, do you feel like Having a product-led growth model helps more with SEO than a classic demo-led business model?

00:10:53: I think that the biggest advantage of PLG is that you get a lot of feedback and you are able to touch into the different parts of the funnel and the different feedback loops that you get.

00:11:04: I think that demo-led tends to bring people to marketing.

00:11:09: into marketing materials more than actually to the product itself.

00:11:13: One of the biggest learnings that I had at Miro was the usage of templates, for example, where these templates are not necessarily just an entry point for someone to sign up, but it's actually a collaboration point, something that allows you to actually bring more people in, not necessarily just that one visit transforming to one conversion.

00:11:32: So I think that PLG and the focus in SEO in this part, opens up a lot of opportunities.

00:11:40: I believe that SEO has historically been almost cornered into that position.

00:11:47: But if we really think about it with PLG type of companies, you can really have a lot of retention through that.

00:11:54: Once again, the template is a really good example where A person can start their journey with Miro to a template, but it can also restart it, can also continue and bring more people and bring that priority to the focus.

00:12:07: I think another good example of that is Canva.

00:12:10: Canva definitely is one of the companies that has opened up that idea to bring the people within the product, not necessarily at the top.

00:12:20: So to me, really, PLG is the dream of an SEO team.

00:12:26: where you're able to have complete, pre-sixty ideas that continually bring value and bring customers to the product, not just this is.

00:12:38: Have you read the book, product led SEO from Eli Schwartz?

00:12:42: Yeah, definitely.

00:12:44: I like Eli's idea.

00:12:45: He's a great guy, also high-five to him, and generally think that it opened up the eye, or at least it made it a name.

00:12:54: I think that sometimes naming things and establish them within the industry can lead to more adoption.

00:13:00: I also think that it's something that is easily misused or misunderstood, to be honest.

00:13:07: I also like the thoughts in there and the idea of just trying to make your product the obvious answer for a user's query.

00:13:18: And I feel like... There are so many cool SEO stories out there.

00:13:21: So my role, Canva, you also named them.

00:13:24: I'm always like astonished by Canva when I look them up into Simrush, then they have like a hundred of a hundred authority score, which is like basically just Google or YouTube has that.

00:13:35: But I also liked, for example, the story of Rippling, where like they're this HR payroll, they have a lot of things, obviously management software in the US, and they had this approach of just constantly launching new products within their range and expanding.

00:13:55: And I think they call it the compound startup.

00:13:58: And I feel like it makes so much sense because from an SEO perspective, you always get a completely new set of clusters and intents and user groups you can target.

00:14:11: So how do you feel about expanding your product?

00:14:17: in a way that you also open up new demand through people that are interested in other things that you haven't tapped into before.

00:14:25: Yeah, I completely agree with you.

00:14:28: Going back to Eli's concept, I think that this means to let the product tell you where the demand is.

00:14:33: Because when a company focuses into creating a feature or creating a specific product, there's a lot of research behind it.

00:14:39: We are not creating things because nobody wants them or because nobody has the intention to use them.

00:14:44: So I think that being able to be in that journey very early into the stages of planning and strategy, making sure that we build products that are also built for acquisition.

00:14:56: I feel that as the industry evolves and as there are more platforms being open, SEO or organic need to become part of the PLG process and almost like fade away within it, where the product by the inception itself is the one that it's creating the demand, but also answering the demand.

00:15:15: I think that at least that mirror the way that I'd like to implement.

00:15:19: that was really understanding beyond, does it rank, beyond, is it visible or be inside it?

00:15:27: It's more about, right, if somebody wants to brainstorm and finds our template, Do they bring more people to us?

00:15:35: Do they actually use it?

00:15:36: Do they end up paying for a subscription or increase the number of boards that they're creating, ranking or being disabled?

00:15:44: I think that nowadays is just not enough.

00:15:48: And the real way to do this is to be quite embedded into the product journey, I think.

00:15:55: I like your idea about the templates and I think templates is a great example for PLG where like product led SEO, you mentioned Miro, you mentioned Canva.

00:16:07: Now I see an increasing amount of companies that also try to embrace this idea of like free mini products or like parts of their product that they offer for free.

00:16:20: So for example, I looked at Eleven Labs, it's this voice AI company and they have a big portfolio of voice to text.

00:16:30: Exactly.

00:16:30: And a lot of small mini features from their big suite that you can immediately experience as a user and get to know the product.

00:16:39: So how do you like this approach and what's your overall take on these mini products, mini tools?

00:16:45: I love it, to be honest.

00:16:47: One of my favorite things when I was a kid was to go to the supermarket and have little samples of things because it really puts you in a position to have to buy them.

00:16:54: So to me, it's a bit of a digital translation of that because my approach at least is whenever a product or a feature is being launched, it must have three things.

00:17:05: To me, it has to have a really good detailed explanation about what this is.

00:17:09: Sometimes I feel that particularly PLG companies tend to fall too much into the trap of being quite inventive or being quite creative with their namings and the way to position the product.

00:17:20: that doesn't necessarily reflect search demand or how people want to find these things.

00:17:26: So having a good explanation is number one to me.

00:17:29: The second one is what can it do for you?

00:17:32: Sometimes I feel that it's not quite obvious what a certain feature a product can do for the daily person or for the decision maker.

00:17:40: One very good example, using Miro could be just a cool thing that you do and put little stickers with a lot of flashing lights.

00:17:48: But in reality, it helps you out to shorten all of your path between idea production and to actually bring to market those ideas.

00:17:55: So demonstrating to decision makers or to the everyday person that what is the value of that, that you can spend fifty percent less time ideating.

00:18:04: and bring into your markets to product is quite important.

00:18:08: The last one is a way to use it.

00:18:10: So the templates is, we're going to go back to that example a little bit too much, but I think that having a way to use these products, show you that the art of the possible needs to happen because sometimes if you're searching for something, it's because you do not know it, right?

00:18:25: And sometimes the answer cannot be that obvious.

00:18:27: And if you give me a way to use it, it's quite good.

00:18:31: Another really good example, it's pre-built websites that WordPress started to use a gazillion years ago.

00:18:39: That is directly PLG.

00:18:41: You have a pre-built website that you just basically need to adapt a little bit.

00:18:45: And you're right on in the internet pretty much using the WordPress platform.

00:18:50: So to me, having that trifecta is probably the best way to approach it.

00:18:56: Let's stick into the point you made about PLG companies being very innovative and maybe with naming, etc.

00:19:04: Maybe not taking into account the demand or how users actually name the things a little bit.

00:19:11: I feel like for SEOs or people generally in organic growth, it's always important to win a seat at the table, to have their perspective heard when, for example, also... product strategy is shaped because we bring this important view for the customer, for the demand and also a little bit as a feedback loop from the market.

00:19:36: What's your advice for people that maybe are still figuring out how to get the seat at the table?

00:19:44: How can you justify it?

00:19:47: Yeah, that's a really good question and definitely one of the best experiences at Mira was working together with the PMM team to precisely marry positioning with search intent.

00:19:58: And one of the things that helped me out to achieve this was the power of friendship.

00:20:04: I believe that creating an actual personal relationship with the person that is building these things and understanding what are their problems, how do they approach things, it's a lot more important.

00:20:14: I think that SEO can be a really good partner to them by feeding information.

00:20:19: the same way that PMM teams will do user research, will do whole groups.

00:20:24: SEO, it's also a reflection of the market and how can people start finding these things.

00:20:29: That's one way, the power of friendship.

00:20:32: The second one, I'm alluring a little bit to it, so the data ultimately numbers reflect reality and being able to demonstrate what is the capacity of acquisition.

00:20:42: or retention that these products can have, it's something invaluable for any kind of planning or strategy.

00:20:48: If you know and what to expect about how many actual people are actually trying to find this and use it today, it can lead you to very different decisions on feature prioritization, on launching times as well.

00:21:00: It can be that you're building something that is really good for Christmas, you know, and you just put it up for Christmas.

00:21:06: So those two things held me out.

00:21:09: The last one, I think that creating a good way to showcase the impact of SEO, it's always a big win.

00:21:19: Everybody wants to be with the winners, right?

00:21:20: And if the strategy itself and the steering of the naming of products and the positioning of products does lead to more acquisition and to more adoption of those products, I think that making that visible is extremely important.

00:21:35: The worst thing that we can do is to do a great job making a product very discoverable and being able to be adopted, but not tell anyone.

00:21:46: In enterprise levels, I think that it's a common pitfall for SEO teams, where we struggle to translate jargon and specific metrics into business impact, something that every team can understand, at least within my approach of, within Uber Eats, for example.

00:22:07: that translated to orders, to basket size, to number of restaurants being signed up, and that's something that the company across would understand.

00:22:15: If I would speak about impressions click through raids, even if for us sounds like the most basic of things, a sales department will probably won't have visibility on that, you know, and it's important that everybody understands the value.

00:22:31: You already mentioned a couple of really great points and I want to maybe summarize it or like get this point across again for the audience because data and looking at the right data and measuring success I think is more critical than ever.

00:22:48: because we basically had this introduction of AI overviews and all the different like new things in the ecosystem where suddenly a lot of people learned that Just looking at traffic might not be enough.

00:23:03: For some, it was no surprise at all.

00:23:05: But for some, it actually exposed some outdated strategies that weren't supposed to work for quite some time.

00:23:13: But they, I don't know, fortunately or unfortunately, still worked.

00:23:18: So if you would have to define a set of data points or metrics to really understand if you're moving into the right direction with your organic growth strategy.

00:23:32: What would you look at?

00:23:33: Yeah, that's a great question.

00:23:35: And before you get into it, I'm not sure if traffic has ever been an enough metric, to be honest, because it almost sounds like vanity.

00:23:46: And it is great that a lot of people are seeing your things.

00:23:49: But if you think about SaaS, or generally any product outside of advertising, outside of editorial, I think that it's all about the bottom line, what generates value for the company itself, if that's revenue, if that's subscriptions, if that's following.

00:24:06: I think that right now what you're facing is just an extreme amount of noise and confusion because I've seen a lot of approaches about reverting to impressions, for example, or river or now I see a lot of noise about as well.

00:24:23: of counting the number of times that the people agent of ChatGPT thinks your website.

00:24:30: Cool.

00:24:31: And so what?

00:24:33: Ultimately, what we are here to do is to provide growth and to provide business value.

00:24:38: So my number one recommendation and where I try to point at everyone, it's really don't get into the noise.

00:24:48: Don't drink the Kool-Aid.

00:24:49: focus into what generates value for your company and continue going.

00:24:52: Now, being pretty specific, I really like to separate two things.

00:24:56: One, those are the business metrics.

00:24:58: How can we communicate to stakeholders, decision makers?

00:25:01: What is the valuable part of the company?

00:25:03: And again, as I was saying most of the time, that revenue and that's value added through the overall business.

00:25:11: The second part is working metrics.

00:25:12: Like what helps us out as you were saying to understand if are we moving into the right direction right now?

00:25:18: To me, I think that the most sophisticated part is the visibility that you can have within any category.

00:25:24: Right now, I've been working a lot with a tool called EverTune that I really like because it flips the story from chasing prompts, which prompts is just a very long-tailed keyword for all folks like me.

00:25:36: Long-tailed keywords and prompts are just the same.

00:25:41: But EverTune flips the story.

00:25:43: It's about the visibility of your brand, regardless of the actual prompts that people are using.

00:25:48: And to me, showing up when the user needs it, that is a really valuable thing.

00:25:53: So visibility comparing to a market full size into the market itself, that helps our luck.

00:26:01: The second part, I do think that people need to come to the website to do an action.

00:26:06: So I'm not saying that sessions is something that is particularly enough, but it is a part that we need to measure.

00:26:13: People ultimately cannot create a template or cannot.

00:26:17: Can people order burgers in ChatsBT already?

00:26:22: I think Uber has not yet.

00:26:24: I don't

00:26:27: know.

00:26:28: Maybe they have already integrated with the Agente Commerce Protocol.

00:26:32: Maybe, I'm not sure.

00:26:33: Yes, I'll ask around.

00:26:38: At least right now, for most of the products, you still need to go to the product itself to do something.

00:26:44: You can find it within the AI tools or the AI overviews, whatever it is, the platform that you're using or the preferred format of your search, you'll ultimately have to go to the product.

00:26:55: So working with sessions as well, it's really important.

00:27:00: The last one, to me, retention now becomes a bigger part of what an organic success looks like.

00:27:06: So making sure that we are acquiring repeatable, that the returning users, it's something that I really like to keep an eye on.

00:27:14: Let's talk more about burgers real quick.

00:27:17: The burgers now being an analogy, but I also put some thought into that and I'd like to get your perspective.

00:27:25: Do you see people anytime soon really like making a whole purchase decision?

00:27:34: And I'm not talking about the purchase of like five euros or something, but let's say maybe a fifty euro plus or maybe a hundred euro plus.

00:27:43: Do you see them making the whole purchase decision and actually completing the purchase in ChatGPT?

00:27:50: Of course, yeah, I think so.

00:27:52: One of the most difficult things is to jump out of platforms and go to another one.

00:27:56: And right now, that means going from ChatGPT to Chrome to a browser or to an actual app.

00:28:03: So I believe that once those use cases are enabled, they will be a massive thing.

00:28:08: to give you a couple of examples.

00:28:10: So right now, I think that some e-commerce has been enabled.

00:28:14: You can not necessarily complete the purchase.

00:28:17: I think that a deal with PayPal was just announced, but it's particularly enabled all of those transactions within the platform.

00:28:24: So we are definitely going there.

00:28:26: Is it a good experience right now?

00:28:27: I don't think so.

00:28:28: I think that chat GPT or generally all of our AI platforms are generalists right now.

00:28:35: They have just any... feature or any kind of use case fit into it.

00:28:41: And if I was one of them, I will definitely start searching more specific use cases and start tackling them one by one and or creating tools for people to solve those use cases.

00:28:52: So e-commerce is definitely gonna be a good one.

00:28:55: I'm definitely looking forward to see what Amazon would think, what all of the e-commerce platforms would think eBay as well.

00:29:03: once you are able to do the full-on research with high detail, with actual information points about everything.

00:29:11: So going back to our burgers, if I was searching for the best burger in London right now and I could order it through ChatGPT, I would one hundred percent do it.

00:29:20: Why do I need to now open my UberEAT account or my Deliveroo account to do so?

00:29:25: So I don't like to see creativity or these platforms as something That has never happened before.

00:29:36: the technology definitely has not happened before or perhaps very little times But it's a new platform and it has its own rules and it's gonna have its own evolution.

00:29:45: You were not able to buy things to Facebook before and for the for now the very little people that use Facebook.

00:29:51: You can still buy things for Facebook Instagram purchases tiktok purchases.

00:29:56: So it's all pointing at the same thing.

00:30:00: these platforms need to generate revenue one way or the other.

00:30:03: And I feel that chat GBT has opened up a lot of business use cases, a lot of enterprise level type of monetization.

00:30:13: Now it's gearing up to monetize from the everyday folk.

00:30:18: Okay, I'm playing a little devil's advocate here because I think that a couple of months ago, Meet meta actually announced yet wasn't July.

00:30:34: So we're recording this in October, so not too long ago.

00:30:39: They announced that they will end their native checkout for shops.

00:30:44: So for Facebook and Instagram, because like they said, they will no longer support order management, post purchase experience, blah, blah, blah.

00:30:54: And you have to go.

00:30:56: To the sellers website.

00:30:58: So as I said playing devil's advocate here.

00:31:00: Oh, yeah.

00:31:01: No good point.

00:31:02: Why do you think that it will be different and a different story with chat GPT and that we won't read the same announcement?

00:31:11: So like a gente commerce blah blah blah being discontinued in about five years?

00:31:18: Yeah, well good good points and I was not aware of that.

00:31:21: nevertheless Facebook.

00:31:23: It's a dying platform I'm sorry to say that.

00:31:26: It's definitely not growing in users or use cases as well.

00:31:29: The audience that Facebook tends to have at the moment is not necessarily the one that was raised with e-commerce embedded, right?

00:31:37: So I think that now being able to manage all of that purchase journey for younger audiences, it's a lot more important.

00:31:44: So that would be one of my arguments.

00:31:47: The other one is mostly the necessity of to create revenue true.

00:31:52: AI platforms.

00:31:53: Even if it doesn't last, it's something that they will definitely try.

00:31:58: The last point is I think that social media is an incomplete journey.

00:32:02: You need to make an entire purchase decision based on five seconds of a video into three kind of posse words in a bunch of fake likes and fake shares.

00:32:12: Never the best.

00:32:12: I think that a journey that is enabled within AI platforms is at the research phase.

00:32:19: You can have that all very well compressed in maybe those same five seconds, but at least with a lot more reliable information and with different sources.

00:32:29: So there's, to me specifically, there's a higher level of trust of the things that I would buy through SiteGBT.

00:32:37: I don't think that I've ever gotten more to your point.

00:32:40: I don't think that I've ever bought anything through Facebook.

00:32:43: or any social media.

00:32:45: But that may be me and a lot of other people.

00:32:48: Yeah,

00:32:49: quickly stepping out of my devil's advocate role here, totally plus one, especially to your last argument.

00:32:56: I also feel like the experience is completely different because you can ask.

00:32:59: chat.gbt and also maybe other AI systems.

00:33:02: You can ask follow-up questions.

00:33:04: You can ask to pull reviews and to compare with other products.

00:33:07: It was never possible with Meta and with Instagram.

00:33:11: Now, stepping back into my devil's advocate role here, I started to see or hear people being fearful about ChatGPT losing its magic in the experience and getting biased towards this commercial impact.

00:33:30: Like as you said, the necessity to drive revenues.

00:33:33: So there has been a report.

00:33:35: I don't know.

00:33:35: I didn't check it.

00:33:36: So for everybody that's hearing this, please go and check it yourself.

00:33:40: That OpenAI will probably lose a half a trillion dollars by twenty twenty nine due to their investment into data centers, blah, blah, blah.

00:33:50: So obviously they have to make more revenue.

00:33:53: But do you also fear that the user experience will become worse due to this commercial necessity for OpenAI?

00:34:08: Yes, clearly.

00:34:09: I think that we need to look back into the previous chat GPT Google.

00:34:15: Ten years ago, the Google server was very clean.

00:34:18: It was really wide.

00:34:19: and you know you would see a lot of empty space pretty much and you would get your answer, your blue link and it would arguably give you better answers than nowadays.

00:34:30: So if we look at how the SERP evolved to more specific use cases traveled.

00:34:36: AI modes, images, videos, rich snippets, all this sort of rich snippets of the world.

00:34:42: Now we are seeing that it's a crowded space with a lot of more components, with a lot of different interactions trying to satisfy specific use cases.

00:34:51: So I don't particularly see a way out.

00:34:54: Arguably the open AI folks are really smart people and probably will find a way to create less friction.

00:35:00: I think to me the key point would be to... generate that revenue with a less amount of friction.

00:35:06: We're using an example.

00:35:08: I've been looking into cars lately.

00:35:11: And in the use case, in the use cars market, it's obviously quite difficult to know if, you know, at a brand X versus another one, it's how good that they are, what kind of common problems do they have.

00:35:25: If I was able to actually book a drive test, I would have done it in the last couple of days.

00:35:32: So how to get you to that point to trigger, to actually click and generate the revenue, it's going to be quite interesting.

00:35:42: Arguably, also OpenAI is making a lot of different revenue streams from the API usage, from the different custom GPTs that you can do, premium plans.

00:35:55: There's definitely a lot of it to your point.

00:36:00: There's trillions of dollars to be generated.

00:36:03: Whenever these companies are going to be taken into the market, it's going to start with a bar this big, like literally off the charts and off the camera.

00:36:11: So that's going to generate a humongous amount of pressure to start driving and milking as much as they can.

00:36:20: Right now, the adoption has been fantastic.

00:36:24: Can we still count it as a new kid in the block?

00:36:26: A little

00:36:27: bit, probably, yes.

00:36:28: the newest kid would call it in the block.

00:36:31: And there's going to be a critical point in which acquisition or at least new discovery is going to end.

00:36:41: There's a very good, I recommend everyone to watch the Kano model of product where basically the wow factor starts dying out over time because people assume it as one of the basic things.

00:36:54: In this Kano model, there's a very good example about Wi-Fi.

00:36:58: like how Wi-Fi.

00:36:59: back in the day, fifteen years ago, it was a plus that you would see a hotel with Wi-Fi like right on.

00:37:05: Right now, if a hotel doesn't have Wi-Fi, you will probably question why are you there.

00:37:09: So, that inevitably happens to old products and the innovation part in keeping things fresh, it's going to be a challenge for open AI or for any AI platform.

00:37:22: great example with Wi-Fi.

00:37:23: actually I had this morning.

00:37:24: I was stepping out of the subway and I was calling my girlfriend and I looked at this historic building here like in Berlin Germany in the city center we have a lot of historic buildings and I was feeling this excitement about wow how crazy it is that I can like call her now and like I was going through the subway.

00:37:42: so but it's it's honestly a very very rare moment to still be amazed.

00:37:49: Yeah,

00:37:52: correct.

00:37:53: And I think that OpenAI or chat specifically brought that wow factor to the internet itself, right?

00:38:00: Like how complex can it be to put all of this tool together?

00:38:05: It was quite amazing, particularly for all, when I was at Uri, we started to play with the Lee one with the Lee two.

00:38:12: We started playing out with GPT, like high-five tool of the data science team there in the Uber team because we were really seeing how this thing was evolving.

00:38:22: One of the very first projects that you got the chance to use this technology was to generate about five million restaurant descriptions in multiple languages.

00:38:32: Surprisingly, we didn't use GPT.

00:38:34: GPT was bad at that time.

00:38:38: seven years ago, six years ago, where this thing was still called machine learning or NLP.

00:38:44: And we use Alexa system actually to reverse engineer the tax that certain restaurants or dishes had.

00:38:55: So we can generate descriptions based on that, but there was an LLM behind Alexa pretty much.

00:39:01: So it's been a lot of fun to see the internet.

00:39:07: to be cool again.

00:39:09: That's a very good bridge you built with the Uber Eats descriptions because I just wanted to say that we already tapped into AI and we spoke about AI like as a channel marketing channel distribution.

00:39:21: But I think like we marketers especially we also use AI a lot.

00:39:25: as a tool.

00:39:26: And as I said in the introduction at MIRU, you also experimented with generative AI workflows for programmatic content creation.

00:39:35: And I just like to understand your thoughts more about how you approach this, because I see a lot of people who also not necessarily always with the best intentions say, yeah, I can just use AI to create hundreds and hundreds and thousands and thousands of pages.

00:39:54: And yeah, I would just like to hear from someone with a lot of experience.

00:40:00: How did you approach it?

00:40:01: How would you approach it?

00:40:02: Like, what are your key learnings?

00:40:04: Yeah, that's a really good question.

00:40:07: I do definitely think that it's so magical, the experience of generating content and words with AI, that it almost oversimplifies the task.

00:40:17: And I think that's a very big, common pitfall.

00:40:19: Number one learning to not over rely on AI right now.

00:40:22: to me that the sweet spot is really on hybrid and assisted models Where you still maintain a human touch point or multiple human touch point depending on how delicate the content is.

00:40:33: To really accelerate.

00:40:35: I don't think that the goal right now is to automate.

00:40:37: I think that right now the most productive is to accelerate delivery and Again, perhaps in the mirror spirit to bring ideas to market a lot faster.

00:40:46: A lot of the different tasks have become meaningless or have become not time consuming.

00:40:52: You will research, clustering, analyzing large data sets, making insights, or drawing different conclusions from multiple sources.

00:41:00: All of that is just a tiny little... prompt nowadays if you have your train models.

00:41:06: So to me, one of the biggest learnings is to invest into preparation.

00:41:11: So creating your custom GBTs, creating your own projects within cloud or whatever is the platform that you're using that can help you out with those repeatable jobs is the best thing that we can do.

00:41:22: So for me, step number one, searching to everything that you do until death, like something that you hate, copy pasting, growing things, downloading, emerging data sets.

00:41:32: all of that can go into an automation part.

00:41:36: The second point and one of the biggest learnings is that a human is still needed very much because now the role perhaps of content copywriter or content manager needs to evolve into content strategy.

00:41:52: At least that's my approach and my concept.

00:41:58: where still a person needs to understand the customer.

00:42:02: still a person needs to understand what good looks like.

00:42:06: Any AI platform right now is not able to discern between good quality and bad quality, at least within the content side of things.

00:42:14: So having someone, an expert that is able to make that adjustment based on their experience is to me, invaluable.

00:42:21: A great example of this was a compare the market where all of the content that is generated there has to be compliance to a lot of government rules.

00:42:31: And you cannot use multiple words that are very common within AI platforms like best, or this is what it's for you, very indicative, very determination.

00:42:42: So being able to have these intervention points is quite interesting.

00:42:46: The biggest challenge that we had there was to integrate compliance, like an actual government compliance touch point, to understand if the quality was good enough or not.

00:42:56: So again, a hybrid model to me is one of the biggest things we can do.

00:43:04: I would definitely advise everyone to consider what are the repeatable tasks that you can automate and do not lose your expertise, do not lose the control of what you want to tell to your customers.

00:43:18: You can definitely bring all of the customer feedback, customer success stuff into one melting pot, but ultimately, your expertise is still very much needed.

00:43:29: What good looks like needs to be defined by someone that has the highest level of quality.

00:43:34: And right now, it's a person still doing.

00:43:37: And if Uber would approach you now again and would say, hey, Lewis, we have to create another five million restaurant or whatever descriptions.

00:43:51: Like, I mean, five million descriptions is very hard.

00:43:56: to have the human in the loop, right?

00:43:58: So do you see a scenario where you could actually, I mean, it's super convincing, the content manager to content strategist argument, but do you still, do you see a setup where it's possible to hand over like such a content project to AI, or would you rather argue, maybe guys, it doesn't make a lot of sense to actually create these five million descriptions?

00:44:24: It's a great question and probably a philosophical one Because I think that you need to determine what good looks like as I was saying and what good enough looks like.

00:44:34: I think that even QA can be assisted by automation Pretty much like you can actually create an antagonist model that goes in searches for the mistake that you have been identified.

00:44:46: In this experience that I was sharing we did spend about Two months or three months just really improved restaurant descriptions Jesus Christ, it was a lot of reading, pretty much.

00:44:58: And we bumped into multiple things, a lot of hallucinations, a lot of things.

00:45:02: So eventually we learned what to search for in mistakes, because these models, if they make mistakes, they tend to do the same mistakes repeatedly.

00:45:11: So we were able to capture.

00:45:13: So if your content is not sensitive, if for you and your audiences and customers, eighty percent is good enough, I would say, fuck it, let's go.

00:45:24: Let's let's just get it done in the case of Uber Eats.

00:45:27: I would definitely guys let's go and and now let's.

00:45:30: we have the ability to do it even better because Creating data pipelines to AI models now.

00:45:35: It's a game before.

00:45:37: in that first approach We needed to use some very big computers to dump all of these data.

00:45:43: We needed to actually tap directly into servers to be able to get the numbers from it.

00:45:49: Right now, an API between different things in charge of it is the easiest thing to build.

00:45:54: So I think that we could even do it better.

00:45:58: What I let an AI completely run and publish without any discrimination.

00:46:03: No, I will be very smart and use industrial production techniques of quality assurance.

00:46:10: Those things have existed for, I don't know, two hundred years, three hundred years, and they're true.

00:46:16: That's how you get your toothpaste good enough without everybody checking age of the toothpaste that are being made, right?

00:46:22: It's based on a sample.

00:46:24: Your sample is statistical significance.

00:46:26: And let's go.

00:46:28: That makes a lot of sense also with spot testing.

00:46:30: Like for every hundred piece, you take one out or like every thousand pieces, you take one out and check it.

00:46:38: And if it's going well, that makes a lot of sense.

00:46:41: Yeah.

00:46:41: If you have, if you have used chat empathy enough, you know that there are patterns or at least I can see patterns.

00:46:47: There are definitely repeatable behaviors that you see.

00:46:49: So struggling with a specific language, struggling with access to specific type of files or to word certain things.

00:46:59: Once you read one million restaurant descriptions, you start seeing the pattern, or at least you have to.

00:47:04: So yeah, it's something that can be done.

00:47:06: And again, I see it as an industrial production side.

00:47:12: Very simple question.

00:47:14: Do you have any preference for chat GPT, cloud, like other providers?

00:47:22: and especially the models around certain SEO tasks.

00:47:25: For example, content creation or maybe also analysis of large data sets, interpretation, any sort of task.

00:47:34: Good question.

00:47:34: For content projects, I definitely rely on Claude.

00:47:37: I think that the way that you can create projects and a lot of the different fine tuning that you are able to do their plug and play, it's quite good, particularly because it enables marketing teams to do it by themselves.

00:47:50: two years ago when I was in Comparter Market, we needed an actual data scientist to help us out fine-tuning and put those three things.

00:47:58: Now it's just there.

00:47:59: It was kind of like tidbits a little bit.

00:48:01: So I like Claude for that.

00:48:03: I like Gemini to strategize, to create different understanding and to summarize, to compress different things and come up with nice... narratives, I think that is quite good.

00:48:14: ChatGPT to me is the generalist.

00:48:16: It's the one that I go whenever I have random questions or random answers.

00:48:21: I feel that, particularly looking at a better Gemini, the index of information is definitely larger within Gemini because they have Google, right?

00:48:31: But as the models evolve, there's sophistication.

00:48:35: ChatGPT to me, it's a yes machine, so you need to retell it.

00:48:42: please don't just buy a confirmed bias to me because I've seen myself just saying yes so many times.

00:48:48: it was like I have a big question in the beginning and then every suggestion is so good so proactive that it's a common thing to just yes yeah okay to it yeah and to me that deviates from actually a good strategy or a good outcome if you just end up saying yes to the machine.

00:49:08: Yeah, right now, those two things for measurement more in specific.

00:49:11: As I was saying, I really like Evertune approach.

00:49:15: Very few tools out there are trying to give visibility on brands.

00:49:22: It's a lot more about giving visibility to the prompt itself.

00:49:25: And I see the point.

00:49:26: It's probably a lot more profitable to charge you by prompt than by brand.

00:49:31: Interesting.

00:49:32: Evertune, I haven't heard of it before, but definitely something that we will put in the video description below for people to check out and just see how they like it.

00:49:45: I like to always... make this podcast and the episodes very practical.

00:49:52: So I obviously like to pick the brains of my guests, but also selfishly for myself, but also for listeners and viewers, extract very practical recommendations.

00:50:05: So let's quickly imagine that you're a growth advisor, and I would be a PLG sales company.

00:50:17: I haven't thought of a product, but yeah, you can just make one up if you like.

00:50:23: And I seek your guidance on our organic growth strategy for the future, obviously with all the different changes going on, AI reviews, AI mode, chativity and all the stuff.

00:50:36: And the things I like.

00:50:38: to just keep things focused.

00:50:40: So I say, hey, Louis, I need you to come up with like three priorities for me, like three things that are critical.

00:50:47: What would be your advice?

00:50:49: Yeah, great question.

00:50:50: And to go to the point, first, invest in your data sets.

00:50:53: Make sure that you have available data for the teams to use.

00:50:56: Without the data, the PLG machine does not work, and that is exactly the same for organic growth and for a lot of these machines.

00:51:04: Right now, we are entering the stage of understanding.

00:51:06: We need to make sure that machines understand what we mean and what you're trying to do.

00:51:12: investing that data and that technical foundation of the team.

00:51:17: The second part is question everything.

00:51:20: Right now, I think that most of the organic setups and organizations are not set up for this.

00:51:26: And I'll come to my third point, but I think that right now is the time to really challenge, do titles work, do scopes of work, are the right ones?

00:51:35: Are we limiting too much and putting people into a corner where they cannot do a full, complete job?

00:51:41: So I would definitely look into my art and my structure to make sure that people are in the right place working with the right tools, as I was saying.

00:51:48: And the last one, it's really, really challenged what kind of content are we producing and what kind of content you think that you're doing.

00:51:58: To me right now, what has happened to SEO or organic is a massive diversification.

00:52:03: I imagine all of this as what if Bing and Yahoo would have been successful back in the day?

00:52:08: where you have more platforms to deal with, and those platforms are feeding themselves by other multitude of platforms.

00:52:15: So a content piece is not just a landing page, it's not just, you know, the written words that go into a URL.

00:52:25: To me right now, it's really key to have a three-sixty approach to content that continuously feeds multiple platforms with the same message, and that there's a three-sixty approach to what you're saying to your customers.

00:52:38: So linking to that reorganization part, to me having a team of content that is able to generate multiple formats for multiple platforms, it's one of the biggest drivers of growth.

00:52:52: Interesting.

00:52:55: Another question I had where I'm really curious to hear your thoughts on is there's a lot of stuff floating around on LinkedIn and like about SEO, AI search, etc.

00:53:13: And I feel like with people that are maybe less experienced or maybe they also don't have the time to go into everything in detail and do their own research, they buy into certain things that are snake oil, so to say.

00:53:30: So I'd like to hear what Like from your experience, from your point of view, what are the biggest misconceptions people currently have about the future of SEO and organic growth?

00:53:44: Good question.

00:53:46: The first one is that it's dead.

00:53:49: I'm not dead.

00:53:52: there's no massive layoffs of SEOs around the world.

00:53:56: on the opposite.

00:53:57: I think that precisely all of this snake oil is helping out to bring some attention and to the value that we actually do create.

00:54:04: So to me, the biggest misconception is that this is dead.

00:54:07: It's not dead, it's evolving.

00:54:08: As I was saying, I see social media as right now the closest channel to what organic is because they have dealt with multiple platforms and new impactful innovative platforms like TikTok most recently, where they have to change entirely how they create the content, how they did it, what are the triggers and the ways to implement these things.

00:54:31: So that's one.

00:54:32: The second one is that this is a completely whole new world.

00:54:35: It is not.

00:54:36: We are trying to satisfy customer problems.

00:54:39: And if your SEO strategy or organic strategy was already trying to solve problems for people and trying to drive value to your business, then there is no conceptual change into what you're trying to do.

00:54:50: there's a different implementation, as I was saying, because it's a different platform and they would like different things.

00:54:55: But I don't believe that you need to completely change the way that you think that you're doing marketing.

00:55:03: It's a lot about answering the problems.

00:55:05: If anything, it's going to drive us closer to customers.

00:55:08: It's going to drive us closer to what pain points do people have?

00:55:13: pretty much.

00:55:14: And then probably the last one and relatively controversial.

00:55:23: This is not necessarily a misconception, but I think that this snake oil is useful at the moment.

00:55:30: We do tend to overblow it.

00:55:33: We are definitely a smoke and mirrors industry.

00:55:36: And I remember the first apocalypse that I went through, that's twelve years ago, with the introduction of Penwind algorithmic update.

00:55:49: It helped me out to go deeper into the rabbit hole.

00:55:52: It helped me out to go deeper and connect with other experts to really know what was going on.

00:55:58: It's all snake oil, it's all smoke and mirrors.

00:56:01: So I think that right now, what all of the noise in the industry, it's being helpful for more folks to understand that there's a lot of value that we can create and that the more integrated organic strategy is within the creation of your product and the... in the business itself, the better results you will get.

00:56:22: Yeah, again, slightly controversial, but don't overdo it.

00:56:27: Don't drink the Kool-Aid, folks.

00:56:29: Definitely, you know your industry better.

00:56:32: You know your data.

00:56:34: If you have a question about is this really the end of times, go to your report, go to see your bottom line results and the contribution that you're making to the business, there's the answer for sure.

00:56:46: Awesome.

00:56:47: I think we already have our LinkedIn hook with like the SEO snake hole is actually helpful.

00:56:53: So this will definitely be controversial.

00:56:57: But I think if people hear your whole explanation, it makes so much sense.

00:57:03: Luis, this has been a super insightful conversation.

00:57:06: I think you shared so much knowledge.

00:57:09: It's really great to have you on.

00:57:12: I started asking a final question a couple of episodes ago, and I'd also like to ask you this one.

00:57:21: What didn't we talk about that we should have talked about?

00:57:23: What did I miss in my questions?

00:57:26: That's a great wrap up question and a rabbit hole itself.

00:57:34: Perhaps we didn't touch too much about team organization, about roles, about, you know, skill sets that people need to have nowadays, which I think that it's something that is probably one of the most foundational changes that we are seeing in the industry right now.

00:57:50: To give you a very good example right now, I'm very much attracted to, to, to not necessarily people that have experience within content writing.

00:57:59: but a lot of experience into automation that they are not afraid of going to loveable.

00:58:04: They're going to hug and face and just like try out new things.

00:58:08: High five to my fellow in Miro Federico.

00:58:13: He showed me a lot about how easy can it be to create automation and create solutions at scale, because now coding is also becoming an meaningless task.

00:58:24: Exactly.

00:58:25: Story.

00:58:26: Obviously not to the level but at least to create really annoying things that no developer wants to do For sure it's getting there.

00:58:33: So that's why I think that the skill set that people are needing now to be really successful is changing massively.

00:58:40: I also think that Perhaps that we need to Let go of the old knowledge a little bit and the old bias about how long things are so right now again with the profile perhaps more junior profile.

00:58:56: I've seen a lot of this snake oil as well about their more entry-level jobs.

00:59:03: on the opposite.

00:59:03: I think that an entry-level person is probably on a younger side of things with all of this technology being available to them from the get-go.

00:59:12: I did not have a chat GPT when I started doing SEO and that's why my timelines are slightly different.

00:59:18: But just for someone that started working with this, their first interaction or their first career opportunity was through ChatGPT.

00:59:28: The timelines are completely different, right?

00:59:32: Completely research in multiple languages is not going to be a long task.

00:59:35: It's probably going to be medium type of thing.

00:59:39: So I think that... The profiles are changing, and I think that the organization is changing as well.

00:59:46: The way that teams are structured right now, in my opinion, is to be completely different.

00:59:52: Back in the day, you would have your traditional's content, your traditional's technical side right now.

00:59:59: What do you mean with technical?

01:00:02: So to me, automation, AI, exploitation, data mining, intelligence itself, it becomes a bigger necessity to have it embedded within the SEO or within organic teams.

01:00:16: Something that I'm really proud to have done in Nero is to recreate all of that structure and to move it from a traditional SEO org to a automated AI-first structure.

01:00:29: And that specifically meant, or at least one of the biggest changes in my perspective, was to have an actual team that cares about intelligence.

01:00:38: There are so many data points.

01:00:39: There's a lot of different platforms to be looking at.

01:00:42: that right now is not as simple as looking at your rankings, looking at your revenue, looking at your sessions.

01:00:46: You need to connect a lot more dots and that's a full-time job.

01:00:51: But if you are able to identify through your structure where the biggest opportunity are, where are your users adopting better the product across all of the different platforms, then you have better prioritization.

01:01:04: Yeah, that's a... That's a little bit more about the organization and skillsets.

01:01:09: I know exactly why I started asking this question due to these insights that then just suddenly pop up.

01:01:17: Awesome.

01:01:17: Thanks so much.

01:01:19: If people feel like me that are listening or viewing and they want to hear maybe more of your thoughts or follow you around, what's the best place to go?

01:01:30: Reach out to me in LinkedIn, send a message.

01:01:32: I'm always available.

01:01:33: I'm always happy to have one-on-one conversations.

01:01:36: I've been participating a lot into coaching and development of the early stage marketeers nowadays.

01:01:45: So yeah, reach out to me.

01:01:46: Don't be shy.

01:01:47: Nice.

01:01:48: Luis, we will put your link to the link in the profile in the description.

01:01:51: People go to Luis and don't spam him, so only politely reach out.

01:01:59: Send me a lot of snake oil.

01:02:01: I would love to see more snake

01:02:03: oil.

01:02:03: Nice.

01:02:05: Luis, this has been a really great conversation.

01:02:07: Thanks so much for doing this.

01:02:11: I wish you all the best for the future.

01:02:13: Thank you so much.

01:02:14: very intrigued to see where you're heading, very intrigued to see where this whole AI search ecosystem, organic growth ecosystem is heading.

01:02:25: And yeah, maybe in a year or so, maybe it's time for an update episode, depending on how the world looks like by then.

01:02:34: But until then, thanks so much and all the best to you.

01:02:39: Thank you so much, Nick.

01:02:40: It was a great conversation.

01:02:41: And yeah, let's rain check in a couple of months to see where we are, if we are alive, if the AI has not taken over the world.

01:02:48: We'll see.

01:02:49: So, Luis, all the best to you.

01:02:52: Bye-bye.

01:02:52: Thank

01:02:52: you, everyone.