I’m your host, Gianluca Fiorelli, and in this episode of The Search Session, I’m joined by Amanda King, a growth consultant and enterprise strategist based in Sydney.
Together, we unpack one of the biggest tensions in today’s industry: how to adapt to AI and LLM-driven search without losing the fundamentals of marketing, strategy, and real business impact.
From the return of keyword obsession and the limits of tracking LLMs, to entity-driven content and human-first strategy, this conversation is full of grounded insights.
What's in this episode
SEO fundamentals in the age of LLMs: optimization hasn't changed much, but uncertainty is pushing the industry back toward keyword tracking.
SEO is marketing: shortcuts stopped working, and now the industry is being forced back to strategy, quality, and real business thinking.
Marketing includes operations: understanding technical infrastructure isn't optional; it's part of the marketer's role.
LLM bot behavior: tracking bots is complex, inconsistent, and often not worth the effort due to unclear and obscured behavior.
Bot access in the LLM era: blocking bots isn't practical and allowing access is what keeps you visible in AI systems.
Human-in-the-loop content in the AI era: scaling content with AI risks losing perspective, expertise, and authenticity, making human oversight essential to build trust and meaningful connections.
Entity-based SEO: covering the right entities and aligning with real business impact matters more than keyword volume.
Product- and business-led SEO: using the business’s goals and market reality as a framework to prioritize work, not industry checklists.
Working with data: getting comfortable with data is one of the highest-leverage skills an SEO can build.
Listen to the full conversation to learn how to navigate AI-driven search while staying grounded in marketing fundamentals.
Topics covered: SEO strategy · AI & LLM search · bot behavior · content strategy · entity-based SEO · product-led SEO · brand & marketing alignment · data mindset
About the Guest
Amanda King
Marketing, Growth & SEO Consultant at FLOQ
Amanda is an SEO consultant and strategist with over 15 years of experience. She brings a strong marketing and business perspective to SEO, focusing on growth, prioritization, and measurable impact rather than isolated SEO metrics.
In 2011, she founded FLOQ, where she works with enterprise and scaling brands to connect organic search with broader business strategy.
Amanda contributes to Search Engine Land and has also written for Advanced Web Ranking, sharing insights on how to prioritize SEO through a business-centric lens rather than keyword volume alone.
She is also an international speaker, with appearances at SMX Munich and SMX Paris and an upcoming talk at the Croatia SEO Summit.
Video Chapters
Transcript
Full conversation between Gianluca Fiorelli and Amanda King.
Gianluca Fiorelli: Hi, I'm Gianluca Fiorelli, and welcome back to The Search Session. Today we are going to have a guest from Australia, a strategic SEO consultant and the principal of FLOQ, an independent consultancy based in Sydney.
With over 15 years of experience across the US, Australia, and Europe. She specializes in “retranslating” organic growth into enterprise business goals, moving beyond vanity metrics like search volume to focus on a strong product-led strategy and stakeholder buy-in.
She writes regularly for Search Engine Land, and if you are living in Europe, you have probably seen her at SMX Munich. She is very well known for her analytical approach and brilliant mind. Our guest is Amanda King. Hi, Amanda. What are you doing?
Amanda King: Hey, Gianluca. I'm good. How are you? Thank you for having me on today. I appreciate it. It's always good to talk shop with people and share opinions, particularly as things are changing so quickly at the moment.
Why "New" SEO is 90% Standard SEO
Gianluca Fiorelli: Yes, indeed. In fact, for this reason, I always ask my guests, "How is SEO treating you lately?”
Amanda King: SEO isn't treating me too bad at the moment, but right now I'm trying to remind a lot of people that large language models are not taking over the world quite yet and that doing things well in large language models is 80 or 90% of standard SEO. I'm sure a lot of other consultants, and you, would probably say the same thing. It's a conversation that ends up needing to be had in most instances these days.
Gianluca Fiorelli: Yes, in fact, I agree. All the things, even if, let's say, because it's somehow a necessity, we already, even the people who didn't really like the terms, started to talk about AEO, GEO, and LLMO.
But also, distinguishing between the nuances of the three acronyms, I always say that everything is SEO, and all these new things are new verticals. Substantially, it's doing SEO for answers inside search engines (AEO) and answers in LLMs (GEO, which also includes the training data, so the answer without citation).
Amanda King: Yes.
Gianluca Fiorelli: LLMO is something in between, because it's not really something that can be clearly defined. In fact, it is the acronym that is totally losing its meaning. But yes, everything is substantially SEO.
And talking about SEO, I mean, it is true that if we want to talk not about the container, which can be the ChatGPT chat or the AI Mode chat, but what is contained, which is the answer, which we know is based first on machine learning and NLP, and then everything large language model. But we know that machine learning and natural language processing, all the algorithms of NLP, have been used by Google for years. So it shouldn’t be considered something totally new or revolutionary.
Amanda King: Yes, actually, I posted not too long ago about an article that Justin Briggs wrote back in, I want to say 2018, about natural language processing and how to write for natural language processing. And I was like, nine years later, and this is still one of the most relevant pieces on how to write for natural language processing. So things haven't changed; it's just that people are starting to catch up to it.
What frustrates me, though, and I don't know if you've had this experience, but I think in some ways it has put the focus back on keywords, because everyone is trying to figure out what keywords they need to track in Profound or Peec AI or whatever large language model tracker they're using.
That is something that I've been trying to guide my clients away from as a metric for so long, only to have it come back in the conversation now with large language models. It makes me so sad, Gianluca.
Gianluca Fiorelli: I always want to declare very openly that I hate doing keyword research, but I think prompt tracking, which can be called keyword tracking 2.0, is a symptom of insecurity on both sides. On the business side, businesses want to have something to measure. They knew that keywords, even if we were telling them that keyword ranking was somehow a vanity metric, because depending on what you had before position one, for instance, or in these very frenetic search results pages with all the features, position one wasn't really an indication of traffic, depending on what features you had before, the CTR was different. But I think it was a comfortable metric.
For SEO, it was a comfortable metric in two senses. One, it was easy to explain everything to a client using keyword rankings, and it was easy for them to find some sort of correlation: “If this group of keywords is going down, maybe we are going to have some problems.” Now it’s even worse. I think people are trying to find the same kind of comfort with prompt tracking, even though I’m sure many SEOs know it’s not really a good metric. It can be a good metric, but only if you have a very deep wallet.
Gianluca highlights a hard truth: clients crave the comfort of ranking reports, even as search becomes more fragmented.
The key isn't to stop tracking, but to track smarter.
Advanced Web Ranking’s Visibility metrics show your actual share of voice, helping you turn "vanity metrics" into business-centric growth stories that stakeholders actually understand.
Try Advanced Web Ranking for free and see how your visibility aligns with actual business performance.
Amanda King: Yes.
Gianluca Fiorelli: And you would want to spend thousands of dollars literally to track a very large amount of prompts so you can have some sort of statistical idea. Something that is ultimately wrong because it’s a neutral search, and we know that personalized search is the norm in LLMs.
Information Gain and the Hidden Timeline of LLMs
Gianluca Fiorelli: You were citing 2018, and in 2018, Google filed a patent that people talk a lot about, which is the information gain. That patent was granted in 2022, which is not a strange date if we correlate it with what’s happening. People talked a lot about information gain when it came out, but they are not talking about it now, even though it is even more important to understand what information gain is.
Amanda King: I think the easiest thing is to look at the timeline. It was 2022 when the information gain patent came out, and that was basically at the same time that ChatGPT was really gaining ground, because their first model was released in September 2021 commercially.
It got swept under the rug, right? Because it was the not-sexy version of, “Hey, here’s a thing you need to pay attention to, and here’s a thing where we actually need to understand what’s happening with how we present information to people, and how we do that kind of 10x type thing.“ I think it just got buried in the excitement of large language models.
Gianluca Fiorelli: Yes, but the fact is that people - I don’t know, I’m sure you have the same sensation - go in two ways. I’m a big Lord of the Rings fan, so it’s like the dwarfs digging so deep that they free the Balrog.
There are people who are digging so deep into every mechanism of LLMs that, sincerely, I ask, "Why?" I don’t care what an LLM is doing for some obscure reason. Tell me what the consequence is. I want to know why it starts doing certain things, but more importantly, what the ultimate consequences are, so I can act in terms of marketing, not as an engineer.
And then there are others who are so superficial that they are the most dangerous people. The classic AI bros, as I call them. The ones who think that with a one-line prompt, you can create a constellation of content. We are already seeing what happens when content is created like this at scale. And there are very few people in the middle.
People who are not superficial but also don’t go to the root. If LLMs are doing something so complicated that you have to explain, it’s because they are trying to fulfill an objective. So if you understand the objective, you can work backwards.
For instance, if they are rewarding information gain, then don’t write the same things your competitors are saying. Try to find a balance between consensus and information gain.
If information gain is becoming a defining factor in visibility, then understanding what competitors are (and aren’t) covering becomes critical.
Advanced Web Ranking enables you to benchmark visibility across topics and identify gaps where unique value can drive differentiation.
Try AWR free to uncover where your content can create real information gain.
Amanda King: Yes.
Gianluca Fiorelli: You don’t need to understand all the super complicated mathematics behind it.
SEO is Marketing: Relearning the Fundamentals
Amanda King: But the thing is, Gianluca, people don’t want to do marketing anymore because it’s hard. It takes work, you have to think about your business, and it’s not something where you can click a button and have someone write an article for you. That’s part of the difficulty of my perspective when I come in as a consultant.
I have a business background. I’ve been studying marketing. I was a marketer before I was an SEO, and for whatever reason people have put digital marketing into a different space than marketing. But we are marketers. We have to follow the same business plans, strategies, targets, goals, and the three- to five-year plan. Those are all the stuff we still need to align to.
We can’t put the same 300-word informational explanation of what running shoes are on our website and expect to get the same kind of traction that we did before. The problem is that at one point that kind of content did work, because Google was trying to understand the market as a person would. I’m looking at different competitors and trying to understand who I want to trust. Google didn’t have the technology to do that well until probably about five years ago, after trying to do it for 20 years before that.
There’s this disconnect between what works and what makes sense, and that gap is closing. Large language models have really accelerated that. People like Jono Alderson and Tom Capper have been doing really good talks about how to measure that and how we need to shift the way we think about measurement and why we can’t keep doing the same things anymore and why we can’t still be creating those 300 to 500-word articles.
In fact, some of the work I’ve done on studying content and content consolidation indicates that one of the best things to do is actually getting rid of and consolidating some of that content. It’s an interesting play between what works and what doesn’t in that space.
It’s one of those things where I think we’ve fundamentally forgotten that SEO is marketing, and we don’t want to do the work we haven’t done for the last decade because it was easier to not be marketers online. Now we’re in that retraining process.
The Dual Nature of SEO (Technical vs Content)
Gianluca Fiorelli: Yes, I think it’s also related to the dual nature of SEO as a profession. I’ve always said that SEO moves like a pendulum. There are phases where everything is technical, very deeply technical, and then phases where everything is content-driven. You know, when all SEOs were changing their LinkedIn profiles from SEO to content marketers.
We are like shape-shifting people. We can shift from one role to another, but we are still SEOs. So there is always this balance, and sometimes a difficult relationship, between the technical and the content side of SEO, to simplify it.
In reality, we should think of ourselves more like people working on a movie or a TV series. There is a director, which could be the client, and the production could be the CEO of the company.
And we are like the director of photography, who needs to understand how the camera works, how light works, and what to use to achieve what the client wants in terms of representation in the frame, but also being creative with the use of light and how that helps in creating content, we should find a balance between these things.
Whenever we don’t find that balance, we go back to what we know. If we focus on the technical side, we can go very deep into trying to reverse-engineer everything. If we focus only on content, many of us have no real idea of content, and that can be a disaster.
Amanda King: Yes, and I think as well, because the internet is still relatively young from an integrated business perspective, very few people think of the website and its infrastructure as operational. But it is.
If you think about the role of a traditional marketer, understanding how things work operationally in a business is really important. For example, when I was working at Optus as their mobile phones lead, if I wanted to sell thousands of Oppo phones, I couldn’t, because we didn’t have the stock.
So there is that balance as well. The infrastructure, the technical side, and the website, if you are a marketer, are things you need to understand as part of your portfolio. But I don’t think it has ever really been seen or categorized that way, as operations and infrastructure. But it is.
And so I think even, regardless of your DP framing, which I think is fantastic, at a fundamental marketing level, we still need to understand the operational side of things. And the operational side now includes the website and its infrastructure.
So the technical side is still something that would be under our remit regardless. I think all of these different titles are generally just people chasing trends to make quick money, which is frustrating to witness. I will always say I’m a marketer first, which is not the sexiest thing.
Gianluca Fiorelli: Maybe it’s not the sexiest thing, but it’s the correct one if you want to be taken seriously by clients and the people you work with. If you are not clear in saying that you understand marketing and that you are a marketer working in this specific medium, then the client will start wondering what the difference is between you and a very advanced developer or between you and someone who is a content strategist or a content marketer.
Why Log Analysis Isn’t Practical (Yet)
Gianluca Fiorelli: Anyway, talking about technical stuff, you recently spoke at SMX Munich about bots and logs.
Amanda King: Yes.
Gianluca Fiorelli: As a general topic, which is very interesting, because earlier we were talking about prompt tracking. There is another version, another “faction” that says the real way to understand if your content is at least being scouted by LLMs is to go directly to the logs.
So if you had to summarize it in a few minutes, what was the main point you were trying to deliver to the audience at SMX Munich with your talk?
Amanda King: To be honest, it was actually a bit of a negative message. I was saying that the level of effort required to fully read your logs and understand whether or not an agent - this was specifically in relation to agentic search - was actually performing the desired action on your website, you would actually have to sit down and follow the journey of that bot over the course of a 15- or 30-minute session, trace its path, and see if it got to the checkout or got to the newsletter subscription or whatever the goal was.
My advice at the end was that if you have the time and you’re curious, you can sit down and do it. But the average person, the average SEO, doesn’t need that level of deep understanding for a channel that is a percent of a percent and is still very cutting edge.
It was more about understanding that if you want a full picture of how bots are interacting with your website right now, it’s actually quite complicated. It’s not straightforward. It’s not as simple as going into your logs and saying, “Oh, the OpenAI bot pinged my website 1,200 times.”
It might ping once for every resource, not necessarily per page. Sometimes it’ll ping the same page more than once. There is no real consistency in behavior or reporting. And again, for agentic search, you would have to actually sit down and follow the path in the logs.
We are still in the Wild West when it comes to bot crawling, especially for large language models and agentic search. It’s not easy to pin down or understand, and nine times out of ten, it’s probably not worth the effort, regardless of what your CEO asks for.
Gianluca Fiorelli: Yes, I agree, because sometimes one mistake I see is that people think every time a bot visits you, it means you are being used. But people who are supposed to know how query fan-out works should know that being visited might just mean you are in the consideration phase. It doesn’t mean you are actually used or cited in the answer.
To Block or Not to Block? The Brand Bias of Training Data
Gianluca Fiorelli: Maybe what is more interesting, for those who call themselves GEO, is to observe the training bots, if they are visiting you, and how frequently. That could indicate interest. For instance, if you have created something new, maybe in the next release of a model that content could be included in training.
Amanda King: Maybe, and not even then. A lot of bots for large language models have mixed purposes. Some are for visiting and training, and others, like Google, I don’t think they separate their bots by purpose.
So again, it’s still not consistent. You can’t necessarily point to a bot visiting and say it was for training. It’s very obscured and hidden behind a curtain in terms of what is actually happening. It’s a bit of a minefield.
Gianluca Fiorelli: Yes, and I have a question about this kind of minefield and the difficult decisions to take because one of the many recommendations I’ve heard - which is not very common but tends to be, especially in the news space - is to allow access to LLM bots for research, the ones doing live search, but block access to training bots.
The idea is that if they use your content, they would need to cite you and give you a link. Usually, when answers are based only on training data and consensus, there is no link at all. So this is a recommendation I’ve heard. I can see from your reaction that you don’t agree, and I don’t agree either.
Amanda King: No, I mean, in a perfect world, I would say that large language models should have done all of their initial training with consent, right? You should have had the ability to opt out from the start. But that’s not what happens.
We are in a world where everything we put on the internet before 2021 is part of the training set, whether we want it there or not. And the only way to update, correct, and continue to enhance it, and to build our brand, is to allow all types of bots access to our site. And we are not at a point where we can trust that any of the major players are transparent about what bots are doing, what jobs they have, or whether they respect these directives.
So my unfortunate advice to all of my clients right now is, “Do not block bots unless there is a genuine safety concern, for something like the ByteDance bot, because it’s a relatively large vector for bad actor attacks." Otherwise, you just have to let them loose and do their thing and spend as much money as they’re going to spend crawling your website and pinging your server however many times. It’s an unfortunate reality that I wish weren't the case, but it is.
Gianluca Fiorelli: Yes, I agree. I usually use a very practical example and say, “Okay, if you want to block a training bot, let’s look at a prompt you might be interested in.” For example, how to paint war armor miniature figures, and I show them the query fan-out, whether it’s from ChatGPT or Gemini, and I explain, "You see this? This is Army Painter because in the query fan-out, Army Painter is compared to Citadel, or Scale75 is compared to something else. These brands are part of the model because they exist as entities in the training data. This is BS the model has on entities that are in the training data. If you are not in the training data, you are not going to appear in the query fan-out. And if you are not in the query fan-out, you are probably not going to appear in the list that is going to propose the synthetic answer. Do you want this? If yes, then fine, but I take no responsibility. Don’t come back blaming me if you decide to block the bots.”
Amanda King: Yes.
Gianluca Fiorelli: This is something I’ve seen discussed recently. I read a little test by Dr. Pete from Moz about this, and it’s something I’ve also noticed and mentioned to Duane Forrester. And he also observed that brand bias exists in the training data, not afterward.
Amanda King: Yes.
Gianluca Fiorelli: This is one more reason not to block training bots, at least in certain niches. I can understand the position of news publishers, where it becomes a matter of "You pay me to be included in your training data, or not.” That is fair.
Amanda King: Yes.
Moving Beyond "Helpful Content" Cliches
Gianluca Fiorelli: And talking about bots, let’s talk about humans. You have also written quite a lot about humanizing content, which sounds strange, because content should be human by default.
But we know that since 2022, with OpenAI first and then ChatGPT and other large language models, I think more than 50% of content is now created very poorly using AI. I like the concept of human-in-the-loop. In fact, I use AI for writing content, but with strong constraints and guardrails and many stops to check myself or the copywriter.
Looking to refine your approach to human-in-the-loop content and strategy? In the Search Session conversations with Gianluca:
Purna Virji explores why human-first thinking is essential in an AI-driven world and how to create content that truly connects with audiences.
Himani Kankaria breaks down how to unlearn outdated SEO content practices and build audience-centric, full-funnel strategies that drive real business results.
Shelley Walsh explains how to craft smarter content in the AI era, focusing on purpose, creativity, and building direct relationships with your audience.
Gianluca Fiorelli: So what do you mean by human-in-the-loop for content? Or not just content. What is the importance of the central figure of a human in controlling the output?
Amanda King: It's massive, and the way I look at it is very much like - and I hate the phrase - the “helpful content" concept. Because a big part of what makes content helpful, especially in Google’s eyes, is perspective and expertise, and that can only come from a human.
What I generally advise, whether your content starts with a large language model or with a person, is to begin with input from your sales team, your service team, or anyone who is directly speaking with the people using your product or service. Because they will be able to tell you the unique pain points and the very specific questions people are asking that you want to make sure you answer. It could be, for whatever reason, the shoelaces that don’t have good elasticity, and people want to know if they can switch out the elastics. These are niche questions you wouldn’t necessarily think to address otherwise.
And I feel like, I don’t know about you, but I want to see people. I don't necessarily want to see stock images. I want those Will it blend?-type videos. I want to see someone doing something silly that still explains the thing that they’re trying to get across. I want to see someone cutting through a shoe to show the cross-section, how much rubber there is, or how springy it is.
We want to connect. And I think in this very AI-driven, heavy cadence of just getting things done and publishing quickly, we’ve forgotten that the reason we’re publishing on our website is to get people to believe in us as a brand. That requires more time, effort, and care.
Even with kind of purely large language model-generated stuff, you want to do it in batches, ground it in your own content, your voice, and your brand guidelines, and even add interviews with your CEO, your CMO, or other people. Ground it in the human, because otherwise, what’s the point?
One of the things that I’ve said for a very long time in my presentations and the things I speak about, when it comes to strategy, is, “Would you do this if Google didn’t exist?”
And I think that can be reframed into, “Would you do this if ChatGPT didn’t exist? If Claude didn’t exist?” There’s such an obsession with creating so much content that I think we all need to take a deep breath and slow down a bit and ask, “Why are we creating this content in the first place? What question are we trying to answer? Who are we trying to solve for? What problem are we trying to address?”
Because I think that the disconnect between traditional marketing and digital marketing has made us forget that we are trying to get customers who believe in us. And believing in us and building trust requires us to create content and build our websites from a genuine place. You know what I mean?
Gianluca Fiorelli: Yes, but I think you’re touching on something really important for me. Sometimes, people don’t really understand what Google means by quality content.
On one side, there are people who think quality content has to be very in-depth, something you spend thousands of hours researching, which is fine if you are an academic. But ultimately, quality content is - and you can find the definition in the patent - the content that is able to satisfy the beneficial purpose the person wants to achieve from your content.
For instance, I always use this real example. A client of mine was ranking number one for car insurance with a form, not a landing page, which is very long and complex. No, Google saw that people were always clicking, and we even saw a boost with sitelinks to the form. So Google decided to show the form directly instead of a landing page as the top result.
And I think this brings us back to marketing. At the end of the day, marketing is about understanding how to solve things faster and better for the client. If customer care takes 20 minutes to answer a question, the client will go crazy and leave your company. And it should be the same for a website.
Entities, Brands & Topical Authority
Amanda King: And that’s where things like information gain, entities, and brands come in as well. That’s my overall perspective on where SEO is right now. It’s really a brand exercise.
Gianluca Fiorelli: Indeed. This connects to what you were saying before about the frenzy of creating so much content, when in reality, you need to create content that matters.
That’s why we saw the case of HubSpot that tanked because back in the day, they targeted every possible variation of marketing-related queries. That’s not the right approach to content.
Content should be connected to entities. What are your entities? What is your brand? What kind of entity is your brand? What other entities are connected to? Are you covering all the relevant aspects of those entities?
Want to go deeper on entities? In the Search Session conversations with Gianluca:
Dixon Jones breaks down how internal linking, schema, and topic hubs build the entity structure machines can read.
Lazarina Stoy dives into entity search, query fan-out, and how to strategically position your brand in AI-driven search.
Beatrice Gamba explores how knowledge graphs, ontologies, and structured data define a brand’s core entities and ground AI in reliable, meaningful information.
Gianluca Fiorelli: So, creating content should now be about filling the gaps to demonstrate that you are “the source of truth” on a topic, which is a new phrasing that is coming out.
Amanda King: Yes, and again, that’s all brand. That’s thought leadership. It’s about doing the hard things and looking at your data and seeing what you can publish about, for example, the 20-year trend of mountain climbing in Spain or whatever it may be.
It’s hard work. And digital marketing has felt easy for the last 10 years because we’ve relied on signals that don’t necessarily reflect the impact on the wider business, in the sense that we’re not looking at the holistic business numbers. We’re still kind of looking at our niche.
And now there are a lot of people recommending looking at things like share of voice or share of market, which are traditionally brand measurement metrics, but they look at the whole market.
I think it will be a humbling experience for a lot of people, because organic digital, not even just SEO but all organic, is still a percentage of a percentage of the overall share of voice a business has in any given market.
So there is a rebalancing that SEO and digital need to have in terms of where we sit in the hierarchy of the rest of the business and how we align the work we’re doing, like thought leadership and content, with the reality outside of the internet.
One of the examples that I give when I’m talking about entities is this: would I trust a travel site if I were looking for recommendations about Paris, but they didn’t talk about the Louvre, or the Eiffel Tower, or croissants, and instead just mentioned Paris 120 times? I wouldn’t trust that.
And that is where we need to shift more toward entity and business thinking, rather than traditional keyword or SEO thinking. I still think a lot of people are in the process of making that transition.
SEO is shifting toward Brand and Entity authority. To measure this, you need to look at your market share across entire topic clusters, not just isolated keywords.
Advanced Web Ranking’s competitive analysis tools allow you to group entities and track your brand's "Share of Voice" against competitors in real-time, proving your status as the "Source of Truth" in your niche.
Try AWR for free and measure your Brand’s Share of Voice against any competitor.
Product-Led SEO: A Framework for Prioritization
Gianluca Fiorelli: Yes, and when I was introducing you, I mentioned that your vision of SEO is the product-led SEO. So maybe this also relates to how we think about metrics.
If you are product-led, the metrics might be related to things like share of voice for product names or different products. Entity coverage could be an internal metric to understand if we are covering all the entities related to the product.
So what is your precise vision of product-led SEO, and why is it, for you, the right approach? Because it's usually associated with the SaaS industry, but actually, and tell me if I’m wrong, you can expand this way of thinking about SEO to other niches as well. Am I right?
Amanda King: Yes, and I mean, it is product- and business-led. It’s about focusing on where we see the business going in the next three to five years and working toward that. It also includes business considerations.
For example, if I wanted to migrate from client-side rendering to server-side rendering because it’s considered best practice, but doing that would take three years and five million dollars, then maybe that gets moved down the priority list.
Instead, we focus on things we know we can impact within the next 12 months. So it’s really a prioritization framework, rather than just doing tactics because they are considered best practices or important in the industry.
It’s about using what you actually see within your business and on your website to frame out what will actually have the most impact. And with the product-led side of things, that is also quite relevant in the large language model and ChatGPT side because we want to have as much detail as easily ingestible as possible.
And that’s also, to a certain extent, where the technical side of SEO comes in, with things like schema markup, how HTML is formatted, when JavaScript is rendered, how things are rendered in the DOM, and whether things are server-side or client-side, and things like that.
But at the base of it, it’s about using your business, your product, and your market as a framework for prioritizing your work and what you focus on, rather than following a checklist of what is considered important in SEO at the moment. So for me, it’s really a focus on the business and the product.
Gianluca Fiorelli: Yes, that’s an interesting perspective. I like it. And one question.
Making Data Less Intimidating
Gianluca Fiorelli: So we are getting closer to the end of this episode. If I remember well, you also wrote about the relationship between veteran SEOs and new SEOs.
But instead of asking how we can help new SEOs grow correctly - I mean, we are BS, so it depends on what we mean by correct growth - I want to ask you something else. What would the Amanda King of today tell the Amanda King back when she was just starting in SEO?
Amanda King: I started doing SEO as a copywriter, writing exact match keyword articles for a couple of years, so that ages me a bit. But I would say I would tell my younger self to get friendlier with data quicker.
Gianluca Fiorelli: That’s a very good point, because data can feel scary. So how can you make data feel less scary? Maybe by having an investigative mindset. Seeing data as something to explore and use, or something else?
Amanda King: I’m very much an exposure therapy kind of person when it comes to making things less scary. I would just go into the data, play around, and see what I can find and discover. That’s how my brain works when it comes to getting more comfortable with data.
But part of it would also be learning Excel or SQL and learning how to get slices of large data sets easily. That’s usually the intimidating part. You have a hundred thousand rows of data and don’t know how to look at it all.
So things like pivot tables to summarize data are really useful. And this is also where tools like Claude can help, because you can ask them to process data, identify patterns, or group queries.
I think it really comes down to getting in there and doing it, even if it’s not the most exciting answer.
Gianluca Fiorelli: No, it’s the right answer. It can be summarized as playing with data. I think the biggest block is the learning part. But you’re right; now with tools like Claude, Gemini, with MCP, all these connectors, the boring part of working with data is becoming easier.
It’s like when you play a tabletop game. The boring part is setting up the board and placing all the pieces. And then the fun part is actually playing the game. The boring part can now be delegated to machines, and that’s important.
Amanda, it was a real pleasure to finally have the time to talk with you and deal with all the time zone challenges. Thank you; it was a great conversation.
Amanda King: Thank you for having me. I really appreciate it. It’s always a pleasure.
Gianluca Fiorelli: And thanks to all of you for being here with us, Amanda and me. And now let me do the classic YouTube SEO influencer thing: subscribe to the channel and ring the bell. Thank you and bye-bye.
Podcast Host
Gianluca Fiorelli
With almost 20 years of experience in web marketing, Gianluca Fiorelli is a Strategic and International SEO Consultant who helps businesses improve their visibility and performance on organic search. Gianluca collaborated with clients from various industries and regions, such as Glassdoor, Idealista, Rastreator.com, Outsystems, Chess.com, SIXT Ride, Vegetables by Bayer, Visit California, Gamepix, James Edition and many others.
A very active member of the SEO community, Gianluca daily shares his insights and best practices on SEO, content, Search marketing strategy and the evolution of Search on social media channels such as X, Bluesky and LinkedIn and through the blog on his website: IloveSEO.net.
stay in the loop