Episode 493AI in MarketingContent StrategySEO

How to scale content with AI without losing your point of view, with Rafael Moiseev

Rafael Moiseev, CMO at Customer Times and principal of the fractional CMO consultancy Apex Strata, explains how his team kept scaling content after losing their content manager by replacing volume with intention. He locates intention at the intersection of four questions: what you are genuinely good at, what people are actively looking for, what you can prove, and whether other people already use you as a proof point, which is why his team stopped trying to be everything to everybody and narrowed to a few topics at once. To give AI a real point of view, he builds what he calls institutional memory: interviewing the company's internal consultants about their perspective and who they follow, running a heavy analysis of their previous work to capture writing style, storing the company's bank of solutions and artifacts, and recording every thought leadership event so all of it feeds back in, with the expert moderating the output rather than drafting it. On AI search, he describes AEO as a bullet train where you either pop on or get left in the dust, since buyers no longer navigate and discover, they ask a question and get three or four vendors back, and his highest-performing play is the research report, including competitive ranking reports where he is still cited as the source even when his own company is not ranked first. He also traces one buyer end to end, from an LLM answer to the report, to a repeated thirty-minute read of the website, to the about page, to a form, to a live conversation and pipeline, tracked with nothing more exotic than HubSpot multi-touch attribution and Microsoft Clarity session recordings. His advice to a team without this machine is to audit what they publish and why, ask uncomfortable questions about positioning and product quality, and only then build a strategy, because he will not market something he does not believe in.

Go deeper on this topic

Myca: AI Content Assistant

Masset's AI assistant for finding and generating on-brand content in Slack, Teams, and the Masset app.

Meet Myca

The Content Alignment Playbook

A practical framework for keeping marketing, sales, and customer-facing teams on the same story.

Open the playbook

When Not to Use AI

A grounded look at where AI helps, where it hurts, and how to avoid low-trust content decisions.

Read the article
Rafael Moiseev

Rafael Moiseev

CMO at Customer Times and Principal at Apex Strata

15 min

Key Takeaways

  • 1Rafael's test for content intention is the intersection of four questions: what are you really good at, what are people actively looking for, what can you prove, and are other people already using you as a proof point, which pushed his team to stop trying to be everything to everybody and focus on a few topics at once.
  • 2The content machine predates AI, and his team was running one for the sake of running one, publishing thought leadership because a consultancy is supposed to have thought leadership, until they lost their content manager, never replaced the role, and rebuilt the function around AI plus a much narrower topic set.
  • 3Institutional memory is how he gives AI a point of view: interview the internal consultants who are the firm's actual asset, ask who they follow, run a heavy analysis of their previous work to capture writing style, store the company's bank of solutions and artifacts, and record every thought leadership event, panel, and outside speaker so it all feeds the same memory.
  • 4The human stays in the loop as a moderator rather than a writer, so the expert reviews the AI's draft and says what is true, what they never said, and what they do not believe, which still takes time but far less time than writing the piece themselves.
  • 5His AEO play is the research report, including competitive ranking reports that name every implementation partner in the category, because even when his own company is not ranked first he is still cited as the source of the ranking, and he verified the payoff by tracing one buyer from an LLM answer through the report, the website, the about page, and a form into a live conversation and pipeline using HubSpot and Microsoft Clarity.

About this episode

AI can produce a thousand blog posts. It cannot tell you what you believe. Rafael Moiseev, CMO at Customer Times and principal of the fractional CMO consultancy Apex Strata, joins Content Amplified to explain how his team kept scaling content after losing their content manager, and why intention matters more than volume. Rafael breaks down institutional memory, the practice of interviewing your internal experts for their point of view and their writing style, analyzing their past work, and recording every thought leadership event so the AI has something real to write from. He describes AEO as a bullet train where you either pop on or get left in the dust, and shares the play that has worked for him: research reports, including competitive ranking reports where he still gets cited as the source even when he is not ranked first. He also walks through a buyer journey he traced end to end, from an LLM answer to the report, to the website, to the form, to pipeline, using HubSpot and Microsoft Clarity.

Topics covered

  • Scaling content with intention instead of volume
  • Building institutional memory for AI content
  • Giving AI a real point of view
  • Research and ranking reports as an AEO play
  • Tracking an AI search buyer to pipeline

Notable quotes

“I think the term is institutional memory. We have to extrapolate their point of view in one way or another, in addition to their writing style.”

— Rafael Moiseev(04:10)

“It's a bullet train, right? So you either pop on or get left in the dust. You're either in the huddle or out of the huddle.”

— Rafael Moiseev(07:22)

“Who is the best implementation partner for X implementation if you're searching for it? Well, here's a report and all the competitors, and here's how we rank. And even if I'm not number one in terms of rank, I'm still being used as a reference because I created that report.”

— Rafael Moiseev(08:22)

“I wouldn't work for anybody whose product or service I don't believe in. We're marketers, we're not liars at the end of the day.”

— Rafael Moiseev(12:18)

Resources mentioned

  • Framework

    The Four-Question Intention Test

    Before a team publishes anything, Rafael runs the topic through four questions that together define intention: what are you really good at, what are people actively looking for, what can you prove, and are other people already using you as a proof point. A topic has to sit at the intersection of all four to earn a place on the calendar. The test exists because the alternative is what his own team was doing before, producing thought leadership because a consultancy is supposed to produce thought leadership, with no intention behind it. The practical consequence is narrowing: instead of trying to be everything to everybody, pick a few things at once and deliver quality content that scratches a specific itch people already have.

  • Playbook

    Building Institutional Memory So AI Writes With a Point of View

    Rafael's answer to the fact that AI has no point of view of its own is to build one from the people who do. Start at the individual level: interview the internal consultants who are the firm's actual asset, ask what they believe and who they follow, and run a heavy analysis of everything they have written so the system learns their writing style, taking liberties where the source material is technical rather than well written. Then add the company level: the bank of solutions the firm has built over years, the artifacts, and the case material. Then keep feeding it, because he records every thought leadership event the company runs in each city, including the outside speakers and the panels, and inserts all of it into the same memory. The last step is moderation, not authorship: when something goes out under an expert's name, that expert reviews it and says what is true, what they never said, and what they do not believe, which costs real time but far less than writing it themselves.

  • Playbook

    Research Reports as the AEO Play

    Rafael frames AEO as non-optional because buyers no longer navigate and scrape websites to discover vendors, they ask a question and get three or four names back, so the job is knowing exactly what that question is and what proof you can attach to it. His highest-performing format over the last several months is the research report, using AI to run in weeks the kind of research that would have taken six or seven months a decade ago. The strongest version is a ranking report: identify the parameters buyers actually rate providers on, understand where you fit, then rank the whole field including your competitors. He is candid about why it works even when it is not flattering. In his own category as a Salesforce implementation partner, he is still used as the reference for who the best partner is, because he is the one who created the report, whether or not he sits at number one in it. The supporting layer underneath is ordinary proof: implementation counts, case studies, and a lot of content that answers the questions directly.

Full Episode Transcript

Benjamin Ard00:00 — Welcome back to another episode of Content Amplified. Today I'm joined by Rafael. Rafael, welcome to the show.

Rafael Moiseev00:05 — Thanks. Thanks for having me.

Benjamin Ard00:06 — Yeah. Rafael, I'm excited. This is a timely subject and this is going to be a ton of fun. But before we dive in, let's get to know you a little bit. If you don't mind sharing a little bit about your background and who you are, I'm sure the audience would love to know who we're talking to today.

Rafael Moiseev00:19 — Sure, so I'm Rafael. I'm a sitting CMO here at Customer Times. We are a global tech firm specializing in AI automation, end-to-end data solutions, customer experience. I've been here for about five years. I'm also the principal of my own consultancy, Apex Strata, where we help emerging SaaS companies with all of their marketing needs from a fractional CMO basis.

Benjamin Ard00:43 — Love it. That's awesome. So, Rafael, you have some incredible experience. Today we're diving into AI. It's really, really cool, but we're going to talk about scaling content with AI and really kind of how it plays into the idea that you can write content with intention, how that plays into AEO, all that kind of fun stuff. So, for you, how do you scale content but do it intentionally? Like, how are you not just succumbing to the AI machine and just writing masses of content? How do you kind of measure the line and find the balance for yourself?

Rafael Moiseev01:14 — It's a good question. I think the machine was around before the onset of AI. We used to have a content team here, really. We had a head content person who was doing an excellent job, as well as several freelancers, where we tried to kind of build this machine. And our focus was always around thought leadership. But to be honest, a consultancy like us, we're supposed to be thought leaders and we're supposed to project and display our expertise and our experience when it comes to various technologies and industries, but we were doing it for the sake of doing it. So there was no real intention there. And then at one point we lost the content manager and we never really hired anybody to replace them because of the onset of AI. And I think the intention lives at the intersection of what are you really good at? What are people actively looking for? What can you prove? And are other people kind of using you as a proof point? So instead of trying to be everything to everybody, we are really focused on just a few things at once, making sure that we're delivering quality content that's scratching whatever itch people have.

Benjamin Ard02:19 — I love that. So when it comes to scaling and AI is helping you have like more volume of content creation, how do you work intention into the machine? Where do humans insert themselves? Where do the quality standards come from? What does that kind of look like on a day-to-day basis?

Rafael Moiseev02:37 — Well, it comes down to tools and specific tools that we've adopted. And in terms of the human, there always needs to be some sort of moderation. I've been hearing this again and again, AI is not gonna replace people, it's gonna replace people who don't use AI, and that's true. I think ultimately we're gonna move towards one person potentially doing the marketing for an entire organization. I'm seeing the improvements on a weekly, if not daily basis with the kind of tools that are out there, but I'm using specific tools and I'm always experimenting with newer tools. And I'm not doing it just for the content engine, but several other engines that I oversee here as a CMO. We can get into the details of what it is that they do. It's nothing that you can't find anywhere, but like I said, you want to scale and become a thought leader when it comes to a specific topic and write as much content as you can around that topic with your point of view. And AI can provide you, or maybe cannot provide you, with a specific point of view. That's arguable. But that's essentially what we're doing. So we're focusing on one channel at a time, and the few channels that we have been focusing on over the last several months have been rather successful.

Benjamin Ard03:45 — So how are you getting that point of view? And I love this because I do agree. AI by default doesn't necessarily have a point of view. It's adopting points of view, it's learning, researching. How do you translate that? Like what part of the interaction with the tools and the humans are you able to actually say, this is what we believe, and this is our angle. How do you get that in that system effectively?

Rafael Moiseev04:10 — Yeah, so I think the term is institutional memory. And when I say intention, we have specific people within our company who are our, you know, quote unquote consultants. They have specific points of view. If we want to attach them to some sort of AI tool that we're building, and we have built several ourselves, we have to extrapolate their point of view in one way or another, in addition to their writing style and other human-esque type of characteristics. So that's pretty simple. We just interview them. So we ask them questions. We ask them who they're following. We do a pretty heavy analysis on their previous works. Not everybody's a talented writer, so maybe some of their work is more technical. And we will take liberties there, but that's essentially the foundation. So it works around an institutional memory, and the tools and the AI that you built work off of that institutional memory to start generating and producing quality content. And again, if there's something coming out on my behalf, and it does, I'm usually the one who's doing it, but just as an example, if I'm creating something for Ben after interviewing him for several hours and knowing what he's talking about, Ben then needs to moderate it and say, yeah, that's true. Well, hey, I didn't say that. That's not true. I don't believe this. And that's also time consuming, right? But ultimately it's taking up significantly less time than you writing it yourself.

Benjamin Ard05:31 — Yeah. So how are you storing and implementing institutional knowledge? I love the interviewing to collect it, but how are you actually technically saving it and making sure that AI is using that institutional knowledge?

Rafael Moiseev05:46 — So now you're asking me really technical questions. I don't have the exact answer to that question, but I can break it up from what I know. You need to look at it from a company perspective as well as an individual perspective. So if a company is made up of several consultants that are ultimately the asset, they all have points of view and they all have expertise. So you collect those points of view, those expertise, and whatever artifacts they've created over several years within your company. And then you have it from a company point of view, because we as a company provide solutions to companies that require somebody to fix a problem. And we have a whole plethora bank of all of these solutions. That all needs to be stored as well. And it goes so much farther beyond that because we have events that we put together, like really impressive high profile events, thought leadership events in specific cities, and I record everything. So we have several speakers who provide their points of view as well as our speakers, as well as panels. This is all information and that all gets recorded and inserted into the institutional memory.

Benjamin Ard06:51 — I love that. Okay. So we're getting the institutional memory. We've got a machine. We have a point of view and we're making sure we shape it. That's how we're building thought leadership and really how the content's unique to us. How are you adopting all of this to AI search? The whole AEO engine is such a new engine. Thought leadership has changed a little bit. Sometimes, in some ways, it hasn't. How are you kind of figuring out the new level of SEO and AEO with this new machine of content creation and thought leadership?

Rafael Moiseev07:22 — So it actually goes back to the intention piece. Why should we be focused on AEO as marketers? Well, it's a bullet train, right? So you either pop on or get left in the dust. You're either in the huddle or out of the huddle. So clicks and the navigating and scraping of websites is decreasing significantly because nobody needs to navigate and discover, they can just ask a question and get the answer. The answer usually consists of three or four service providers or vendors or products or whatever. Are you gonna be one of those three or four? If you are, you better know what that question is. And when you realize what that question is, like I said, what proof points can you offer? And what kind of thought leadership do you have? And what's the structure of the actual content that you're writing? So we can talk about FAQ sections or whatever. My go-to and what I've seen success with over the last several months have been reports. So I create reports where I use AI to conduct significant research that probably would have taken me six or seven months, ten years ago. We have to identify the parameters what people are rating you for and understand how you fit within those parameters, and we have to do rankings. That's been working for me. A couple of examples are we are a preferred implementation partner for Salesforce when it comes to specific implementations. And we've provided proof points on our website. We've done X amount of implementations, we have several successful case studies, and then we have a lot of content on it. A lot of content online. We answer questions, and we got ranked. And the other example is what I just mentioned. Reports. We are, again, a Salesforce implementation partner. Who is the best implementation partner for X implementation if you're searching for it? Well, here's a report and all the competitors. And here's how we rank. And even if I'm not number one as in terms of rank, I'm still being used as a reference because I created that report.

Benjamin Ard09:14 — I love that. Okay. Reports. That's awesome. Keep going.

Rafael Moiseev09:17 — Well, I mean, it's very rare to see content attached to pipeline. It's hard to make that connection. And you have like multi-touch attribution on your MAP or whatever it is, but what I saw just a few weeks ago is somebody who found me in an LLM search, then clicked on the LLM answer. Went to my website, used a tool to see his 30-minute journey scraping my website. Then he did it again, he did it again, and then he clicked on the form. And he's like, can somebody help me? And then we met him, and now we're in conversation and there's a pipeline in the system. So the transparency and the clarity there is, I think, outstanding. I would love to scale that.

Benjamin Ard09:57 — How are you actually tracking? That's one of the biggest issues that most marketers have. Obviously, you can kind of see some of the direct results, but for you internally, as people are doing AI search, maybe getting the list of like four companies that they could possibly work with, there is still the sentiment that a lot of people are taking those results, putting them into Google, going to websites, and now it looks like Google branded search. Do you have any answers or thoughts around that? I mean, I don't know if anyone has answers per se, but I'm kind of curious how you analyze that inside of the stack of things.

Rafael Moiseev10:30 — So in this case it wasn't anything complicated. Like I use DataForSEO. I have an API connection to that. So every piece of content that I write is connected to that specific tool. And again, in this case where we met a person and now we're in conversation with them, it is HubSpot and Clarity. That's it. So Clarity is, I think, I believe a Microsoft tool, if I'm not mistaken. We have the multi-touch attribution set up in HubSpot. We saw that he went to the report and he read it a couple of times, and then we saw that he went on the website. We have a video of his journey on the website, and then he reached out to us. We saw him go to the about page and read about the CEO and the company and how long we've been around and the other stuff that we do. And it came down to a form and then I have a BDR who's dedicated to inbound. And again, we're a consultancy, so we don't get that many inbounds. This is a relationship-based business. That's the biggest challenge. It's not, hey, here's a video for product-led growth, and here's a free trial and here's my 250 a month. These are quarter million dollar deals, right? So you really need some sort of trust. And I think face-to-face is the best way to do it. And it eventually led to face-to-face. So in this case, it was just really simple tools.

Benjamin Ard11:40 — I love it. Very cool. So, one final question because we're almost out of time. This is something I'd love to end almost every single podcast episode on. Getting really tactical about what you would do this week. So let's imagine you are now inside of a different organization who doesn't have this machine, hasn't really built out the intention of what the content is for, how they kind of put in their institutional knowledge and their point of view, but really wants to get to that level that you just described here. What are like the first two or three things that you're actually going to start to set up so that you can start to move in this direction and kind of get to this machine you just described?

Rafael Moiseev12:18 — I'm probably gonna do an audit. So I want to see exactly what kind of content they're writing. I wanna ask them why they're writing about it. I want to see how many people are reading it, how it ranks organically. A lot of it has to do with the positioning. And again, it comes down to that intention. Why are you writing about this and what do you want to rank for? And sometimes we have to ask uncomfortable questions. If I'm in a new organization and I'm working for somebody, first of all, I wouldn't work for anybody whose product or service I don't believe in. We're marketers, we're not liars at the end of the day.

Benjamin Ard12:49 — Yeah, that is true.

Rafael Moiseev12:50 — But you're not gonna rank if you have a crummy product. You may, but at the end of the day you're probably not gonna see any revenue around it. But yeah, the audit, the tools, and then if you believe in it, then just put a very cogent, proven, unique strategy together and see what the results look like.

Benjamin Ard13:10 — Love it. Very cool. Well, Rafael, this has been incredible. Thank you for your insights. Love how you're approaching content. Love how you're approaching AI. For anyone who wants to reach out and connect with you online, how and where can they best find you?

Rafael Moiseev13:22 — You can find me on LinkedIn. Rafael Moiseev, M-O-I-S-E-E-V, and apexstrata.com.

Benjamin Ard13:29 — Perfect. For anyone listening, scroll down to the show notes and you will see both of those links listed below. So connect with Rafael. Tell him you came from the podcast if you don't mind. That would be awesome. Say hello. Thank him for his time and insights. Again, Rafael, this has been absolutely amazing. Really do appreciate it.

Rafael Moiseev13:46 — My pleasure. Thanks again.

About the guest

Rafael Moiseev

Rafael Moiseev

CMO at Customer Times and Principal at Apex Strata

Rafael Moiseev is the CMO at Customer Times, a global tech firm specializing in AI automation, end-to-end data solutions, and customer experience, where he has spent about five years. He is also the principal of Apex Strata, his own consultancy, where he handles marketing on a fractional CMO basis for emerging SaaS companies. After his team lost its content manager and never replaced the role, he rebuilt content production around AI and a practice he calls institutional memory, interviewing internal experts for their point of view and writing style, analyzing their past work, and recording every thought leadership event the company runs. He believes AI is not going to replace people, it is going to replace people who do not use AI, and that the winning move is to pick a few topics you can genuinely prove out rather than trying to be everything to everybody. He is also blunt about the limits of marketing, saying he will not work for a company whose product he does not believe in, because marketers are not liars. He uses he/him pronouns.

Connect on LinkedIn

Continue Exploring

Myca: AI Content Assistant

Masset's AI assistant for finding and generating on-brand content in Slack, Teams, and the Masset app.

Meet Myca

Get new episodes in your inbox

Join listeners who get episode summaries, key takeaways, and content strategy insights every week.

Frequently Asked Questions

He starts by rejecting volume as the goal, pointing out that the content machine existed before AI and that his own team was producing thought leadership for the sake of producing it. Intention, in his framing, lives at the intersection of what you are really good at, what people are actively looking for, what you can prove, and whether other people already use you as a proof point. That narrows the team to a few topics at once instead of trying to be everything to everybody. AI then handles production volume against that narrow set, while a human expert stays in the loop as a moderator who confirms what is accurate and strikes what they never said or do not believe.

Institutional memory is Rafael's term for the stored point of view, expertise, and voice of the people inside a company, assembled so that AI tools can write from it. He builds it at two levels. At the individual level, he interviews the firm's consultants about what they think and who they follow, and runs a heavy analysis of their previous work to capture writing style, taking liberties where past work is technical rather than well written. At the company level, he stores the bank of solutions the firm has built for clients over years plus the surrounding artifacts, and he records every thought leadership event the company runs, including outside speakers and panels, so that material is inserted into the same memory. The tools the team builds then generate from that memory rather than from generic training data.

He argues AEO is a bullet train where you either pop on or get left in the dust, because clicks and site navigation are dropping sharply. Buyers ask a question and get an answer naming three or four vendors, so the only question that matters is whether you are one of them, which means knowing the exact question and the proof you can attach to it. Reports have been his best-performing format because AI lets him run research in a fraction of the time it used to take, and because ranking reports force him to identify the parameters buyers actually judge providers on. The counterintuitive part is that ranking honestly still pays: in his own category he is cited as the reference for who the best implementation partner is even when he is not ranked first, because he authored the report.

Rafael says the stack for this was not complicated. He connects every piece of content he writes to DataForSEO through an API connection, and for the buyer journey itself he uses HubSpot with multi-touch attribution set up, plus Microsoft Clarity for session recordings. On one recent deal he could watch the whole path: the person found him in an LLM search, clicked through from the answer, spent about thirty minutes on the site, came back and did it again, read the report more than once, went to the about page to read about the CEO and how long the company has been around, then filled out a form, where a BDR dedicated to inbound picked it up and it became pipeline. He notes his firm is relationship-based with quarter-million-dollar deals rather than a product-led-growth motion, so the goal of that visibility is trust that eventually leads to a face-to-face conversation.

Previous Episode

Why AI is both the best tool and a mortal threat to creativity, with Wesley Quock

Related episodes

EP 49019 min

How to use AI for content without failing the sniff test, with Lexi Croisdale

with Lexi Croisdale

Readers can smell AI writing in one line, and once they do they never give the content a second chance. Lexi Croisdale, Director of Global Content at Varonis, joins Content Amplified to map where content teams should sit on the spectrum between fully automated production and fully human writing. She explains why the sniff test is really an authenticity test, why marketers are now pulling dashes out of their own style guides, and how she pushed back when leadership handed the team a Claude Enterprise account and asked for a 10x return. Her answer was to point AI at the admin work instead of the first draft: trained agents that turn meeting transcripts into recurring notes, and a reporting project that takes CSV exports from every channel and returns the monthly report already formatted the way leadership wants to read it. She also makes the case that if a company expects you to use AI it should pay for you to learn it, and that the strongest human signal a brand has is its own employees with their own bylines.

July 28, 2026Listen
EP 48120 min

Why your content should be infrastructure, not output, with Vanessa Mbonu

with Vanessa Mbonu

The most expensive thing you can make in marketing is an assumption. In this episode of Content Amplified, Vanessa Mbonu, Vice President of Marketing at the NAACP, explains what changes when a team stops treating content as output, checking the box on this week's blog post, flyer, or TikTok, and starts treating it as infrastructure: a strategic system the whole organization feeds, from programming to policy to operations. Vanessa shares the leadership moment that sparked the shift, when her CEO said the organization focused a lot on output and not enough on outcome, why audience-first means meeting supporters where they already are, since they care about voting rights and also watch Love Island and March Madness, and her rule for AI: start with AI or end with AI, but never AI from start to finish. She closes with a tactical playbook: run a skills assessment, use the start-stop-continue framework, automate what a human doesn't need to do, and pick a brand mentor outside your category, whether that's Nike, Beyonce, or Labubu. If your content program feels like throwing things at the wall and seeing what sticks, this conversation shows you the smarter way.

July 14, 2026Listen
EP 47317 min

Why getting cited by AI now matters more than ranking on Google, with Jeramy Gordon

with Jeramy Gordon

For 25 years the game was rank on Google. Now no one scrolls past the AI overview. In this Content Amplified episode, Jeramy Gordon, VP of Marketing at Cisive and a 14-year marketer who started in newspaper journalism, breaks down the shift from SEO to GEO and AIO. Jeramy explains the move from fighting for position to fighting for inclusion: roughly 40% of queries are now answered inside AI overviews, organic traffic is dropping accordingly, and Adobe saw a 1200% jump in AI-driven traffic once engines started citing it. He gets specific on how to earn those citations: structure content around the questions people actually ask, write 3000-word definitive guides that go deep instead of broad, tighten your headers, title tags, and meta descriptions, and build off-site signals through press, podcasts, and third-party mentions that train the AI's model of who you are as an entity. He also makes the case for video and podcasts as proof of human-generated content. If you are trying to relearn the playbook for AI search, this conversation gives you a place to start.

June 23, 2026Listen

Get new episodes in your inbox

Join listeners who get episode summaries, key takeaways, and content strategy insights every week.