The sunbeam Show: Episode 1 ft. Mike Anderson – Winning AI search in a regulated $6B+ Business

Mike: Absolutely. I think one of the things that I picked up on shortly after joining Hims & Hers was really the importance of trust with your customers.

As you said, coming from SaaS, I think the bar is considerably higher when you step into healthcare because a lot of the topics that you’re writing about as an SEO or content marketer are very sensitive topics.

As we’ve seen with Your Money or Your Life, there’s a very specific set of requirements and trust signals that are important to be sending when you’re creating content.

Beyond just determining a content calendar and creating content, it’s really important to make sure that you have the level of rigor and editorial standards that undergird your overall program. That was one thing that I observed in working at Hims & Hers from the jump.

At the time that I joined, we actually had a pretty informal brand voice, believe it or not. The company started as more of an erectile dysfunction company and a hair-loss company—still very stigmatized conditions.

There was, as part of the company’s DNA, a real desire to speak to those topics and categories in an authentic but more informal way.

Shortly after I joined, we really took a step back and wanted to interrogate that a little bit more and understand: How do we evolve the brand, and how do we become more credible with our customers?

What that looked like over time was transitioning more into a newsroom or publisher-style model, similar to what you’d find with NerdWallet or Healthline—companies that are certainly more mature as far as their editorial standards.

When you come across their content, there are a number of factors on the page that immediately establish trust. Who’s the author behind the piece of content? When was the content last updated? Who was the piece medically reviewed by?

What you find in looking at those examples is that they’re mapped to the category. The authors and experts contributing to those articles are specialized.

That was the model that we were looking to adopt at Hims & Hers: not just elevating the editorial standard of the company and being a little bit more serious in the way that we spoke about these topics in our content, but also organizing by category.

We had an SEO lead, an editor, writers, and specialized medical professionals all contributing to content for sexual health, hair loss, and weight loss as we expanded into new categories. That was really the model that we looked to take from the jump.

Mike: The first thing that we really had to align on was having a shared set of goals for stakeholders who historically were more siloed within the company.

That included members of our editorial team, SEO, and our medical team. They historically had to collaborate, but within the organization, they laddered up to different leaders.

Teams will sit in different functions within a company—that’s a constant. But making sure that there are shared goals that each team ladders into is really important to being able to achieve that unlock you’re describing.

One important thing that we realized was that we really needed to have very tight feedback loops between SEO, editorial, medical, and legal.

These functions were really critical in advancing content, from defining the content calendar and getting content assigned to making sure that the back-and-forth between the review stages could operate on a very clean and predictable cadence.

It was really about looking at the operational side of it and making sure that the teams were unblocked and laddering up to a shared set of goals so that we could scale content production over time.

Mike: One mistake I think a lot of marketers are making right now, especially in healthcare and finance, is the assumption that more content equals more growth.

I think we’re only seeing that paradigm accelerate as agents and AI-generated content become more prevalent.

The distinction that I would draw is separating the use of AI for content creation from the use of AI to remove friction in the content production itself.

What that looked like at Hims & Hers, for example, was that our medical specialists spent a lot of their time reviewing content and trying to find sources from academic papers, peer-reviewed studies, and medical journals that they could relay back into the content.

That involved a lot of independent research on their part.

We looked at using essentially an AI-supported search engine to help with that part of the process so that less of their time was spent on the back-of-house work of digging for claims and independently verified studies.

That allowed them to spend more time actually poring over the content itself, making judgment calls, and helping the editorial team figure out where we could pull in more of that validation.

Looking at E-E-A-T overall, I think that’s really the paradigm when it comes to using AI and agents.

The ability to create content is no longer the bottleneck. It’s being able to incorporate more of that information gain into the content.

That gets messy and requires a lot of iteration and, ultimately, trust between editorial, medical, and legal to understand what we have on hand that we can incorporate into the content.

You need to make sure that you’re freeing up time for those members of the team to really dig into that on an ongoing basis.

Mike: To be clear, I think that even as recently as one to three years ago, that wasn’t the case. I think that was still the primary objective for SEOs and growth marketers.

Today, as we’re seeing AI take over more of the overall discovery process, AI is capable of answering informational queries—full stop.

If your program is built on pumping out more informational content, AI is solving for that.

When I was at Hims & Hers, we hosted an event with members of our weight-loss community. It was an opportunity for us to hear more about their experiences.

During the course of the summit, one of the questions that I asked almost in passing was, “How many of you are using AI or ChatGPT when you have health-related questions?”

I was expecting maybe a quarter of the hands to go up. Every hand in the room went up.

This was back in mid-2025, before AI-search adoption had accelerated to where it is today. We’re seeing that compound exponentially today.

Your customers are increasingly turning to AI to answer these types of questions.

I think it’s important to understand the goal of AI search. To me, the goal is to understand where your brand is visible.

What are the specific topics or attributes that are important for your brand to show up in? If you’re not visible—or if you are visible—how is AI describing you? Is that accurate? What are some ways that you can improve the overall sentiment?

We look at prompt trackers like Profound, Peec AI, and Scrunch. Those are important measures for understanding how visible you are for specific topics or attributes.

Those can be things like common use cases, customer challenges or pain points, and common questions that surface in your category or industry.

That’s where I would start: getting a baseline read on whether you’re visible and how you’re being described.

Then, when you search for some of the prompts that map to each of those topics or attributes, look at the sources that are coming up. What channels and information sources are being cited most?

Those are going to be the clues that help you understand what tactics to take to start closing those gaps.

For example, Reddit has been one of the more prominent sources cited by ChatGPT, although that has fluctuated significantly by engine and over time.

A lot of the time, what’s being cited might be review sites or publishers. It’s important to understand how you’re being described not just on your own website, but on third-party sites as well.

Mike: To me, I think it’s looking at what your company can say uniquely that is going to be something no other brand can say.

That’s looking for your differentiator and really leaning heavily into that.

Specifically, what that looked like for us at Hims & Hers was looking at what our community was saying about our product and what specific things were surfacing in our customer surveys.

I think it’s looking at the first-party research you have. What are your expert voices able to add? What is showing up in your reviews or your customer-support content?

It’s identifying those specific things and pulling them into the content that you’re writing.

At Hims & Hers, we referred to a lot of our information-gain content as first-party point-of-view content.

That may be writing an article about what I learned throughout my hair-loss journey. It’s pulling in those customer voices. It could be pulling in perspectives from your doctors and incorporating that into the content that you’re producing.

I think that’s going to be more challenging to produce at scale, but that’s where the quality bar is really important.

You need to contribute something meaningful to the content that you’re producing on a regular basis. That’s going to be the thing that really separates what your brand is creating from what your competitor is creating.

Mike: It’s a great question because, going all the way back to the early days when I first started in SEO, I’ve always viewed SEO as a team sport.

One of my favorite things about SEO is the fact that it doesn’t have a natural home.

I’ve had portions of my career where it’s lived in ecommerce, product, content, or growth. There are so many places where it can theoretically sit and make sense.

I think that speaks to the fact that there are so many natural collaborators with SEO.

Over the past five to ten years, what has frustrated me a little bit is that I think it’s gotten too predictable.

A lot of SEOs are comfortable with the fact that there’s a checklist they need to adhere to in order to optimize content, rank highly, and generate backlinks.

I think that’s a very flat way to view SEO.

From my experience throughout my career, good SEO is good execution.

Being able to execute SEO within a company setting requires strong relationships with members of functions that are different from yours. Being able to speak their language and get things done is extremely important.

Looking ahead, I think that we’re seeing a real convergence between SEO and brand in ways that span measurement and the types of tactics required, such as specific earned-media pushes to boost AI visibility.

From the measurement standpoint, the specific tactics, and the overall strategy, what I’m hearing from clients and marketing leaders is the expectation that marketers are more multifaceted.

Even if they’re not specialized in brand, product marketing, or public relations, they at least understand the language and are curious enough to work and collaborate with those different teams and functions.

That’s only becoming more of a premium and more of an expectation.

Ultimately, I think what that looks like is: Let’s say our visibility for one of our categories is really low.

Let’s go in and see what ChatGPT is saying about us. If it requires more of a brand push or more information-gain content, then start those conversations with brand, public relations, or even your customer-support team.

You can say, “We see a real opportunity here. What can we focus on this month to make more of a concerted push to expand our visibility in this area?”

Mike: I think that, relative to SEO, there is a lot of overlap when optimizing for AI search—especially for the technical piece.

A lot of the overlap that we see is crawlability, good structured content, authority, and things like optimizing different fields in the codebase. All of that is really important.

But so much of AI search involves answer engines looking for consensus across both owned and earned surfaces.

That fundamentally requires a lot more investment in understanding what’s happening off the page.

The real challenge, as more of the focus shifts toward AI search, is being able to understand your presence beyond the boundaries of your website and how to influence that.

If you think about a shortlist that ChatGPT, Claude, or Gemini is making—if you’re looking for the top GLP-1 providers—the first question to ask is: Are we showing up?

The second question is: How are we being described?

What the LLMs are looking for is whether your brand is credible, trustworthy, and worth recommending.

If the answer to that is no, I think there are two distinct paths of exploration to take.

One is: Is there a technical issue? Is Claude or ChatGPT unable to fetch our site at that moment of intent?

If so, you want to make sure that your important content is reliably accessible to search and AI crawlers. Depending on your stack, server-side rendering can be one way to address that.

The second piece is: Are we trusted—not just as far as what our brand is putting out, but based on what’s being said on independent review sites or by publishers?

If you’re finding that, time and again, you’re being ranked at the bottom of lists produced by publishers, or your reviews are really poor on places like Trustpilot or G2, then that could be more upstream.

That could be a product or customer support issue.

Those are really important conversations to have in ways that extend well beyond SEO.

If you’re wanting to build more of your own credibility as an SEO, go to your leadership and say, “We need to work on our reviews,” or, “What are we currently doing to improve our response time for customer-support tickets? Could issues like that be contributing to some of the negative sentiment that’s showing up downstream in the answer engines?”

That’s definitely something you want to get ahead of.

Those are the two areas that I would look at.

Number one: How optimized are you from a technical standpoint?

Number two: What are some potential gaps that are creating real issues as far as your perception, trust, and ability to be recommended at the moment of intent?

The operating assumption I would stress to any marketer is to assume that AI knows everything about your brand and work off of that assumption.

What are some of the tactics that you’re taking that you would approach differently knowing that?

Google has historically been built around information retrieval. AI goes beyond retrieval—it’s making judgments and recommendations based on what it understands about the world and about your brand in moments of buyer intent.

I really think it is incumbent on marketers right now to steer away from short-term tactics that might move the needle.

If we suddenly know that Reddit is one of the most cited sources in answer engines, that doesn’t necessarily mean that you need to be spamming subreddits.

I think it’s really important to take the long view, show up authentically in the places where your customers are spending time, and build durable relationships with your customers.

Mike: I think, from my perspective, the job to be done has really evolved from where SEO started.

Again, I think that Google is fundamentally an information-retrieval system, whereas AI is much more about understanding the broader world and looking across both what you’re saying about your brand and what’s being said about you.

One interesting stat from Searchable’s analysis is that 52% of the Claude answers they tested came from model memory without a web search.

I think that is a reflection of the fact that the paradigm has really shifted.

When a searcher goes to Claude with a question, a lot of the time it’s pulling from what it already knows about a brand or product.

That covers both your website and third-party domains. That’s going to include review sites, forums, and social media.

LinkedIn is a really important one, particularly if you’re in SaaS.

It’s increasingly looking at the whole picture beyond just optimizing for your website, which I would say SEO has historically been more indexed toward.

I think it’s really separating how to optimize for a single property from understanding where the areas are that we can influence what AI systems understand about our brand.

Mike: It’s constantly changing, certainly, but my view is that there is a fundamental distinction between the two.

I think the history of Google as a search engine is fundamentally different from how ChatGPT and Claude grew to prominence.

I do think that the way the systems themselves function is fundamentally very different.

I referenced that 52% finding from Searchable’s analysis, where Claude answered without performing a web search.

That’s really interesting to me because, when you think about Google, Gemini, AI Overviews, and AI Mode, those are fundamentally built on top of the search engine and the algorithms.

You tend to see a lot more overlap as far as the pages and domains that are showing up in AI Overviews compared with traditional search.

I think you see a lot more variability with ChatGPT and Claude because these systems weigh different channels and sites differently.

They also behave very differently in terms of how often they’re performing a web search. Claude is partnering with Brave, for example, so a lot of what you’ll see in Claude is very different from what you’ll see in ChatGPT.

I think it’s important to have a very clear-eyed view of how different any given answer can look in each one of those engines.

Mike: I think so.

Ultimately, I think the experiences may start to converge, but I think that the foundation of each of the models is starting from a very different point, if that makes sense.

Mike: I think the best marketers five to ten years from now are going to be the ones that are most resourceful and most curious.

What we’re already seeing with v0, for example, is that the bar to produce a top-one-percent website has never been lower.

As long as you have the curiosity and the drive to tinker and have the creative sense, it’s extremely easy, I think, to build something very polished.

Applying that to the marketer’s perspective, I think you’re laying it out exactly right: Agents are going to be able to work more autonomously.

I think they’re really going to cut down on a lot of that friction that I mentioned earlier in our conversation.

They’ll address a lot of the grunt work that technical SEOs and engineering teams have had to dedicate a tremendous amount of time to historically.

It’s really going to shorten the window between being able to identify an opportunity and act on it, essentially in real time.

I think the things that will really separate marketers going forward are being able to ask the right questions, identify high-value and high-signal opportunities, and work seamlessly with agents to act on those opportunities in real time.