What’s Old Is New Again: How Digital Health Marketers Are Winning in the LLM Era

by Fire&Spark ⏐ June 30, 2026

Key Takeaways

  • LLM marketing builds on traditional SEO, it doesn’t replace it. If forced to choose between Reddit and SEO right now, choose SEO. The expansion channels stack on top of the foundation.
  • The patient journey is compressing, not disappearing. Awareness, consideration, and conversion are collapsing into fewer touchpoints; sometimes a single LLM conversation.
  • Reddit, YouTube, digital PR, and affiliate content are among the primary inputs LLMs draw from when patients ask health-related questions. These channels are more relevant than ever. While attribution is difficult, you have to believe in them anyway.
  • The “more with more” opportunity is real, but it requires a deliberate leadership choice. Most organizations are asking how quickly they can get to a leaner org chart. The winners are asking what they could build if they treated AI as an expansion of capabilities.

Digital health marketing teams are chasing a misunderstanding.

Boards want to know why they didn’t show up in ChatGPT. CMOs are hiring GEO agencies. Agencies are rebranding the same services with new acronyms.

The result is a lot of organizational friction. The process isn’t linear, and the playbook that answers the board question rarely matches what actually works. 

Traditional SEO remains the bedrock. Authority, E-E-A-T, site structure, technical optimization, quality content on your own domain. What’s changed is that LLMs synthesize information from across the broader web. The channels that expand your reach beyond your own site now feed directly into AI systems where patients make care decisions.

LLM marketing includes traditional SEO. It doesn’t compete with it.

Katelyn Watson, CMO of Talkspace, has been operating inside this shift for 18+ months. Her team has delivered 30% quarter-over-quarter growth without increasing marketing spend.

The Channels That Stack on Top of SEO

Reddit, YouTube, digital PR, affiliate content. These are among the primary inputs LLMs draw from when patients ask health-related questions. They’re also the channels digital health marketers spent years struggling to justify internally. Attribution was difficult. They sat in the gray zone between brand and performance.

Attribution is still difficult. But these channels are more relevant than ever.

What tends to work is additive. The brands gaining meaningful presence in LLM results have invested in SEO for years and layered Reddit, YouTube, and digital PR on top of that over time. When forced to prioritize, the website comes first in almost every situation. The expansion compounds when there’s a strong foundation underneath it. Without one, these channels have a harder time doing what you need them to do.

The Funnel Is Compressing

The patient journey hasn’t changed: awareness, consideration, conversion. It’s just compressing into fewer touchpoints than most digital health marketers are used to planning around. A single ChatGPT conversation can carry someone from venting about relationship stress all the way to comparing therapy platforms that take their insurance. That whole sequence, before a single website gets visited. 

At Fire&Spark, we’ve been tracking this across digital health brands. Healthcare is experiencing steeper organic traffic declines than almost any other industry. The brands holding ground expanded their distribution beyond their own site while maintaining strong SEO foundations. (More in our guide to AI search marketing for digital health.)

Talkspace’s Channel Strategy

Reddit

Watson’s team was early to Reddit once it became clear how heavily LLMs were drawing from it. The first attempt felt promotional in spaces that don’t respond well to that, and the communities made that clear. The execution that followed was more deliberate — content calibrated to how each subreddit actually works, what belongs there, what earns trust, what gets removed. That kind of platform fluency takes time to build.  

Metrics they track: direct traffic from Reddit citations, and the percentage of posts that remain live. That second one matters. It’s a proxy for whether your content serves the community or extracts from it.

YouTube

YouTube has reportedly surpassed Reddit in LLM citations. The shift is thinking about it as a citation source for AI systems, not just a video library.

For most digital health brands, this is a more accessible starting point than Reddit. The content usually already exists. 

Talkspace takes content they already have, brings it into YouTube, embeds videos on-site. The content exists. Repurposing it is a distribution decision, not a production one.

Digital PR

Recency carries meaningful weight in how LLMs pull from digital PR coverage. A recent piece in a smaller, credible publication often does more than something years old in a widely read outlet. 

Talkspace has built a process around this, using AI tools to match content to the right publishers at scale and keep fresh coverage coming without depending on traditional PR relationships for every placement. 

Attribution

These channels are more relevant than ever. While attribution is difficult, you have to believe in them anyway.

Watson’s toolkit: holdout groups, lift studies, publisher-run studies, updated “how did you hear about us” surveys that include LLMs.

“Hope is not a strategy. But you put the art and science together, look at the balance sheet, and it should come out on the other side.”

Don’t change everything at once. Keep enough of your existing mix stable to isolate what’s moving the needle. We see this across the digital health brands we work with—the ones with the clearest picture of what’s working are the ones disciplined about what they change and when. (See Bicycle Health and Form Health.)

Building the Organization

Watson took her SEO and website leader and named them the internal AI champion. Their job: lead tool adoption, educate the team, be the person others go to without escalating to the CMO.

“I’ve spent zero dollars learning all of this. I’m not an LLM expert. I’m a CMO trying to make sense of things.”

What’s driven the results at Talkspace has less to do with LLM expertise and more to do with attitude. A willingness to experiment, learn from what didn’t work, and stay patient with a process that doesn’t move in a straight line. Getting started costs less than the conversation around it suggests. 

Testing budget: Talkspace allocates 10-15% of marketing budget to testing. For Reddit or digital PR: $5K-$10K per month to start.

Three principles for leading through this:

  • Be the teacher. When you run into misconceptions, be patient. That patience builds the trust you’ll need when you want to move fast.
  • Assign an internal innovation leader. Find the person who derives excitement from confusion. Empower them formally.
  • Make failure low-stakes. Trying something and having it not work is initiative, not a miss.

Opportunity and Reality

Watson’s arguments: 

  • This is the best time in a marketer’s career. 
  • AI opens room for high-level strategy. It automates the drudgery. 
  • The question isn’t “how do we do the same with less?” but “how do we do more with more?”

And yet.

The same week we recorded this, we were talking with digital health brands that had cut their marketing teams significantly and expected the same output from whoever remained. According to Challenger, Gray & Christmas, Becker’s Hospital Review & Forrester 2026 Marketing Predictions, April job cuts rose 38% over March, while AI led the stated reasons for the second month in a row. The marketing job market at entry and mid levels is difficult. Burnout is real. A lot of the chaos isn’t the productive kind—it’s organizations hopping on the AI cost-reduction train without a strategy for where it’s going.

Which is true? Both, depending on where you’re sitting.

The “more with more” opportunity is real, but it requires a deliberate choice. AI does create leverage for marketers who use it well. 

The brands getting the most out of this are using AI in both directions: to move faster and to go deeper than they could before. That requires a team that trusts where it’s being led. Used that way, it changes what a skilled team is actually capable of. 

For teams built the right way, the ceiling keeps rising. The brands that recognize this will build real competitive advantage, especially relative to competitors using AI to shrink headcount and calling it a strategy.

The problem: “more with more” requires a leadership decision. It requires someone who asks what could we build if we treated this as an expansion of capabilities rather than a reduction in costs. A lot of the organizations moving fastest on AI right now are optimizing for a leaner org chart, not a more capable one. There’s a different path available, and some digital health brands are already on it. 

What that path requires is pretty simple to describe and harder to act on: a clear point of view on what AI should be building toward, and a team that’s been set up to move in that direction. 

If you’re executing: The marketers who will be most valuable are the ones who become orchestrators—people who set long-term vision, ask the right questions of AI tools, run measurable experiments, and focus on impact over flash. The ability to tell the difference between AI output that moves a business forward and AI output that just looks impressive is going to be rare and valuable.

If you’re leading: The question is whether your AI strategy is building something or just reducing something. Cost savings are real. But if your competitors are all making the “same output, fewer people” bet, the brand that figures out how to do materially more—more content, more experimentation, more personalization, more speed—is going to look very different in three years.

The other piece is trust. The marketers navigating this well are doing so because someone helped them see it as opportunity rather than threat. That’s a leadership choice too.

Watson left us with an optimistic frame, and she’s right—for the people and organizations willing to act on it. The question is whether that’s you.

Looking Ahead

The dominance of traditional SEO tactics isn’t going away, but the landscape is expanding. LLMs are synthesizing information from across the web—Reddit, YouTube, publishers, your own site—and patients are making care decisions inside those conversations.

The brands that will win over the next year are the ones that:

  1. Maintain strong SEO foundations. Authority, E-E-A-T, technical optimization, quality content on your own domain. This remains the bedrock.
  2. Stack expansion channels on top. Reddit, YouTube, digital PR, affiliate—these channels feed into AI systems. Build presence deliberately.
  3. Optimize for conversion, not just traffic. As AI tools vacuum up informational queries, lower-funnel content becomes more valuable.
  4. Make the leadership choice. Decide whether your AI strategy is building something or just reducing something.

For more on how we’re thinking about this at Fire&Spark, read our guide to AI search marketing for digital health, explore our digital health case studies, or talk to a strategist.