SODP

SODP Media

    Menu
    Menu
    SODP logo

    Admiral CEO: Why We’re Embracing Generative AI for Visitor Management

    Dan Rua is CEO of Admiral, The Visitor Relationship Management (VRM) Company. Admiral VRM helps thousands of digital publishers worldwide grow visitor relationships and revenue. Since its launch late last…
    Updated On: December 1, 2025










    ChatGPT


    Grok


    Perplexity Streamline Icon: https://streamlinehq.com Perplexity





    Claude
    Vahe Arabian

    Created By

    Vahe Arabian

    Vahe Arabian

    Fact Checked By

    Vahe Arabian

    Andrew Kemp

    Edited By

    Andrew Kemp

    Subscribe to AI insights

    • Trending pubtech resources
    • Review of pubtech and adtech tools
    • Valuable pubtech strategies

    By Vahe Arabian

    Founder at SODP






      Dan Rua is CEO of Admiral, The Visitor Relationship Management (VRM) Company. Admiral VRM helps thousands of digital publishers worldwide grow visitor relationships and revenue. Since its launch late last year, ChatGPT has captured the imagination of both the media and the general public. Rocketing to 100 million users in record time, the Generative AI platform has caused a stir rarely seen in this day and age. As the dust settles, digital publishers are coming to realize that they’ll need to leverage this technology or risk being left behind. Many publishers have begun to experiment with the tool, searching for ways to integrate it into their editorial workflows. The widely held belief is that Generative AI will turbocharge content creation. At least one company, however, believes Generative AI has the potential to boost business publishers’ productivity outside the creative space. Admiral, a visitor relationship management (VRM) software solutions provider, has just announced the integration of OpenAI’s GPT engine within the visitor engagement layer of its platform for publishers. “Admiral’s intelligent visitor journey builder now automatically drafts conversion-optimized messaging based upon multiple factors, such as page context, value exchange offers, desired sentiment, and more,” Admiral CEO Dan Rua said in the announcement. Admiral says its solution allows publishers to automatically generate CTA offers for their audiences, solving “challenges across the full visitor journey.” To better understand the company’s motivation for integrating AI in its VRM platform, State of Digital Publishing (SODP) sat down with Admiral CEO Dan Rua to discuss the company’s goals and aspirations for the product.

      How has Admiral transformed from an ad block tech company to a platform helping publishers grow visitor relationships and revenue?

      Admiral’s founding MVP way back in 2016 was to size and solve adblock losses. We size them with free analytics and recover them with full-stack adblock recovery, AA-based recovery, or the revenue-maximizing combo. We’ve monitored and protected more than a trillion impressions since our founding. As we did some root cause analysis in 2018-2019 on why ad and data blocking exploded, we stumbled onto the idea that the Internet’s core business model of tracking visitors for ads resulted in weak relationships between publishers and visitors. When there’s a lack of relationship, both sides eventually abuse the relationship — publishers with bad ad experiences and visitors not caring whether publishers live or die. Therefore, our mission of “Saving the free Internet, one publisher at a time” is focused squarely on growing relationships between publishers and visitors. How we do that is called visitor relationship management (VRM) — similar to a CRM and marketing automation, purpose-built for media publishers. VRM encompasses the entire visitor journey, including adblock, email capture, registrations, social follows, mobile app downloads, donations, paywalls, privacy consent, first-party data collection, and more — all from a single tag, with no coding required to launch all modules.  Most publishers start by using us for a single pain point like adblock recovery, email capture or paywalls; and VRM integrates well to coordinate various point solutions. Eventually, publishers love shedding their Frankenstein multiple tags monster, improving page load speeds, and optimizing visitor UX by leveraging our full VRM capabilities. The goal of VRM is ultimately to maximize visitor relationships, data, and average revenue per visitor. Publishers that leverage VRM — whether using Admiral or cobbling together their own multi-vendor VRM stack — will separate themselves from the pack as privacy and user empowerment continue to disrupt the core business model of the Internet.

      Following the launch of chatgpt, has the adoption of generative AI translated into publishers’ daily workflow? Can you provide a few examples?

      Generative AI is hitting all parts of the publishing business, and we’re seeing a few categories of key usage emerge:
      1. Content Creation
      2. Operations 
      3. Marketing
      Although there are some concerns in SEO forums about full ChatGPT article creation, many writers are leveraging Generative AI to save time in researching and outlining articles, something that was happening well before ChatGPT. The primary goal of that use is content creation speed and cost savings. The second area is helping operations. For example, INMA published an example of The Financial Times asking an AI model to plan a new product to segment audiences by reader depth and willingness to pay for news. That is also largely a cost savings application of AI.  The third area, and why we call our work Generative AI that Generates Revenue, relates to revenue growth via marketing content, intelligence and optimization. AI-powered engagements and emails can go beyond cost savings, accelerating revenue through automated messaging and revenue-optimized iterations. We already launched AI-powered smart journeys last year, based upon surging and popular articles. However, when emails or on-site engagements relied upon manual creation, the ROI of testing 100 or 1,000 diverse messages was unclear. But, when those messages are generated via GPT, an intelligent system can constantly A/B test, promote winners, generate new messages, “rinse and repeat.” 

      What are the considerations for integrating open AI’s GPT into the admiral platform? How does integration differ from the current personalization engines focusing on contextualized advertising or content recommendations/conversations?

       When I first saw DeepMind beat all the Atari games it seemed like some form of wizardry. We wanted to capture that same feeling of magic with our innovation — we call that MaiGIC. Most recommendation engines for content or advertising are making a choice between known creatives, or with minor tweaks for placement and keywords. That is really just the tip of the iceberg for what’s possible, so we unleashed full AI-powered message creation, with much cleaner, easier point and click UX than ChatGPT’s prompt writing.  You don’t have to be an expert prompt writer to make MaiGIC happen. Just choose a sentiment (e.g. Funny, Friendly, Happy, Hopeful) and message length, and we’ll automatically incorporate your site’s content (e.g. Sports, Gaming, News, etc.) and offers (e.g. paywall, adblock recovery, email, etc.) into a unique, relevant message that drives conversion. Our team of revenue experts already knew a bunch of visitor journey best practices for messaging, but these MaiGIC messages even surprised us for how well they integrate the context, sentiment and goal. If you ask ChatGPT about the three parts of a magic trick, it will tell you about the “patter”, the “performance” and the “prestige” — the most surprising and memorable part of the magic. Therefore, if our point and click config is the “patter” and the resulting AI-powered message is the “performance”, then the “prestige” is the optimized revenue that results from improved and repeated A/B testing. The goal isn’t just to be wowed by a machine writing better messages than a human, but instead to truly deliver Generative AI that Generates Revenue. Generate multiple personalized messages, compete them against each other BattleBots-style, and repeatedly replace the lowest performing messages with new, automated messages — driving up conversion rates and revenue automatically, forever. That is the “prestige” that is the most memorable part of our MaiGIC. Before I wax too poetic about the power of what we’ve just launched, I’ll also provide an important disclaimer. This is still a very new technology, so we’ve built in guard rails to protect our publishers, allowing humans to edit or approve any message being generated.

      Your announcement also indicates the platform will “leverage visitor and segment-level first-party data to deliver the right message, to the right visitor, at the right time.” do you have any more information to provide at this time that can help plan publishers’ first-party development and setup?

      Although I don’t want to disclose too much about our second act, A/B testing is a simple case of what’s ultimately multivariate testing — both in terms of message/engagement variables and in terms of visitor variables, including their activity, past conversions, and rich first-party data. No audience is homogenous, so ultimately, audiences become segments, and segments become visitors. When messaging can optimize on the fly — for each segment or visitor — then we’re ultimately talking about delivering the right message, to the right visitor, at the right time. This flows naturally from Admiral’s original vision of Visitor Relationship Management, empowering publishers to “meet visitors where they’re at” to grow relationships instead of trying to force a single value exchange to diverse segments.

      What is the right balance between A/B testing offers and CTAs with running amendments on autopilot?

      Most publishers are still at the crawl or walk stage of personalized offers — with many still focused on fairly basic content metering instead of multi-dimensional metering across visitor journeys. We’ve staffed our Customer Love team full of revenue experts to help publishers take the next step with a trusted AI advisor instead of throwing darts blindfolded. Just as Tesla is learning from autopilot assisting drivers before full self-driving, publishers will learn from MaiGIC assisting their optimization, before launching full self-driving CTAs. However, if we’ve learned anything from our past investments in AI, it’s that the progression from assistance to autopilot is coming faster than we could’ve all imagined a few years ago.

      What is your perspective of generative ai for the year ahead, and how does this inform admirals’ product roadmap?

      I’m confident that Generative AI will quickly move from the cost savings focus to revenue generation. In fact, the cost savings aspects of AI could cause some pain in publishing if revenue generation doesn’t follow quickly behind. Our VRM platform was already like a “Swiss Army knife of new revenue” via adblock recovery, paywalls, donations, email revenue, and more; so we’re focused on infusing MaiGIC into every revenue stream we power for publishers. Ad rates are already down year-over-year, and there’s still macro uncertainty about the recession and death of the cookie, so publishers need every bit of new revenue they can get. The lens we use for every new innovation we deliver is our mission statement of “Saving the free Internet, one publisher at a time.” We’re pretty excited about how this launch and future advances in Generative AI will help us deliver on that mission.

      Related Posts

      STATE OF DIGITAL PUBLISHING – COPYRIGHT 2026