Robinson Club Scales €4M+ Experimentation Success With Opal

Robinson Club, part of TUI Group, operates premium all-inclusive resorts across Europe, Asia, and Africa — destinations where guests expect seamless experiences from the moment they start browsing to the day they check in.

Michael Richter, Manager of Conversion Optimization and User Experience, came to the role without knowing what A/B testing even was. He'd spent years as a developer implementing what he thought were brilliant ideas, sometimes to see them fall flat. Then, suddenly, a role had opened up at a company he had been with for most of his career, outlining something he was always looking for.

That was over a decade ago. Since then, Michael has built an experimentation program that generates millions in proven revenue — simple button tests evolved into complex personalization across five TUI brands in ten European languages. He's created custom JavaScript templates, developed his own Chrome extensions for QA, and pushed the boundaries of what's possible without ever hitting a wall.

Now, Michael details how Opal is scaling everything from content creation, competitor comparisons, and landing pages for Robinson Club's digital strategy. Optimizely has grown with Michael, and now, together, they're tackling AI. After ten years of partnership, he says, "I'm not interested in ever changing systems".

• Michael Richter manages conversion optimization and experimentation across five TUI Group brands — Robinson Club, TUI Magic Life, TUI Blue, and the Maura — covering over ten European languages, making his role significantly more complex than typical single-brand optimization.
• His approach combines both digital analytics and in-person user research at the resorts, where guests are eager to share feedback not just about the website but about the entire brand experience.
• Michael uses a mix of quantitative methods (surveys, revenue mining, analytics) and qualitative research (moderated and unmoderated user testing) to build a complete picture of customer behavior and needs.

• When Michael joined as Conversion Optimization Manager, Optimizely was already purchased by TUI but nobody was actively using it at scale — he started like most do, with simple copy changes and button color tests before evolving into more complex experiments.
• Michael came to the role without any background in A/B testing — he'd spent years as a developer launching ideas, so discovering Optimizely felt like finding the software he'd been searching for his entire career.
• The breakthrough came when Michael removed his personal bias from the ideation phase by directly asking customers where they struggled and what problems they encountered — this shift from opinion-based to customer-driven testing made experiments more complex and yielded significantly better results.
• After years of customer conversations, Michael no longer needs personas — he can simply imagine the real faces, voices, and family situations of guests he's met at the resorts, fundamentally changing how he approaches experimentation from conversion metrics to being genuinely helpful.

• Michael's proudest experimentation wins both came directly from observing real customer struggles — one guest couldn't find family rooms during her booking journey, and survey feedback revealed customers wanted to enter coupon codes earlier in checkout than the team expected.
• The family room test solved a visibility problem by showing all room types available at a resort, even if unavailable for selected dates — guests could see what they were missing and adjust their plans accordingly, generating €2.5 million in additional revenue while the test was running.
• Moving the coupon code field from the payment step to the first checkout step (based on direct customer complaints) delivered €1.7 million in additional revenue during testing and is now deployed to 100% of users.
• These wins highlight a critical insight: teams become so familiar with their own processes that they miss obvious friction points, which is why direct customer observation and feedback, plus testing and experimenting, are essential for discovering high-impact opportunities.

• Michael started with Opal like most users — asking for simple copy and testing ideas — but quickly realized generic AI-generated content wasn't enough, so he began systematically training Opal on Robinson's unique tone of voice, product details, golf opportunities, and sustainability initiatives.
• The training evolved from tone of voice to full context: landing page frameworks for different purposes, psychological principles and behavioral patterns with examples of when and how to apply them, and even the specific CMS components available (FAQs, USP lists, teasers) so Opal could structure pages correctly.
• Now Opal generates complete, production-ready landing pages in Robinson's tone of voice, applying appropriate frameworks and psychological patterns, structured with available CMS components — work that can be immediately copied into the CMS with reasoning explained for every decision.
• The upfront work was significant, but the payoff is massive: what used to take days now happens in seconds, and Robinson can generate multiple versions for A/B testing to find what works best, dramatically speeding up their entire content creation and experimentation process.

• AI adoption at TUI Group isn't optional — it's mandatory for staying competitive with major players like Booking.com, with Robinson's upcoming strategy day heavily focused on AI implementation and every employee encouraged to experiment with how it can improve their performance.
• Michael has built custom Opal agents that analyze landing pages — both Robinson's own content and competitor sites — identifying which frameworks and psychological patterns are being used, what the page is about, and providing improvement recommendations.
• Taking it further, Michael created benchmark agents that analyze multiple URLs across competitors, comparing storytelling, visual appearance, SEO, and accessibility, then generating executive summaries and specific recommendations for how Robinson can reach the top of the benchmark.
• Tasks that previously took Michael one to two days now happen in seconds with Opal doing the heavy lifting upfront — he describes it as training an "artificial intern" that learns more over time and provides starting points for creative work and test hypothesis development.

• Beyond content creation, Michael trained Opal agents to audit existing pages for tone of voice consistency, factual accuracy, typos, and grammar — catching errors like outdated geographic information (Aventura incorrectly placed in the Mediterranean) and inconsistent formatting (three different ways of writing "all-inclusive" on one page).
• These quality control agents provide reasoning and probability scores for their assessments (like "80% confidence this is a PAS framework"), helping editorial teams work faster while maintaining consistency that would be impossible to manually audit across all content.
• Agentic workflows automate time-consuming manual tasks, freeing human colleagues to focus on what they do best: creating emotionally appealing, on-brand content that captures Robinson's unique experience — work that requires human creativity and brand understanding, not just technical accuracy.

• Michael uses Opal for publicly available brand and product content because it lives inside the Optimizely ecosystem — meaning he doesn't need to constantly provide context through system instructions, and can access it directly within Web Experimentation and collaboration workflows.
• Opal's power comes from being aware of all the experiments Michael has run, the statistics collected, and the content already used — as he teaches Opal more about Robinson's metrics, workflows, and page flows, the results become increasingly sophisticated without requiring constant re-contextualization.
• Michael's background as a computer scientist who never wanted to be a developer merged with content systems experience and mathematics education to create the perfect foundation for experimentation — Optimizely became the tool where all these disciplines converged into one role.
• The platform's accessibility from multiple angles (development, marketing, analytics) means anyone can start working and continuously learn, with AI potentially reducing the need for front-end developers at scale while creating opportunities for more specialized roles.

• Michael's experimentation journey evolved from simple point-and-click tests to custom JavaScript templates that sped up his program, then to building custom Chrome extensions for quality control and leveraging APIs — pushing boundaries constantly without ever encountering limitations.
• The platform enabled whatever Michael wanted to do, even if sometimes "not in a way Optimizely likes" — this flexibility combined with reliable support (including immediate responses from the Berlin team) created a partnership where disappointment never entered the equation.
• When a critical hackathon faced technical issues two days before launch with the recommendations catalog, Michael's CSM contacted the US team who worked through the night to solve it — the kind of support that reinforces Optimizely treats experimentation as core business, not a side product.
• After burning through six CSMs over a decade (all great people who moved to different roles), Michael remains committed: with no disappointments and a platform that continuously enables growth, there's simply no reason to change systems.

Michael Richter's journey with Optimizely isn't just a success story — it's a blueprint for what's possible when curiosity meets capability. He joined Robinson Club as a conversion optimization manager without knowing what A/B testing was, Googling "Optimizely" before his first day. Over the next decade, he built an experimentation program that's generated over €4 million in proven revenue, scales across five TUI brands and ten European languages, and now leverages Opal AI to turn multi-day tasks into seconds of work.

The transformation came from a simple but powerful shift: removing personal bias by listening directly to customers. A guest struggling to find family rooms generated €2.5 million in additional revenue. Survey feedback about coupon code placement delivered €1.7 million more. These weren't clever growth hacks — they were solutions to real problems that only became visible through genuine customer observation.

Now, with Opal trained on Robinson's tone of voice, landing page frameworks, psychological principles, and CMS components, Michael's team produces production-ready pages in seconds. Custom benchmark agents analyze competitor sites and generate executive summaries. Quality control agents catch factual errors and inconsistencies that would be impossible to audit manually. And agentic workflows free his editorial colleagues to focus on emotional, brand-defining creative work rather than repetitive tasks.

What makes Michael's story remarkable isn't just the metrics or the efficiency gains — it's the partnership. After burning through six CSMs (all great people who moved to bigger roles), experiencing immediate support from the Berlin team, and watching the US team work through the night to fix a hackathon crisis two days before launch, Michael has never encountered a wall. The platform grew alongside him, from simple button tests to custom Chrome extensions to AI-powered workflows, always enabling whatever he wanted to build next.

As TUI Group mandates AI adoption across the organization and Robinson's strategy day focuses heavily on staying competitive with Booking.com, Michael's foundation is already in place. He's not scrambling to catch up — he's teaching Opal to be his "artificial intern," training it to understand metrics, workflows, and page flows so it becomes more sophisticated over time.

The lesson from Robinson Club? When you combine customer-centric experimentation, platform flexibility that never disappoints, and AI that amplifies rather than replaces human creativity, you don't just optimize — you transform. And after a decade of growth with zero regrets, Michael isn't interested in changing systems. Why would he? He's never hit a wall.