• A reflection on the decision to embrace artificial intelligence in hotel distribution: not as a technological race, but as a way to create lasting value.

The journey no longer necessarily begins in a search engine. More and more often, users ask an assistant instead. The data confirms this trend with remarkable clarity. Phocuswright reports that travel research through traditional search engines has fallen from 51% to 36% of travelers. Yet no statistic was really needed to sense the shift. When the environment in which booking decisions are made changes, a distribution company has two options: chase the phenomenon or stop and ask what purpose it serves. We chose the latter, taking time to reflect and reassess what we believed the future impact would be.

We did so amid considerable noise. Almost the entire industry seemed to be moving at once. According to Wyndham’s Owners Trend Report, 98% of hoteliers had already begun incorporating artificial intelligence in one way or another. Yet only 32% had truly integrated it into their operations, and nearly three out of four admitted they wanted to do more but did not know where to start. Meanwhile, MIT estimated that 95% of generative AI initiatives in business were failing to produce measurable impact on results. The picture was clear: broad adoption, yet surprisingly superficial. Plenty of activity, little consolidation. And perhaps the most dangerous temptation of all: confusing movement with progress.

This is why the first question we asked ourselves was not whether we should embark on an AI immersion journey. To us, that was the wrong question. The real question was for whom and for what purpose.

The only honest answer always pointed to the same people: those who entrust us with their hotel and, through it, their guests, our clients. This simple shift in perspective changed everything. It turned every decision into a demanding yet clarifying filter. An AI capability that does not improve the experience of the person booking or managing a hotel, no matter how impressive the demo or how compelling the narrative, simply takes us away from our long-term objective. Hotel distribution has long struggled with a well-known problem: booking abandonment rates approach 80%, and more than half of travelers leave because of a poor digital experience rather than price. This is where artificial intelligence has something truly valuable to offer: anticipating needs, personalizing without losing consistency, and responding equally well across every channel. It is no coincidence that among those already using AI effectively, nine out of ten report measurable improvements. The value was there, but only if measured in terms of realistic customer experience rather than novelty alone.

Recognizing this did not make us naïve about the business opportunity, which is enormous. Between 2018 and 2024, AI attracted nearly 65% of all technology investment in the travel and mobility sectors. The AI market in tourism itself, valued at around $3 billion, is expected to more than quadruple before the end of the decade. Yet precisely because the opportunity is so significant, so too is the temptation. For us, innovating with purpose, rather than through inertia, meant resisting the headline and choosing robustness over apparent speed time and time again. A tool is only valuable if it is reliable: if it behaves predictably, maintains data consistency across channels, and is prepared for the emerging reality of agents operating directly within the distribution layer. Building that reliability, and preparing for an agent-driven distribution environment, is less spectacular than launching eye-catching features. It is also the only approach that stands the test of time.

There was another decision we made early on, almost without debate: this immersion would not happen against our people, but through them. Once again, evidence supported intuition. When examining real-world AI usage, most AI-assisted tasks focus on augmentation, with humans remaining at the center of decision-making, rather than replacement. In customer-facing environments, augmentation consistently delivers better outcomes than pure automation. As a result, people remain where judgment matters most: in business decisions that cannot be delegated, in training and refining models, and in handling the exceptions that no machine can adequately resolve on its own.

Alongside this, we nurture a discipline that borders on obsession: documenting the reasoning behind every automated decision and every limit we establish. It may sound bureaucratic, but it is quite the opposite. It is the difference between a capability that we truly understand, and therefore can maintain, audit, and improve tomorrow with a new team or a new model, and a black box that works until, suddenly, it no longer does, with no one knowing why. Accumulating that silent debt was simply a risk we could not accept as a company with more than twenty years of history and a long-standing commitment to the future.

Data Governance to go further

If there was one thing that confirmed we were on the right path, it was discovering what truly represents the greatest obstacle to AI adoption in our sector. It is not a lack of ideas or technology. It is data privacy and security, together with cost. In other words, the real barrier is trust. And trust is not declared; it is built. It must be built from the very beginning, not added at the end as an afterthought. That is why we treat data security and quality as foundations rather than finishing layers. The certifications and compliance framework we already had in place are not badges to display. They are the environment within which innovation can progress without becoming a risk. On those foundations, we have built safeguards that we never consider final or complete: validation at source, traceability of data origins, control over the information used to train models, monitoring of deviations, and more. Data evolves, models evolve, and the industry evolves. The useful paradox is that strong data governance did not slow us down. It allowed us to go further.

So why did Dingus immerse itself in artificial intelligence? Not to win a technological race. After all, such a race can never truly be won; it can only be run. We immersed ourselves because we became convinced that a well-governed AI transformation creates a different kind of value from the one promised by shortcuts: a value that accumulates instead of being exhausted. Better service for hotels. Tools that can be trusted. Data that is properly cared for. People freed to focus on the things only people do well. And knowledge that remains within the organization instead of evaporating. Ultimately, this is what we believe AI will bring: that its value will not be measured by a momentary efficiency gain or a headline, but by trust that compounds over time. The trust of customers, the trust of the industry, and the trust within our own company, as we move toward a horizon of increasingly capable agents and more general forms of intelligence that no longer seem like a distant future.

In the end, investing in artificial intelligence is an investment in that trust. And for us, it was the only investment truly worth making, because it also ensures the continuous delivery of value to the market and to our clients.

Jaume Monserrat
CEO & Co-founder, Dingus