Sumaya Oneill: What AI Actually Changes About the Guest Experience

Sumaya Oneill has been reporting on the intersection of AI and hospitality for Hospitality121 since the technology moved from the back office into the guest-facing layer of hotel operations. Her account of what has changed, and what vendors are overselling, is one of the most grounded perspectives on AI in hospitality you will find anywhere in the industry press.

AI That Optimises Versus AI That Enables

She begins by drawing a distinction that she argues most industry commentary misses: the difference between AI that optimises existing processes and AI that enables new ones. Most of what the hospitality industry has deployed to date falls into the first category, revenue management algorithms that make faster and more accurate pricing decisions than human analysts, demand forecasting that processes more variables with greater consistency, and housekeeping scheduling tools that route staff more efficiently. These are genuine improvements, and their commercial value is real. But they are fundamentally doing what hotels already did, faster and at lower marginal cost.

The more interesting territory, Oneill argues, is the AI that creates new guest experience capabilities rather than optimising existing ones. The most compelling examples she has encountered in her reporting are in pre-arrival personalisation, specifically, the ability to synthesise data from multiple sources (booking history, stated preferences, loyalty tier, social media in markets where data permissions allow) to arrive at an arrival experience that is genuinely tailored rather than segmentation-based. The difference between giving a guest their preferred room type because they booked it and anticipating that a guest who has stayed twelve times and always requests extra pillows might also appreciate early check-in is a meaningful one. The second behaviour requires intelligence; the first requires only compliance.

Where Conversational AI Still Falls Short

Oneill is direct about where AI falls short in the guest experience context. The conversational AI deployed in hotel chat functions and virtual concierge systems has improved significantly in its language capability but has not solved the fundamental problem of guest intent recognition. Guests communicate in elliptical, contextual, and often emotional language that current systems handle poorly. A guest who messages to say they are celebrating their anniversary and asks for a recommendation is communicating far more than the semantic content of their words, and the properties where that message is handled by a human who understands the full context are delivering a materially different experience than those routing it through a chatbot trained on FAQ data.

AI as a Staff Support Tool, Not a Replacement

She is particularly interested in the role AI is beginning to play in staff support rather than staff replacement. The properties she considers most sophisticated in their AI deployment are not those trying to reduce headcount through automation, they are those giving frontline employees better information, faster. A check-in agent who can see a guest's complete preference history, any service failures from previous stays, and a real-time alert that the guest's room is not yet ready is in a fundamentally better position to deliver a warm, informed welcome than one working with a name and a booking reference. The AI is in the background; the human is still the experience.

Where AI Will Have the Biggest Impact in the Next Three Years

**Where do you think AI will have the biggest impact on hospitality in the next three years?**

Personalisation at scale, specifically, the ability to move from segmentation-based service to genuine individual-level personalisation across the full stay lifecycle. The technology to do this is largely available. The constraint is data quality and integration, which is an operational and organisational problem rather than a technology problem. The hotels that get their data infrastructure right in the next eighteen months will have a personalisation capability that is very difficult for competitors to replicate quickly.

The Most Common AI Misconception in Hospitality

**What is the most common AI misconception you encounter in your reporting?**

That the goal is to reduce the number of people involved in the guest experience. The most commercially successful AI deployments in hospitality are augmenting human service rather than replacing it. The properties that approach AI as a cost-reduction tool tend to deliver a worse guest experience and see it in their scores. The ones treating it as a capability-building tool for their teams are seeing both better guest outcomes and better staff retention.

What Hotel Leaders Should Ask AI Vendors

**What should a hotel leader ask a technology vendor who is pitching AI capabilities?**

Two things. First: can you show me this working in a live hotel environment, not a demo? Second: what does the integration with my existing data infrastructure look like, and who is responsible for it? The answers to both questions will tell you everything you need to know about whether the technology is ready and whether the vendor is serious.

Frequently Asked Questions

What is the difference between AI that optimises hotel operations and AI that enables new guest experiences?

Optimisation AI makes existing hotel processes faster and more accurate, pricing algorithms, demand forecasting, and housekeeping routing are the clearest examples. Enabling AI creates genuinely new capabilities, such as synthesising guest data from multiple sources to deliver pre-arrival personalisation that would be impossible for a human team to execute at scale. Both have value, but the second category represents a more transformative shift for guest experience quality.

Why do hotel chatbots and virtual concierge systems still struggle with guest communication?

Current conversational AI systems have improved substantially in language fluency but continue to struggle with guest intent recognition, the ability to understand the emotional and contextual layers behind what a guest is actually asking. Guests communicate with nuance, implication, and feeling that trained FAQ-based systems are not equipped to interpret accurately, which is why the properties delivering the best digital guest communication are still combining AI routing with human response for anything beyond straightforward transactional requests.

How can hotel leaders evaluate AI vendors before making a technology investment?

The two most diagnostic questions are: can the vendor demonstrate the technology working in a live hotel environment (not a controlled demo), and who owns the integration work with the property's existing data infrastructure? Vendors who cannot answer both questions clearly are typically offering a product that is not yet operationally mature, regardless of how compelling the technology appears in a sales presentation.