‘The P&L Doesn’t Lie’: Hotels See More AI Costs Than Bottom-Line Gains

What happened

Hotel companies are spending more on artificial intelligence across guest service, revenue management and back-of-house operations, but that investment has yet to show up as profit growth, according to Skift's analysis of hotel groups' financial reporting. The outlet's review of profit-and-loss statements finds that AI-related costs are rising faster than the bottom-line gains operators had expected from automation and AI-driven tools.

The reporting points to a widening gap between the promise of AI in hospitality — faster guest response, sharper pricing, leaner staffing — and what is actually showing up in hotel groups' earnings. Investment continues across multiple functions, from chatbots and virtual concierge tools to AI-assisted revenue management systems and operational software, but the financial case for that spend is not yet proving itself out in reported results.

Why it matters

For an industry built on thin margins and highly seasonal demand, the finding is a useful reality check on AI adoption more broadly. Hospitality has been among the more visible sectors experimenting with AI for guest-facing and operational use cases, and this data point suggests that deployment alone — even at scale and across multiple functions — does not automatically convert into measurable financial return.

The implication for digital transformation leaders beyond hotels is similar: AI spend needs to be tied to specific, measurable operating outcomes rather than treated as a general capability upgrade. Where that discipline is missing, costs accumulate — licensing, integration, change management, retraining — while the offsetting gains in productivity, conversion or retention take longer to materialise, or may require more fundamental changes to how teams and processes work than a tool rollout alone provides.

The Renascence take

This is less a verdict on AI's usefulness in hospitality than a verdict on how it is typically introduced. Most AI deployments in service industries are layered onto existing processes rather than used to redesign them, which caps the upside while the costs stay real.

The P&L gap Skift describes is a service-design problem wearing a technology label. Bolting AI onto an unchanged guest journey or operating model will show up as cost before it shows up as value, because the tool is doing old work faster rather than removing the need for that work altogether. Operators who are seeing returns are typically the ones who redesigned a specific journey — check-in, pricing, complaint resolution — around what the AI makes newly possible, rather than asking AI to quietly automate a step inside a process that was never rebuilt. The lesson for any sector watching this: measure AI against the operating model it was meant to change, not just the budget line it sits on.