Iceland is an accelerated education in revenue management. Demand is international, highly seasonal, and sensitive to airline capacity, exchange rates, volcanic activity, and global sentiment. Normal market conditions rarely stay normal for long, which makes pricing both high-stakes and hard to automate well.

Before RoomPriceGenie, Keahotels had tried both ends of the RMS spectrum. A traditional rule-based platform gave them control in theory, but Iceland's demand can shift dramatically within a single week — a setup that made sense on Monday often needed intervention by Friday. A more sophisticated AI-driven platform solved the operational load but created a different problem: its pricing decisions didn't consistently reflect Keahotels' commercial strategy. As Ernir puts it: "The technology had changed; the workload had merely changed clothes."

The root issue was transparency. Without visibility into how the model weighed different signals, the team faced risk in both directions — reacting too slowly, or pushing rates beyond what the market could support. They needed automation that could handle the volume no human team could match, while remaining explainable and strategically aligned with each property.