We run sentiment analysis across every verified review. Not to count stars, but to understand what actually happened on the day. Patterns in language reveal what no rating can: whether the guide was present, whether the itinerary held up, whether the operator delivered what they sold.

Operators write their own descriptions. Travelers write the truth. We cross-reference both — systematically. When review language consistently contradicts what the listing promises, that gap is a signal. We treat it as one.

A listing that hides behind vague itineraries and buried exclusions is making a choice. We score how clearly an operator communicates what you're buying — the route, the inclusions, the real price. We also measure inclusions density against comparable tours. Ambiguity costs points. Always.

Price alone means nothing without context. We measure what's included against what you pay, cross-check it against how travelers described the experience, then compare both against similar tours in our database. When reviews say overpriced and inclusions are thin — that's not an opinion, it's a pattern. It costs points.

Roam Benchmark

Most platforms show you a price. We tell you if it's fair.

Every tour is measured against our database of thousands of Japan tours. Normalized by city, category, and experience type. This is the Roam Benchmark: a cross-check on whether what you're paying is fair, and whether what's included is standard for this type of experience. It's the one signal no other platform can compute, because they don't hold the data we do. The benchmark acts as a modifier on the final Roam Score, rewarding tours that over-deliver for their price, and penalising those that don't.