Tabist: Dynamic Pricing

Impact

270+

Hotels use Dynamic Pricing

15,000+

Rooms dynamically priced every month. 

10%+

Increase in RevPAR (revenue per available room)

75%+

Reduction in time spent on pricing and market research

My Role

I worked with product and engineering to design Dynamic Pricing, a B2B AI tool that automatically adjusts hotel room rates across OTAs to improve occupancy and revenue. My focus was translating a complex predictive pricing model into a dashboard hotel managers could understand, trust, and override — balancing automation with the control operators needed to feel confident handing pricing decisions to an algorithm.

Problem statement
Hotels manage pricing across dozens of room types, rate plans, and dates simultaneously — a scale of decision-making that’s impractical to optimize by hand. Yet handing that decision to an algorithm introduces a new problem: operators need to trust a system they can’t fully see inside, and retain the ability to override it when their own market knowledge says otherwise.

Research Findings

We interviewed hotel owners operating a range of property types across Japan—business hotels, ryokans, and luxury hotels—to understand how they approach room pricing. Despite their differences in scale and guest experience, one issue came up universally: pricing is hard to get right.

The Problems

  • No clear answer on optimal price. Owners lack a reliable way to know what a room should be priced at on any given night to balance profit and occupancy.
  • Overwhelming volume of combinations. With multiple rooms, plans, and dates in play, the sheer number of price points is difficult to track and manage.
  • Too many variables, too much cognitive load. Day of week, seasonality, location, competitor pricing, and other factors all influence the “right” price—forcing owners to juggle far more than they can reasonably process manually.
  • High time cost. Researching market rates and updating prices accordingly consumes significant time and effort that owners would rather spend elsewhere.
  • Low trust in automation. Owners are skeptical that an algorithm can price as accurately as they can, and often prefer to keep pricing decisions manual rather than delegate them to an AI they can’t see inside.

Who are the users?

The people responsible for setting hotel and ryokan room prices generally fall into two roles: Hotel/Ryokan Owners, who often manage pricing directly alongside running daily operations, and Revenue Managers, who are dedicated to pricing and revenue strategy at larger properties. Based on our research, we distilled these users into the following personas

Goal

Design a pricing interface that makes AI-driven decisions legible and controllable at scale — allowing operators to understand why a price was set, adjust it in a few clicks when needed, and manage pricing across their full room inventory without becoming overwhelmed by the volume of decisions involved.

Solving the Problem

Ideas & Solutions

  • A calendar-style view so users can locate the price of any room, for any date, at a glance
  • A breakdown of how the AI calculated each price, for any AI-priced room, to help build confidence and trust in AI pricing
  • Visual indicators showing which pricing model is active for a given room on a given day
  • Bulk editing — letting users change the pricing model or price preferences for multiple rooms across a date range in one action
  • A searchable list view of all room prices and plans, filterable by date, party size, price, and discount

Design Constraints

  • High information density had to coexist with visual simplicity — operators needed to scan a lot of pricing data without feeling overwhelmed
  • Because operators update many prices per day, minimizing clicks per action was a priority
  • Given the volume of price data on screen, the design needed to reduce the risk of accidental or incorrect price changes
  • The interface had to remain usable on smaller displays (1280×720), since some operators work on older or lower-resolution monitors

Wireframes

Early concept for the calendar-style room pricing view. Dates scroll horizontally, letting operators scan pricing across a wide date range without leaving the page.

In the price override modal, users select a pricing model and apply it to specific room types over a defined date range and days of the week. Minimum and maximum price limits can be set per room and guest count, giving operators a guardrail against the AI pricing outside an acceptable range.

Final Designs

Calendar view of rooms

A single view showing all room prices for any selected date, built for fast daily scanning.

Room Pricing Model

Operators can change the pricing model for any number of rooms across a date range; selected rooms are highlighted so the scope of the update is clear before it’s applied.

Breakdown of AI Price Computation

Clicking any room price reveals the underlying factors the AI used to compute it, giving operators visibility into a decision they’d otherwise have to trust blindly. 

Listview of Room Prices & Plans

A searchable, filterable table view for operators who prefer to work across many rooms and plans at once rather than date-by-date.

Impact & Validation

Dynamic Pricing is now one of Tabist Compass’s most-used products, adopted by 270+ hotel and ryokan properties.

Users report a 75% reduction in time spent on pricing and market research. Previously, operators tracked pricing manually in Excel, cross-referencing competitor rates across OTA pages by hand — Dynamic Pricing eliminated that process entirely.

Trust in AI pricing improved substantially: 90% of users who were initially skeptical of AI-driven pricing now report confidence in it, largely attributed to the price breakdown feature giving them visibility into how each price was calculated.

Adoption also correlated with a 10%+ increase in RevPAR (revenue per available room) among hotels using the tool.

“Among them, I feel that dynamic pricing is Tabist’s flagship system. Optimizing pricing using AI is extremely effective not only in terms of sales but also in significantly reducing operational workload “

Dream Inn Hakata KUBIKAI