Enterprise Reservation & Revenue Analytics at Scale

About the Customer

A leading global hospitality company operating at massive scale

With a wide-ranging portfolio of hotel brands serving millions of guests worldwide, the company has a strong digital presence. Its brands include more than 1 million rooms across 6,600 hotels and resorts across global destinations, reflecting both the breadth of its footprint and the scale of its operational data. The organization manages complex reservations, customer loyalty, financial and operational data across multiple brands, regions and channels.

Challenges

Legacy architecture limiting cloud-scale reservation and financial analytics

As part of its cloud modernization initiative, the organization moved its enterprise data to Google Cloud Platform and sought to replace Essbase while minimizing data movement. Analytics teams initially ran queries on BigQuery directly from Tableau and a custom front end. While this approach supported basic reporting, it could not deliver high performance, multidimensional modeling and centralized governance required for free-form ad-hoc data exploration at enterprise scale. This resulted in:

Performance constraints: Slow query performance significantly deterred ad hoc analysis and slowed new analytics development. This limited delivery quality to “one and done” analyses instead of being able to iteratively optimize and refine results. It also increased delivery timelines and costs. As concurrency increased, revenue and property teams struggled even more to analyze demand, pricing and occupancy in real time, limiting the organization’s ability to scale reservation and financial analytics across teams.

Lack of governance and control: BigQuery functioned as the centralized data repository, yet direct access to raw tables without a platform-independent semantic layer led to business logic being recreated across tools, causing inconsistent metrics across finance, revenue and operations teams and resulting in conflicting results and repeated reconciliation.

Fragmented logic: To compensate for performance and usability limitations, teams relied on extracts, fragmenting enterprise logic and increasing maintenance effort.

Needed to support multidimensional analytics outside of Essbase: Inability to model complex hierarchies and cross-dimensional calculations made it difficult to analyze performance across brands, regions, properties and channels. This limited deeper insights for pricing, forecasting and capacity planning.

Buisness Goals

Fast, governed reservation and financial analytics at scale

The organization aimed to establish a modern analytics foundation that could support both reservation analytics and financial reporting, without increasing operational complexity.

Achieve faster ad hoc analytics across Tableau and a custom front end

Scale analytics across teams while enforcing security and access controls

Maintain excellent query performance during high-concurrency periods

Support advanced hierarchies, calculations and period-based analysis for finance use cases

Minimize data movement and reduce dependence on extracts

Improve delivery quality and timelines

How Kyvos Helped

Blazing-fast, multidimensional analytics with centralized security

The organization selected Kyvos based on its ability to deliver high performance, fine-grain multidimensional analytics with enterprise-grade governance.

Single trusted source of truth: Kyvos semantic layer delivered a curated and governed view of enterprise data, eliminating conflicting results and the need for tool-specific BI extract steps. It maintained consistent performance at scale across analytics use cases—significantly reducing validation cycles and rework.

Advanced multidimensional modeling: Kyvos supported sophisticated hierarchies, aggregations, and complex financial logic through advanced capabilities. Also, calculated member support eliminated the need to create 100’s of individual period-over-period metrics.

Instead, the periods could be defined within the calculated members themselves. This enabled single Kyvos columns to do the work of 100s on another platform. For example, instead of calculating “Last Month’s Profits” and “Last Month’s Sales”, a single calculated member could accurately group any metric by any required time period or category.

The calculated members were able to do this while also accurately respecting complex financial rollup logic within individual aggregations. This led to shorter delivery timelines and accurate, user-friendly consumption.

Cross-dimensional calculations
Ensured KPIs remained accurate across brand, region, property, channel and time, so portfolio-level views always reconciled with property-level performance
Conditional CASE logic
Applied different business rules for scenarios such as prepaid vs refundable bookings or direct vs partner channels, improving accuracy of revenue and margin analysis
Solve-order control
Guaranteed calculations like currency conversion, discounts and taxes were applied in the correct sequence before higher-level KPIs were computed, keeping financial rollups accurate
Period-over-period and scenario analysis
Enabled built-in actual vs budget, forecast and year-over-year comparisons, supporting faster planning cycles and more confident forecasting

Consistently fast ad hoc analytics at enterprise scale: Kyvos enabled reliable, high-performance querying across reservation and financial data, maintaining responsiveness even during peak concurrency periods.

Seamless experience across BI tools: With native connectivity to Tableau, Oracle Analytics Cloud, Excel and custom front-end applications, Kyvos delivered high performance across the consumption layer, while supporting plug-in functionality for navigation, filtering, member selection and formatting.

Operational and financial reporting at enterprise scale: Kyvos supported pixel-perfect reporting and automated, scheduled distribution for operational and financial use cases.

Enterprise-grade security, governance and access control: Kyvos enforced centralized governance and user access, with consistent permissions across tools through full LDAP, RBAC and Okta integration, enabling row- and column-level security at both user and group levels.

Eliminated dependence on extracts: The high-performance semantic layer combined scale and advanced functionality in centralized platform, eliminating the need for tool-specific extracts.