Booking Holdings Scaled AI Tooling for Engineering

Managers should monitor how AI-driven coding efficiency impacts total IT expenditures per project.

Updated on Oct. 2, 2026 in Corporate Finance

Bold flat-color editorial illustration showing a repeating geometric server rack structure, representing institutional tech-infrastructure and AI efficiency. — Booking Holdings has reported a 30% increase in code production after scaling AI tooling across its engineering division using model cost routing. AI Illustration. Upload story photo >

Live Poll

Do you trust that using AI tools will make your own daily tasks easier or faster?

Booking Holdings is scaling AI across its 9,000-person engineering team to boost productivity, reporting a 30% increase in code production. The company currently utilizes model cost routing to manage AI expenses while maintaining its focus on high-frequency customer transactions.

Why it matters

The company aims to transition from a transactional model to a high-frequency business while utilizing AI to optimize its $8 billion to $9 billion annual marketing spend. Leadership views these investments as essential for maintaining competitive parity in the travel sector.

Booking Holdings reports that its 9,000 engineers have increased code output by 30% using AI tools, while merchant bookings reached 73% of total gross bookings. Over 30% of its active customers are now in higher-tier Genius membership programs.

The players

Booking Holdings

A global digital travel platform owning brands including Booking.com, Agoda, and Priceline.

Ewout Steenbergen

The Chief Financial Officer of Booking Holdings responsible for the company's financial strategy and AI investment framework.

The details

Booking Holdings implements model cost routing, which directs simple engineering tasks to open-source models and complex requests to higher-cost, proprietary models. To track efficiency, the firm measures total IT cost per merge request, accounting for both human and AI-token usage. This mechanism allows the firm to scale AI deployment while strictly managing the overhead associated with large-scale software development.

Q2 2026 data was reported during August earnings disclosures.
Ewout Steenbergen detailed the AI strategy at Fortune's AIQ Summit on October 1, 2026.

Market Landscape

Booking Holdings' approach reflects the broader industry trend of integrating generative AI to drive developer productivity in large tech organizations. This moves beyond early-stage AI experimentation into structured cost-routing frameworks designed to protect margins.

Operators should evaluate their own IT cost-per-output metrics to determine if AI-driven coding increases developer velocity without inflating token-related costs. This strategy requires establishing clear cost-routing policies to ensure that expensive LLMs are only utilized for high-complexity tasks.

The takeaway

The firm’s focus on measuring total IT costs per merge request provides a replicable benchmark for managers overseeing large-scale engineering teams. Leaders should prioritize tracking the specific return on investment for AI-token consumption versus human developer time.