Holiday Extras says Codex delivered a working app rebuild in two weeks

Holiday Extras says it reached a fully working rebuild of its main travel app in two weeks using Codex, achieving progress that previously would have taken more than six months. OpenAI’s November 12, 2026 customer case study describes a changed division of work: customer specialists helped shape software while engineers built the technical foundations.

Customer knowledge, reusable building blocks

Holiday Extras helps customers search, book and manage trips. It had spent years developing its mobile apps, with major changes to the main app centered on an annual release. Company-wide access to ChatGPT Enterprise preceded the use of Codex for customer-facing apps and internal systems.

The rebuild paired Liam Thompson, vice president of AI innovation and formerly head of design, with engineers. Thompson brought his understanding of customers into the development work. Small, documented software packages provided reusable foundations, rather than requiring each piece of the app to be built separately from scratch.

Engineers focused on the app’s architecture, security, development workflows and APIs—the connections through which software systems interact. The case study describes a shared effort, not a business specialist independently replacing an engineering team. Employees were encouraged to experiment within packages, security controls and workflows that engineers supplied.

The rebuilt app used customers’ bookings, trips and history to offer a more personal experience, powered by OpenAI models. Codex also enabled features for iOS and Android to be developed together, according to the case study, instead of Android improvements arriving behind those for iOS.

More bookings, with a measurement limit

Holiday Extras reported a 32.5% increase in open-to-book conversion following the rebuild, alongside the lower error rate. Conversion here tracks the move from opening the app to making a booking. These are company-reported outcomes presented in an OpenAI customer story, rather than independently verified findings.

Both the conversion and error comparisons cover different periods for the old and rebuilt apps. That makes them before-and-after measures, not a controlled test isolating Codex’s contribution. The reported gains concern the rebuilt product as a whole, including its changed customer experience.

The delivery schedule changed too. In the first month after the rebuild, the main app moved from an annual major release to weekly releases, each containing multiple new features. The separate Countdown app went from adding a feature roughly every six months to multiple additions a week.

A customer complaint becomes an update

One Countdown example shows how feedback entered that faster process. After a customer complained about a missing list view, an agent prepared a fix. Thompson approved it and the update was submitted. The customer then changed a one-star review to five stars, according to the case study.

The same effort to turn operational knowledge into software extends beyond app design. Director of Innovation and Growth David Lee describes an experienced accountant rebuilding finance systems around business knowledge and colleagues’ feedback. CEO Matthew Pack also builds with Codex himself, using that experience to understand what employees need.

Pack’s next ambition is for more Holiday Extras systems to be administered through AI, with employees directing the work and shaping outcomes. That remains an ambition, rather than a completed company-wide transition. He is asking the leadership team to reconsider which opportunities the business can pursue across its products and markets.