• Six hundred domain names
• Eight hundred ft.com subdomains
• Three separate digital channels

Problems

• Repeating work
• Unable to let go of old tech
• Experts needed everywhere
• Square pegs, round holes
• Learning 'cliff'

Single channel future

Many others (Guardian, BBC, Gov.UK, Boston Globe) have built single-channel sites that serve all devices with a single website. Ultimately, we should too, because

• Same URL for an article, regardless of platform
• Much lower cost to maintain
• More consistent experience for users
• Easier to comply with accessibility obligations
• Better prepared for Web Components to change how we build websites

How to get there

• Build reusable components that we can share between products
• Design at component level, not page level
• Have a standard, so things are easy to understand and use
• Align design with implementation using a living style guide
• Use powerful and flexible versioning and packaging
• As far as possible, use public standards

Package dependencies

Consultation

• BBC (Simon Stevenson)
• CBC (Barbara Bermes)
• GDS (Frances Berriman)
• Sky (Harry Roberts)

Working in public

http://financial-times.github.io/ft-origami

In summary: catching up

• Build components, not pages
• Respond to content, not devices
• Integrate design with dev in a living style guide

In summary: pioneering

• Make everything modular
• Package management for SASS
• Fully scalable architecture

What we've done so far

• Tested packaging solutions
• Written a standard
• Taken extensive advice
• Built about 10 sample components
• Started creating tools to support the framework

Roadmap

• Finish build service and registry tools
• Build a living style guide
• Build more components
• Encourage internal teams to identify component use cases
• Provide support internally to component authors
• Use components in production: replace legacy code, remove inconsistencies, use in new projects