In 2005, the bank was entering mobile as a channel with no established playbook — mobile best practices didn’t yet exist, and the device landscape was fragmented across wildly different phones, screens, and capabilities. No one had validated which banking tasks actually mattered enough to justify the constraints of a phone screen, and the strategic bet cut both ways: build a custom application tuned to a narrow set of devices, or build once in standard HTML through a translation layer and reach broadly, trading away some native polish. Getting it wrong risked alienating early adopters on one side, or overbuilding for a market that hadn’t yet arrived on the other.