NCBA LOOP

I’m not the only designer on LOOP, the consumer app, Biz, POS, and Mshwari each have their own focus areas. This case study covers what I own directly.

The existing experience was a list of bank products dressed up as an app, account types and menus mirrored the bank’s internal structure rather than how people actually use money.

Kenyans split bills, run chamas, send money across borders, and save in three places at once, often before lunch. None of that showed up in the old information architecture.

Banking interfaces carry constraints most products don’t. Every flow needs compliance sign-off, every number has to reconcile exactly with the backend ledger, and you’re redesigning behaviour for people who already trust the brand with their salary, so risk tolerance for change is low.

LOOP also isn’t one app, it’s four: Consumer, Biz, POS & Cards, and the Mshwari integration, each maintained by its own team with its own legacy decisions. A real constraint was building a design system that could hold across all four without forcing every team onto an identical visual language overnight.

My ownership centres on LOOP Biz, specifically the Supply Financing, Term Loan, and Grow Limit flows, plus the LOOP Merchant component library and design system.

I also lead motion and interaction work that spans the wider ecosystem: the dark mode redesign, the font migration off Poppins onto Inter for cleaner iOS rendering, and Lottie and Rive work in flows that used to be static.

We rebuilt the structure around tasks instead of accounts: pay, save, borrow, grow, the things people actually open the app to do, with the underlying products organised behind that instead of in front of it.

That meant restructuring navigation and information architecture before touching a single visual detail, the hierarchy had to be right before it could be made to look right.

Financial interfaces fail quietly. A confused user doesn’t complain, they just leave, or they call the contact centre, which is worse for everyone. Clarity was treated as a hard constraint, not a nice-to-have.

Every number on screen carries a label. Every action gets a confirmation before it happens, not just an explanation after. Every error states what to do next, not just what went wrong.

Most banking apps treat animation as garnish, a fade here, a bounce there. On LOOP, motion has to earn its keep: a live computation card recalculating a credit limit as you move a slider, a bottom sheet that confirms before money moves, a transition that tells you a state actually changed rather than just looking nice.

I prototype these in ProtoPie and Rive before they reach engineering, mostly to get the timing and easing right. A confirmation that feels instant builds more trust than one that’s technically correct but reads as sluggish.

Banking in Kenya is social by default. Chamas, shared savings goals, family remittances, money rarely moves between just one person and their account. The product had to carry that, not work around it.

We built group features that feel as lightweight as starting a WhatsApp group but carry the reliability people expect from a bank transfer, the same instinct that shows up later in LOOP Biz’s lending and savings tools.

The Supply Financing and Term Loan flows are live in production for LOOP Biz. The rebuilt design system is now the shared foundation for LOOP Merchant, replacing a set of components that had drifted out of sync across screens over time.

The font migration and dark mode work have shipped across the consumer app and are in active daily use. I don’t have clean usage metrics to share publicly for this one yet, the honest proof here is what’s shipped and stayed shipped.

LOOP is where I learned that financial UX seniority is mostly about saying no: no to an icon that’s slightly off-system, no to a shortcut that breaks the audit trail, no to a ’just this once’ exception.

The unglamorous discipline of consistency is the actual craft. The flashy redesign moments are rare, most of the job is keeping four products feeling like one bank.