The same app, two very different weeks. What changed in the design after we watched both.
Why two cities at once
We say KinHolo is designed for the European Union and for Africa at the same time. It is easy to write that on a website and hard to mean it. So before the foundations were finished we put the same build in front of households in Lagos and in Berlin, for the same two weeks, and watched.
We were not testing whether people liked it. We were testing whether the same design could hold two very different weeks.
The Berlin week
The Berlin households were on fast home wifi and unlimited mobile data. Their pain was coordination: two working parents, a teenager with a timetable, a grandparent who helped with pick-ups and did not want to be on a group chat. Everyone had email. Everyone had a recent phone.
What they used most was tasks, and what they talked about most was the fair-load view. The shopping list was a distant second, because the shop was close and the trip was short.
What we changed: the grandparent role. Our first version treated extended family as a lighter kind of adult. The Berlin grandparents did not want to be a lighter adult. They wanted to see the calendar and the pick-up tasks, and nothing else, and they wanted that to be obvious to their children. The role now starts with exactly that scope and the household can widen it.
The Lagos week
The Lagos households were on mobile data, often metered, sometimes absent. Two of the phones were four years old. One member did not have an email address and had never wanted one. The pain was not coordination so much as money and errands: who bought what at the market, whether the month’s food budget was holding, and what to do when the usual brand was not there.
What they used most was the budget and the shopping list, and what they talked about most was logging an expense with no signal and finding it already on their partner’s phone when they got home.
What we changed: phone-number sign-in went from a fallback to a first-class path with its own screens, and the offline state got its own visible label rather than a small icon. Expense entry went to a keypad, because typing an amount into a text field on an older Android was slower than it needed to be. And we added a second currency that can be recorded at a rate the member enters, or kept as its own line, because “convert everything silently” was wrong for a household that thinks in two currencies.
What was the same
More than we expected. Both sets of households wanted their children to have their own logins with their own limits. Both wanted to know who had changed what. Both hated notifications that arrived at dinner. Both used the out-of-stock swap within the first week without being shown it.
The design held. The interface language, the white surfaces and teal structure with coral reserved for the one thing that needs a decision, read the same way in both cities. Where the weeks differed, the difference was in which module carried the day, not in what the app was.
What we took away
Designing for two regions at once is not two designs. It is one design with a wider set of conditions, and it makes the product better everywhere. The offline state that Lagos needed is the same one that saves a Berlin parent in a basement supermarket. The keypad that helped an older Android is faster on a new one too.
We will keep testing this way. Four households in each city is not a study, and we do not pretend it is. But it is the fastest way we have found to catch the assumptions a single city would let us keep.
Want the next one?