The Problem
A Fortune 500 agribusiness’s digital venture studio had a thesis to test: consumers are drowning in choice and can’t find great new snacks and drinks. On the flip side, brands are starved for real consumer insight. Pairing these two problems together, the Nosherie team set out to build an app to allow customers to swipe snacks they might like, providing valuable customer insights.
The challenge
The parent enterprise for Nosherie had a clear constraint: test your thesis as fast as you can. Investment depends upon user growth, and market indicators. This meant Nosherie need ed to get in front of real users, quickly, without cutting the corners that will make it fall over.
At the same time, Nosherie required a two-sided product:
- A consumer app where people could indicate preferences and enjoyment for snack options (and ultimately be rewarded for their perspectives later); and
- An insight engine that the brand side could learn from.
Add on the admin capabilities and unglamorous operations suite, and we had our work cut out for us.
The approach
We built a product with a clear job on each side. For consumers: a simple, one-thing-at-a-time swipe experience to discover new snacks and drinks, where daily activity unlocked a free themed sample box. For the brand: those in-app choices and ratings became real consumer-insight data on an anlytics dashboard.
To ship fast and run lean, we paired a cross-platform mobile app built in React Native with a modern cloud backend in Azure. We stood up a machine-learning recommendation engine driven off users’ own choices to provide an improved scrolling experience.
And because we were moving so quickly, we wired an in-app feedback funnel straight into the team’s chat. This made every single day start with fresh customer signal feeding the next sprint.
One deliberately unglamorous piece: the sample boxes. Early on, the team packed and shipped them by hand. As it grew, we integrated a logistics partner and automated the daily reports and shipping confirmations, because for us, building a product means building the little world around it, right down to the box that shows up at someone’s door.
The result
A real, ML-backed consumer product (mobile app, backing API, market research dashboards, recommendations, and operations). Built and deployed in 10 weeks. Enhancements continued after launch, and additional funding unlocked.
Who this is for
If you’re a corporate innovation group or venture studio testing a thesis through or facing a stage-gate, this is the shape of the work: get something real into users’ hands fast, and own a build you can bring in-house the moment you’re ready. Speed to first ship, and a product that doesn’t depend on us to keep running.
Have a thesis that only counts once it’s in real users’ hands? That’s the kind of build we love to start.