Algo

The 48-hour brief

Algo was created by a five-person team during the SUEDE Designathon. We explored a familiar frustration: people know recommendation systems shape what appears in their feeds, but they have little visibility into how repeated activity narrows that feed over time.

I led the research sprint, helped synthesise the product direction, contributed to the brand and interface, and co-presented the final pitch. Because the event ran for 48 hours, every decision traded depth for speed.

From research to one clear frame

We used short interviews and a survey to gather fast directional evidence, then grouped observations through a live affinity map. The recurring need was not to remove recommendation systems. People wanted a legible account of what the system appeared to be learning from them and a way to reflect on whether that matched their intent.

The concept became Spotify Wrapped for your feed: a familiar storytelling format for otherwise opaque behavioural data. A standalone app would summarise patterns across social platforms, then present themes, diversity, and filter-bubble signals as a monthly story.

This frame helped the team move quickly, but it also concealed a major feasibility question. The prototype assumes access to consistent cross-platform activity data. We did not establish that technical access during the event, so Algo should be read as a product and interaction proposal rather than an implementable service plan.

What we prototyped and tested

The mid-fidelity Figma prototype covered onboarding, a current Wrapped dashboard, an Insights view, and a six-card story. The visual language borrowed the energy of music recap products so the experience felt reflective rather than instructional.

We ran two Rapidata comparisons between visual variants. The more colourful dashboard and Insights treatments were preferred in those comparisons. These were preference tests, not task-based usability tests: they helped us choose a pitch direction, but they did not show whether people could interpret the metrics, change their feed, or trust the underlying data.

Outcome and limits

The prototype received the SUEDE Designathon People’s Choice Award. That is meaningful evidence that the idea and pitch resonated with peers at the event. It is not evidence of product-market fit, technical feasibility, or a usable production experience.

Algo remains useful in my portfolio as a compact experiment in research velocity, framing, and team decision-making. It sits beneath the featured case studies because those projects carry evidence further into implementation and verification.