Flux

Jan Johannes and his cofounder came up with the core idea years before launching, during university discussions about data silos. When Facebook entered the German market and displaced StudiVZ (a German social network), they experienced firsthand the pain of having all messages and friends trapped in one company's ecosystem. They wanted to build something that let people own their data while still connecting with mainstream social networks—initially exploring federated social networks similar to Diaspora before pivoting to focus solely on messaging.

Flux attempted to be a unified messaging client that could connect to multiple platforms through a modular architecture. Jan built a DSL in Erlang to handle REST API integration, experimented with microservices using Erlang, Go, Ruby, and JavaScript, and invested heavily in technical infrastructure like CouchDB-based distributed applications. The team gained traction with a waiting list of "a few thousands of users" through public talks, pitches, and exposure from StartupBootcamp Berlin accelerator. However, they never entered a true growth phase and failed during private beta.

Email support became surprisingly powerful and allowed them to pivot when platform APIs became restricted. Twitter's API was initially open, but Facebook and Google shifted to closed garden strategies right as messaging became a major focus—bad timing that devastated a resource-constrained startup. However, the biggest issues were internal: Jan over-engineered multiple components (the Erlang DSL connector system, complex view models, early microservices adoption) because he wanted to impress investors rather than solve immediate user problems. His cofounder relationship deteriorated despite initial synergy, and they were mismatched in what the startup actually needed. Most critically, they spent enormous energy pursuing VC funding when they should have focused on bootstrapped consulting and custom solutions from the start. After finally pivoting to target businesses instead of consumers, they were severely under-resourced. A contract negotiation with a major German company became an "endless rabbit hole"—by the time they reached acceptable terms, the company restructured, their contact person changed, and the new contract demanded exclusivity that made the product unusable for other clients. With zero runway left, they couldn't recover.

Jan spent another €15,000 to wind down the company properly and open-source all the software they built. He transitioned to consulting and freelancing for companies needing messaging platforms, using Flux's codebase as a foundation. He credits the experience with massive personal growth as a developer, architect, and CTO, though he acknowledges the learnings came at a high cost: €70,000 in personal savings plus €70,000 in angel funding, all burned with zero revenue achieved.