Learning from LangfuseObservations of what makes Langfuse so successful

Note: this is not a culture document as published on the Langfuse website, the team might frame things differently, this is just what I have observed over time:

On team and culture

Everyone on the Langfuse team loves their craft and really likes to work. Yes, while it's ok to not make work too important in your life I think at Langfuse we actually want work to be a big part of it. We like to create and get better at things.

There's this shared pull to ship something big by the end of every week. People get impatient when things drag, or when someone reports "completed tasks" instead of "user value shipped." Seeing the others ship pushed me to ship more, and better. I have the feeling the whole team works like this, and it's contagious.

Langfuse loves focus. Guests are always surprised by how little we talk. Whole mornings would pass without a word. Everyone knows what they're working on. Probably not for everyone, but I love this.

That focus comes from a few things:

  • Very little ownership overlap. Each person owns their slice end to end. You figure out what to work on, then do it. People help each other, yes, but you rarely need to sync. There's real beauty in a big chunk of work living in one single brain. It works because everyone knows what they're doing and has the drive to get it done. No one needs motivation.

  • (Almost) no goals. No strategy. Shipping is the strategy: I had to learn this one. My first weeks I went looking for the strategy, the KPIs, the quarterly goals. There weren't any. The only focus is a strong weekly shipping cadence. Shipping fast is the strategy. This works because the market moves so fast and everyone gets user feedback multiple times a day. It's powerful because it kills the "will this move the needle?" debate. We'd just ship three things. We never worried about what each small change did to some goal. We care about revenue to some extent, but it rarely drives a specific product decision. Revenue follows from many tasteful improvements that compound into real user value.

  • Few meetings, no documents, no waste. Many companies create so much waste on documents and internal alignment. At Langfuse, almost everything lives in code. No slides, no docs. Very few meetings means time to actually think and work. Lunch is the time for a daily casual sync if needed.

  • Small changes that compound over big projects. This didn't come naturally to me either. Marc is a beast at it - he has an allergy to starting big complex long projects. I love shipping refined, ambitious things. It's a strength and often a weakness. Langfuse does the opposite. Small, achievable projects beat big ones, every time. Big projects create surface area for things to break. Small ones compound and create value while you sleep. If something was months of work, it rarely got started.

On picking directions / what to work on

Marc, Clemens and Max have incredible taste on the big decisions. Here are 3 things that I think make Langfuse especially successful:

  • Build for a global market, sell in the US, from day one. Focusing on a single European market just slows you down. If you build in or for Europe, please try it globally or at least for all of Europe, from the start.

  • Competition is good. It's not a reason to avoid a market. It's proof there's money to be made. When Langfuse started, many people said "Datadog will just win this." They were wrong. I love how VCs and fellow founders like to disqualify early ideas based on competition. "Oh there is a player so and so already" - when there is a great market, there is competition. Note: while Langfuse got started in a competitive space, the team never really thinks about competition that much - direction comes from talking to users not from analysing what others are doing.

  • Winning by caring more. Competition is good. It means there's a market. But it's only fun as long as you're winning. Having watched Marc, Max and Clemens, I don't think winning comes from being smarter or working harder (as an input). The team is successful because they care more. They genuinely love their customers and couldn't imagine building for anyone else. They also love the distribution model of PLG dev tooling. I think founders underestimate how much it matters to pick a customer group and a distribution model they really dig. It's a kind of taste - taste of knowing yourself and what you actually find interesting. If you work on something that genuinely interests you, you end up working harder and smarter (as an output!), simply because you care more. The Langfuse team, and the ClickHouse team, do care a lot.

TL;DR: pick a global (=big and broad) market with competition that you are comfortable competing in because you care more about the users than anyone else.