Growth Engineer (all genders) SaaS/Fintech

Every company in Germany has to keep books, file VAT and produce annual accounts. It is unavoidable, unloved and far more expensive than it needs to be, and it lands hardest on the smallest businesses. mika is the infrastructure that takes it off them. None of this work is optional, which is what makes it infrastructure rather than a product, and the point of building it is to lower the cost of doing business in Germany. From 100 to more than 700 customers in eight months, and still a small team. Which is why this role is broad, why you own an area instead of a slice of one, and why your scope grows at the speed you do. We work in English, we are international, our market is German. y positive surprise. Choosing one is also a big decision: several steps and several touchpoints before anyone moves. One ad and one landing page do not close it. Tasks You run our growth experiments end to end. You bring the questions worth answering, agree the hypothesis with the founders, build the test yourself, and measure it properly. End to end includes the measurement. Instrumentation, attribution and whatever needs fixing in HubSpot for a result to be trustworthy are part of running the experiment, not somebody else's ticket. We are looking for a marketer who can build, not an engineer who supports marketing. Knowing which hypothesis is worth the week matters more here than the code, and you write the German variants yourself instead of waiting for copy. what you will do Shape the test agenda with the founders, and sync with the rest of growth wherever an experiment touches their area Build what the test needs: landing pages, flows, tools, integrations, automations. The simplest version that will confirm or kill the hypothesis, in days (or better hours) rather than sprints. Be self-sufficient Make the measurement trustworthy. Tracking, attribution, the data model and the HubSpot side of it. Our pipeline data is not yet reliable enough to steer by, and getting it there is part of the job. Read results honestly, including when they kill your own idea, and write down what we learned so nobody relearns it in six months Requirements what you bring to the table AI-native, not AI-curious. Claude or an equivalent is your daily sparring partner rather than a tool you open occasionally. You augment your own output with it and you automate the parts of your job that should not need a human. Here it is a baseline, not a differentiator: you would find working without it slow. You are a marketer first. Funnel, messaging, positioning, offer, what actually moves a number. We’re not looking for an engineer who “builds what marketing wants” We have a strong bias towards people who have been entrepreneurs themselves. You have built and launched something of your own, or run a function as if it were your own company. You have run experiments without being handed hypotheses. Talk us through a real one: what you expected, what happened, what you changed. Visual and copy judgment. Y