wtf is data-stack-survey-2025?

Metabase's Data Stack Survey 2025 repository gives you a hands-on way to explore the results of a survey they conducted with 338 companies about how they choose and use their data tools, and how AI is influencing their workflows. Instead of just reading a static report, you get to run the survey data through Metabase (a tool for visualizing and exploring data) so you can dig into the findings yourself, build your own dashboards, and answer the specific questions you care about. The setup is designed to be mostly automatic. You run a single command that starts Metabase and a helper service that loads the survey data into it. Once it's running, you open your browser to a local address, log in with the provided credentials, and the survey responses are already connected and ready to explore. The data lives in a simple file-based database that comes bundled with the project, so there's no external database to configure. This would appeal to anyone who works with data tools, analytics, or engineering decisions, whether you're a founder evaluating what other companies use, a PM researching industry trends, or a data practitioner curious how peers are adopting AI. For example, if you're deciding between data warehouses or wondering whether companies are actually putting AI into production, you could query the raw responses and break them down by company size or industry rather than relying solely on Metabase's written summary. The main requirement is that you have Docker and Docker Compose installed, which is software that lets you run pre-packaged applications in isolated containers on your machine. The README doesn't go into detail about the survey methodology or what specific questions were asked, but it does point to the full report on their website for the editorial take. If the automatic setup hits a snag, there are a few troubleshooting tips, like checking container logs or tweaking environment variables to retry the configuration step.