Metabase

New in v58: Guest embedding 🪄 Drop dashboards and charts into your app without creating a Metabase account for every single user: pick your content, customize the appearance, grab the code snippet, done.

In Metabase 58 we made Documents available to everyone 🎉 As a Developer Advocate I love seeing our community build awesome projects and dashboards. Just this week I've seen cool projects from: Isaac Somuah, Fares Dadi, Rodrigo Soares, Marcio Reesink, Carola Cardozo, Md Abid Sarkar, Ygor Gonzaga, Guilherme Vassoller Daros, Kamogelo Mogoba, Diego Jorge da Silva, Konstantinos Devetzidis, Fágner Antônio Beraldo Rezende, Bodhana Jayawickrama. Can’t wait to see what everybody starts building with Documents. If you make something cool please tag me. Release notes here: https://lnkd.in/g6vM-MBG

New in v58: Metabot AI is out of beta 🤖 Ask questions naturally, let it write your SQL by @-mentioning a data source, table, or model, and explore data with the new AI interface.

If you’ve ever opened a dashboard, clicked a filter, and then waited… …and waited… …and wondered whether the tab froze or you did — you know this problem. And it’s more common than it should be. It takes weeks, months, sometimes years (if you’re really patient), but eventually teams realize their dashboard technically works, but practically doesn’t. We saw this recently with a US-based financial consulting firm and one of their end clients. Their core sales and finance reporting lived in Power BI. Over time, it became a bottleneck: — Slow load times on real data volumes — Filters that froze dashboards — Visuals that looked fine but didn’t support analysis The ask sounded simple: “Can we migrate these dashboards to Metabase?” The reality was more nuanced. We had no direct database access: all SQL, validation, and iteration happened inside Metabase. So instead of copy-pasting dashboards 1:1, we treated the migration as a redesign. What we rebuilt: → Sales performance across customers, teams, and reps → Store-level views (wholesale vs retail, product categories) → A clean, executive-friendly financial dashboard And more importantly, we fixed the structure: → Tab-based layouts instead of one overloaded page → Filters used deliberately, not everywhere → Visuals chosen for decisions, not defaults The result — predictable load times, no freezing under large datasets, and finally dashboards people can actually work in. We wrote up the full case study — including screenshots and design decisions — on our website → https://lnkd.in/dvVA-7CD If you’re dealing with slow, fragile dashboards and want to see how we approached this one, the full breakdown is worth a read.

Starting the year right... Metabase 58 is here 🔛 Improved Documents are now available on all plans (including Open Source!), Tenants keep customer data isolated in embedded analytics, Metabot AI is out of beta, and more! Full release link in comments ↓

Want to become a data analyst but don't know where to start? We put together a practical roadmap 🗺️ for 2026 designed to be actually doable. Most guides try to list everything you might need to learn. That gets overwhelming fast. Instead, we built a clear path through the fundamentals, from data basics to building dashboards. Check it out in the comments ↓

I’m starting with an “easy” topic: asylum applications in Europe. (yes, that’s ironic) 🌍 For those interested in geopolitics - an overview based exclusively on official Eurostat data 🎯 For aspiring data analysts - how a senior analyzes real-world public data in Metabase 🛠 For the Metabase team & users - how a pro user leverages Metabase for real analytical workflows 📺 Full video on YouTube: 👉 https://lnkd.in/dUjihZpd 📊 Interactive dashboard: link in the comments

Making self-service analytics actually happen. Alejandro R. implemented Metabase at roadsurfer and grew adoption from a few users to 400+ across the company. The data team is not on-call. They're not drowning in requests. Self-service actually stuck. Join Alejandro on January 28 to learn what he did to make it work: - Built an internal Metabase Academy - Focused on verified/official reports for governance - Set up alerts for order monitoring and site downtime And more: https://buff.ly/2qbMm0I

Most analytics stacks appear complete. Data is ingested, models run and dashboards are in place. But in practice, teams still struggle to answer basic questions without back-and-forth, manual queries or conflicting numbers. In our latest case study with Metabase, we shared how self-service analytics only became possible after investing in clean, well-defined data models. By aligning business logic in dbt on Google BigQuery, Metabase shifted from being just a BI tool to the interface where real exploration happened. The key lesson was simple but repeatable: •  Visualization wasn’t the bottleneck. Trust and structure were. •  Once the foundation was solid, speed and independence followed naturally. This case study highlights our approach at Datum Labs, where clarity comes first and scale follows naturally. 𝗥𝗲𝗮𝗱 𝘁𝗵𝗲 𝗳𝘂𝗹𝗹 𝘀𝘁𝗼𝗿𝘆 𝗵𝗲𝗿𝗲 👇 https://lnkd.in/dvnjTH6Y #DataEngineering #AnalyticsEngineering #SelfServiceAnalytics #DataModeling #BusinessIntelligence #ModernDataStack #dbt #BigQuery #Metabase #CaseStudy #DatumLabs

If you're sharing Metabase across teams but different people need different database access, connection impersonation lets your database handle it. Set up roles in your database with row-level policies. Map those roles to Metabase user attributes for your groups. When they query (SQL or query builder), Metabase runs SET ROLE first. Database permissions take over from there. Keeps one source of truth, no need to recreate rules 👌 Docs ↓