dbt Labs on dbt: Why this analyst is obsessed with dbt Insights

I’m Rachael, one of the data analysts at dbt. While I’m always excited when we ship new features, I’m not always on the frontlines of the folks dogfooding them. With our fantastic team of analytics engineers, it’s rare for me to get in all the weeds of goodies like snapshot improvements or the latest on iceberg.

As an analyst, I’m often coming in after the models are built, usually navigating what's available and stitching together what I need to leverage for business decisions. But even with the best models, discovering the right data and insights is time consuming. Gaining confidence that I have the specific data I need for a business decision involves the following:

• Identifying which models I need
• Identifying which fields in the models I need
• Identifying which values in those fields I need

Finding the answer has always been a multi-faceted workflow across multiple tabs at every company I’ve worked at: ping ponging between combing documentation in one place (a catalog, a Confluence Wiki, comments in past reporting, etc.) and writing queries to dig in deeper somewhere else (Hex, Mode, Dbeaver, etc.). After a while, you take the multi-tab existence for granted.

…Until now.

Could we be talking data discovery and querying, all in one place?!

Indeed, we are! One of the new features dbt has built is dbt Insights: a place to actually query data from within dbt's governed workspace, without tab hopping. It goes beyond a standard SQL interface with previews and charts to tie your data exploration to all the metadata, semantic layer logic, and documentation, in the same environment as your engineering team.

Likely, you already have a place you are comfortable doing your data querying, but here are a couple dbt Insights tricks that have been leveling up my workflows:

• ➡️ From dbt Catalog to queries, instantly: I can hop directly from a model’s documentation to live querying that model in the blink of an eye…no switching tabs and typing out needed!

• ➡️ Everything in-line: Likewise, I can hop back to dbt Catalog just as easily, or even explore it side by side (my personal preference) to ensure I'm querying trusted, governed models with full context in view: metadata, documentation, and lineage.

The dbt Insights features that have transformed my workflow

My dbt Insights love letter is about to get sappier:

• ➡️ AI efficiency boosts: As a staff analyst, I’m often jumping into new business contexts. When I’m working with an unfamiliar area and not as sure on where to begin, I can chat with our context-aware AI, dbt Copilot, to get some ideas of where to start looking across both dbt Catalog and saved queries. This has proven much faster and customized than my prior catalog search process of throwing out search terms and hoping to strike gold:

• ➡️ Pre-populated dbt Semantic Layer syntax: Another big time saver is that I no longer have to look up or try to remember the nuances of how to query the dbt Semantic Layer, which frankly I’m not in everyday and don’t have memorized. I can just look up metrics, like MAU, and dbt Insights will auto-populate a good starting point for the syntax automatically. Yes please.

• ➡️ Understanding what SQL the Semantic Layer syntax actually represents: I always have trouble remembering what I need to do to see what is behind a semantic model. The product team knows I have mentioned this before! 😂 With dbt Insights, I finally get my pretty out-of-the-box view of what SQL my metric queries are exactly compiling to. No more guessing or bugging my team.

• ➡️ Some extra Fusion goodness coming next: You may have heard of all the Fusion perks we’re adding to the dbt engine, like CTE previews. Plans are in the works to enrich dbt Insights with some of this as well.

Try dbt Insights: a better way to query and explore trusted data

Above, I highlighted some of the reasons dbt Insights has won me over for discovering and querying data (beyond just the fact that I happen to work at the company that makes it).

If you’re already working in dbt, dbt Catalog and Insights can help you move faster and feel more confident, without ever stepping outside the trusted, governed layer your data team has built. It makes it easier for more analysts to explore models, run queries, and contribute without relying on messy workflows or one-off help. If you’ve been the “go-to” dbt person on your team, this may help open up dbt to the rest of your team to move faster, together.

This is just my personal take, and I’m always looking to learn more about other analysts’ workflows! Ping me on the Community Slack with your thoughts if you give it a spin, or let me know what your querying tips and tricks are if you’ve found productivity elsewhere.