dbt Reviews & Product Details

dbt simplifies the process of building a solid data pipeline by offering a lot of features that would be difficult to implement from scratch. In particular, the SCD2 and incremental functionality helps remove a lot of overhead for developers and makes ongoing maintenance easier. There are also many other features that are great and contribute to a smoother overall workflow. Review collected by and hosted on G2.com.

There’s nothing I dislike about it, but I do have one suggestion:adding a feature for backfilling data (historical loads) will help a lot. Right now this can be done using a macro, but having an inbuilt option similar to incremental would make it much easier and help a lot. Review collected by and hosted on G2.com.

I have been using DBT every day in my project for the last two years, and I can say it is one of the best platforms for creating data models through SQL-based data transformations. DBT’s ease of integration has helped us save effort during code deployment. We integrated DBT with Git so we can work in parallel on multiple tasks across multiple models for a single requirement.

Its ease of implementation also makes it simpler to understand DBT’s benefits and has helped improve our efficiency. The command-line interface and the IDE environment make it easier for us to manage data and execute data models. DBT also has its own test case preparation features, which help us validate both our data model changes and the data being populated through DBT models. If we run into any product-related issues, there is a customer support team that is always ready to guide us through our blockers. Review collected by and hosted on G2.com.

In my experience, I worked with DBT for 2–3 years and noticed that, to use DBT effectively, you need prior knowledge of SQL and Python—especially if you want to take advantage of additional features like Jinja templates. This can be a major drawback for data analysts and developers who don’t already have a solid understanding of SQL and Python. Review collected by and hosted on G2.com.

I like best about dbt is how it brings a clean, developer‑friendly structure to analytics work. It makes modeling and transforming data feel organized and predictable, thanks to its simple SQL‑first approach and clear project layout. I also really appreciate how dbt encourages good engineering practices such as version control, testing, documentation. So the entire workflow becomes more reliable and collaborative. Review collected by and hosted on G2.com.

I dislike about dbt is that some parts of the workflow can feel a bit inflexible, especially when you're trying to customize how tests or models behave in more complex projects. It also relies heavily on command‑line and configuration files, which can become demanding as the project grows. On top of that, dbt doesn’t handle ingestion or real‑time needs, so user often need additional tools to complete the pipeline, which makes the setup feel less seamless. Review collected by and hosted on G2.com.

The way it handles large amounts of data, as well as how it integrates into AWS (S3/Glue) is great. This allows me to avoid building custom pipelines which would have been very time consuming and caused additional headaches and due to its columnar database design, all of my complex query requests are processed in a timely manner which means I do not fall asleep while waiting for results. Review collected by and hosted on G2.com.

Vacuuming Tables… seriously, I have to manually vacuum and analyze tables to keep this thing running smoothly? It looks like 2005. Managing the clusters and nodes is also a pain – it’s not true serverless. If you’re not paying close attention to the costs, they will jump up way too high. Review collected by and hosted on G2.com.

I primarily value how dbt shifts data transformation into a software engineering workflow. By materializing code into tables and views automatically, it lets our team focus on the transformation logic rather than DDL boilerplate. The model selection syntax is incredibly efficient for running specific segments of our DAG without wasting warehouse resources.

Macros and Jinja integration have also been game-changers for modularizing our logic and reducing code repetition. I find the YAML-based unit testing to be a robust way to ensure data integrity before it reaches our BI layer. Between the two, I prefer dbt Cloud over Core because the IDE provides immediate visibility into query results and schema changes, which speeds up our development cycle.

I use it every workday. Customer support was quick and responsive when I ran into issues during the initial setup of dbt with Snowflake authentication. Implementation was also straightforward when connecting it to Snowflake, and once the connection was established, I didn’t have any ongoing issues aside from needing to reauthenticate every day. Review collected by and hosted on G2.com.

The most significant friction point is the authentication lifecycle with Snowflake. The session tokens expire frequently (often every few hours), forcing a manual re-authentication process that disrupts the flow of development.

Additionally, there is a noticeable feature gap between the versions. dbt Core lacks the native, instant result-set previewing that makes dbt Cloud so productive. Bringing a similar "live preview" or integrated results pane to the Core CLI experience would make it a much more viable option for local development. Review collected by and hosted on G2.com.

I like the amount of variety that dbt offers because of the Jinja code and the inbuilt functions. The incremental models are very well built-in, offering lots of capabilities at a layer beyond what's in the data warehouse, like Redshift. The initial setup of dbt is very straightforward, which I find really helpful. Review collected by and hosted on G2.com.

I guess the development cycle of dbt is slower as a result. Writing YAML file descriptions and the actual code for every single model can lead to a slower development cycle. Review collected by and hosted on G2.com.

I threw terabytes at DBT and expected the infrastructure to fail but DBT ran the distributed execution on its own with no intervention by me. The ability to run machine learning directly within SQL is strange but better than exporting to vertex.ai and dealing with cluster management myself. I also do not have to worry about cluster management as I can just write the query and wait for the results which in my opinion is very straightforward thing to do. Review collected by and hosted on G2.com.

Billing is a trap as well as if you run a generic query without a filter the costs jump up right away which can be very annoying. I had to re-write all of my stored procedures because the syntax isn't quite like pl/sql. And I really dislike reading the logs when a model fails and/or errors occur. Review collected by and hosted on G2.com.

One thing that I find impressive about dbt is that it promotes discipline in writing of transformations. It transformed my approach towards the way I deal with my work, as I now think twice before imposing changes. I use it on a regular basis, and it has enhanced teamwork since logic has less difficulty in reviewing and discussion. This has saved time on quick fixes and has assisted us in developing more confidence on outputs that may be shared. Review collected by and hosted on G2.com.

What I do not like about dbt is that there is a huge effect of little errors in the models. Some of them may break down under the pressure of having a few downstream pieces broken when there is a slight change. It is time consuming and can even bring several individuals into the same problem when it comes to debugging those chains. In my case, this retards progress and results in context switching which can be annoying when time lines are near. Review collected by and hosted on G2.com.

I use dbt every day to transform raw data in our warehouse into clean, analytics ready tables and my workflow typically begins in VS Code, where I write sql models, then push them to Git for version control and run them through dbt Cloud. And overall it has also made collaboration between our team members much easier because dbt makes the whole process much more simpler. Review collected by and hosted on G2.com.

It's challenging when one change throws an entire run off track and the error messages are at best, vague. I also feel the need to defend is the handiwork of my contributor to dbt cloud. I have also encountered the overly relaxed strucure and the resulting chaotic command and environment specific configurations. Review collected by and hosted on G2.com.

Thanks to dbt, I no longer have to depend on the engineering team to manage and transform the SQL data within our warehouse. It is the first step for me in organizing, testing, and documenting the entirety of our data models. I appreciate that all of this information is in one place in version control. I can track all changes made and the details surrounding each one. Review collected by and hosted on G2.com.

Troubleshooting complex dependencies and build errors can be a daunting task. There are occasions when a model fails and it is unclear which upstream change might be the cause. While the documentation is really good, I have found digging into a Stack Overflow or Slack thread to be the answer for some of the more obscure problems. I also find the visualization of lineage in dbt Cloud to be cumbersome. Review collected by and hosted on G2.com.