dbt Platform
dbt LabsExternal reviews
199 reviews
from
and
External reviews are not included in the AWS star rating for the product.
Analytics Engineering revolution!
What do you like best about the product?
dbt is a potent tool with lots to explore. The data lineage is fantastic, where you can easily see if a small change brakes a model downstream.
Tests are integrated into it, which we use a lot, specially custom ones. Macros are handy and fantastic resources for controlling tests, functions, environment behaviour, etc.
Building new models is effortless because the analyst only needs knowledge of SQL (and a bit of dbt but like any tool).
Finally, I love the dbt community spirit, the excellent documentation and how dbt is continuously improved.
Tests are integrated into it, which we use a lot, specially custom ones. Macros are handy and fantastic resources for controlling tests, functions, environment behaviour, etc.
Building new models is effortless because the analyst only needs knowledge of SQL (and a bit of dbt but like any tool).
Finally, I love the dbt community spirit, the excellent documentation and how dbt is continuously improved.
What do you dislike about the product?
However, I am worried about the scalability of using one repository. There were two of us when we started using dbt, but we are +10 now and this reflects in running times, release management and CI/CD. We would love to see a bit more support on this.
Another thing is the alerting system, we can set up tests as warnings, but you need to enter dbt cloud on purpose and see the alerts inside the job. We are building alerts out of the box because this doesn't work for us.
Lastly, it was difficult for me initially to adapt from the "old" data stack, but I am 100% dbt converted now.
Another thing is the alerting system, we can set up tests as warnings, but you need to enter dbt cloud on purpose and see the alerts inside the job. We are building alerts out of the box because this doesn't work for us.
Lastly, it was difficult for me initially to adapt from the "old" data stack, but I am 100% dbt converted now.
What problems is the product solving and how is that benefiting you?
We previously had an ETL tool which was very difficult to maintain and contribute to.
This was a huge bottleneck, and dbt has allowed many people to contribute to modelling and building dashboards.
This was a huge bottleneck, and dbt has allowed many people to contribute to modelling and building dashboards.
DBT is a great out-of-the-box solution to huge problems found on data-driven organizations
What do you like best about the product?
I like how it enables analytics engineering (building pipelines) with best practices and has other really useful features (such as documentation and testing). I haven't found a product this complete in other competitors.
What do you dislike about the product?
I dislike the multiple indentation errors I get on YML, I dislike having to deal with separate "packages" for basic stuff that should be included on DBT regular release, I dislike the "feel" of the typing on DBT Cloud. I feel I can't type code with the same speed and flow I have when using VSCode.
What problems is the product solving and how is that benefiting you?
DBT is solving SQL standardization, lack of documentation, lack of testing, lack of a good optimized code across companies, which makes us lose less time debugging stuff or trying to figure out someone elses scripts and upstream/downstream tables.
Overall good developer experience and great ELT solution
What do you like best about the product?
- Git integration and software engineering best practices
- Lineage
- Metric layer
- Tests
- Incremental materialisations
- Quite quick response times for cloud support tickets
- Lineage
- Metric layer
- Tests
- Incremental materialisations
- Quite quick response times for cloud support tickets
What do you dislike about the product?
- Lack of being able to run changes on a dev environment without having to create a separate project
- Lack of explanation of best practices for how developer and deployment schemas should be set up/named
- Lack of explanation of best practices for how developer and deployment schemas should be set up/named
What problems is the product solving and how is that benefiting you?
- We needed a more scalable solution to the python scripts we were running to perform ELT on our data
- Batch processing for redshift data which we then use in our prod systsem
- Batch processing for redshift data which we then use in our prod systsem
Great tool to get more visibility into data warehouses
What do you like best about the product?
Being able to analyse dependencies between data models easily by following the lineage graph. Also, being able to embed tests while creating data models in a straightforward and fast way reduces the tendency to look at testing as an afterthought, effectively making data pipelines more robust.
What do you dislike about the product?
It is still not possible to precisely pinpoint the root cause of a certain error. dbt is able to identify the table where the error happens, but not the exact fields involved in it.
What problems is the product solving and how is that benefiting you?
I used to spend a lot of time trying to figure out the dependencies between data models. With dbt I just have to check the documentations, specifically the lineage graph, and I know in an instant how multiple different data models are related.
The era of shaky SQL templates is over
What do you like best about the product?
DAG execution when rebuilding an entire schema/mart, docs, tests (referential integrity especially), dbt utils and packages (redshift utils especially) and many more.
What do you dislike about the product?
There is little that I dislike, IDE version 2 has improved, one can still over-write default macros in order to achieve custom functionality.
What problems is the product solving and how is that benefiting you?
For instance, my first three months with tool were spent on a migration of all kinds of different sql processes which accumulated over 5 year within the company, now all simplified and unified under one repository. The main problems we solve are staging transformations (used by legacy reporting) and building dimensional schema for self-service analytics as well as embedded analytics.
Really helpful for cleaning up data
What do you like best about the product?
The product is really good, but the support documentation and community surrounding the product are really helpful in implementing the product.
What do you dislike about the product?
It takes a lot of training to get competent with the product - not unexpected.
What problems is the product solving and how is that benefiting you?
Orchestrating our data into usable models
dbt review
What do you like best about the product?
Highly functional SQL transformational tool. Simple to start, but with adequate levels of complexity available to add if necessary. Good documentation.
What do you dislike about the product?
Nothing comes to mind as the product provides a great solution
What problems is the product solving and how is that benefiting you?
Internal data transformation in a cloud warehouse. Very few other products accomplish this as well.
Great tool - fast development experience
What do you like best about the product?
Time saving of having fully integrated tool - faster development experience than managing my own IDE, query tool and command line/git
What do you dislike about the product?
git branches that are deleted still stick around for some reason, dark mode glitches and becomes light mode on some tabs, SQL preview window glitches and freezes
What problems is the product solving and how is that benefiting you?
Collaborating on data pipeline design and creation of data assets
dbt is a great tool, I wouldn't want to be without it
What do you like best about the product?
Most helpful is how dbt sorts out dependencies for you. Using references in your models, you can tell immediately if a column name change has caused a problem downstream. The DAG graph is fantastic, particularly in the dbtCloud IDE as it updates immediately. Adding basic tests is easy, leading to increased levels of confidence. I know I can run and rerun things without causing problems. If tests pass we're good to go. Freshness tests help make sure that nothing is run if upstream data is not present. This has helped us considerably recently. And exposures, linking to external reports that use the data, is really useful.
What do you dislike about the product?
Creating a job in the dbtCloud interface and having it default to an hourly schedule. We've only had one minor mishap, but that could be costly with BigQuery. Unless I've missed a setting, that should be off by default.
I don't use the Web IDE. Personal preference, I don't want to switch editors. But the lineage graph is awesome, and seeing the compiled alongside the editor is a great feature.
I don't use the Web IDE. Personal preference, I don't want to switch editors. But the lineage graph is awesome, and seeing the compiled alongside the editor is a great feature.
What problems is the product solving and how is that benefiting you?
Getting data under test so that we know the data makes sense, not just that it has been loaded. The freshness tests have increased confidence that reports are ready to send from our BI tool. We have a history of many similar queries doing the same work. We're using dbt to find and eliminate that duplicated work.
UI based Version Control and ETL Orchestration
What do you like best about the product?
DBT Cloud lets users set up transformations and orchestrate code runs from a UI-based IDE in the cloud for users who aren't as familiar with using the command line. By leaving engineers to do the coding rather than spending time digging through documentation, our team could reach functionality faster and with less fuss than expected. The ability to test in a safe environment by default before you push your changes to live makes version control easy for users not familiar with git principles.
What do you dislike about the product?
The cloud-based IDE has a long-ish startup time (30 seconds or longer) and locks you out of editing while it syncs after you save a file. This leads to a little bit of frustration when making minor edits across multiple files. Some other minor frustrations related to the Cloud IDE come to mind, like how column names aren't sensed while you're typing, leading the user to write out column names that could be easily polled during the lineage population. In addition, the testing documentation was a little confusing, leading to me trying different things to get it to work. In the end, I learned that you have to prefix the name of a test model with "test_" to get the orchestrator to see the model as a test. However, after a couple of speed bumps, most issues I have turn out to be minor and do not impact the product's overall usefulness.
What problems is the product solving and how is that benefiting you?
DBT Cloud has allowed us to migrate from one ETL partner and re-create our transformation files with version control and replicability in mind first and foremost. Additionally, for users who aren't familiar with command line interfaces, we can spin up new users and share work easily while everyone can push to the same deploy environment.
showing 51 - 60