Item: Matillion
Rating: 9
Author: Matthew Belo

Use Cases and Deployment Scope

We replicate data between PostgreSQL and Snowflake for most of our core business operations. We were using various home-grown techniques for doing that which took a very long time to complete due to the growing size of the data. We turned to Matillion for help and started first with their original CDC product that was part of the ETL server. That could not handle the volume that we pushed, so we switched to their developing CDC product in Data Loader.

Pros and Cons

Provides seamless, end-to-end replication
Works tirelessly with the customer if there are any issues

The customer service team needs to improve interaction with the customer
Provide documentation to current customers on new features that are added so that we don't have to find out either ourselves or have to search through the FAQ pages

Return on Investment

Reduced our replication time from hours to now seconds
Direct-to-Snowflake is proving to be the most reliable version yet and very resilient

Usability

It is very easy to do what is needed for replication, but working with the ETL product is a bit daunting. Since it is not our primary tool from Matillion, we have yet to receive any training on it so it is not intuitively obvious what we need to do to accomplish many of our plans.

Time to Value

At the beginning of the ETL CDC process, it took us several working sessions with the teams to get all of the different components constructed in AWS, and then to find out it didn't meet our needs was a bit disappointing. With the CDC in Data Loader, though, we were up and running in a couple of hours after getting everything built and permissions established. Much better product. With the direct-to-Snowflake version, a new version can be stood up and snapshotted in less than an hour.

Scalability

The current connectivity is set to provide information primarily in one direction; from simpler OLTP systems to big data warehouses. Allowing the big data warehouses (e.g., Snowflake) to be used as a source for transmission elsewhere would be a big extension to the product line and provide more flexibility in architecture.

Alternatives Considered

Fivetran, Stitch from Talend and Striim

Matillion ran circles around Stitch and Striim both in functionality, setup, and performance. There was no real comparison. Fivetran massively outperforms Matillion in pretty much every facet of the production from setup, maintenance, visibility, and usability. It already has the ability to connect any data source to a destination regardless of database type. Why we chose Matillion over Fivetran is that, for our current needs, Matillion provides us with the functionality that we need and a much more competitive price for a smaller company.

Key Insights

Do you think Matillion delivers good value for the price?

Yes

Are you happy with Matillion's feature set?

Yes

Did Matillion live up to sales and marketing promises?

Yes

Did implementation of Matillion go as expected?

Yes

Would you buy Matillion again?

Yes

Other Software Used

Apache Airflow, Tableau Cloud, PostgreSQL, Snowflake, Monte Carlo, bodo.ai, Keebo

Likelihood to Recommend

The current evolution of their CDC offering is where it should have been in the first place. That offering provides replication direction from the PG database to SF without any intermediate steps. The previous version had too many steps: ready from PG, write to S3, run ETL to push to SF, and there were opportunities for missed transactions. Where they need to grow for us is to allow end-to-end replication from PG-to-PG, SF-to-PG, and SF-to-SF. Those are handled by competitors, so it would help them close the gap.

Replication with Matillion Saves Time

Overall Satisfaction with Matillion