Fivetran Data Movement Platform

In my current project, the specific use case of Fivetran  is bridging the gap between siloed data, allowing me to extract and have it in the same data warehouse. Fivetran  works with its seamless connections to sources and destinations, helping to avoid reinventing extraction logic from scratch. Fivetran already has data models, so it pulls in data quickly, providing it all in the same data warehouse.

I've worked extensively with Fivetran, mainly used for extraction purposes, and I've worked with the transformation element in it as well. Fivetran not only has built-in connectors but also provides SDK connectors, allowing us to develop our own connectors in an easy manner. I don't have to write raw Python scripts or dumping scripts; it offers straightforward examples and guidelines, making it much simpler to develop custom connectors inside Fivetran. We've been able to develop many custom connectors as well, which is unique and beneficial for having everything centralized instead of having those connectors located elsewhere.

One of the best features by Fivetran is its clean, simple, and intuitive UI. It includes a transformation section where I can deploy my DBT queries and scripts. It also supplies good tracking capabilities for billing estimates and user permissions, allowing for customization to the desired level. The number of connectors it has remains a standout feature, and within connectors, the options available are very helpful. Although it sometimes appears static due to its built-in nature, it offers good flexibility for data transformation and caching, which I appreciate because it saves us extensive script-writing time.

The experience of using SDK connectors can be improved, as it lacks computational power, which is an area that needs enhancement. Additionally, for some connections, I want more flexibility during ingestion, specifically for transformations needed beforehand. When we tried to use a built-in Oracle connector, it didn't allow for the tweaks we needed, which led us to the SDK connector route and caused delays in development.

I've been working for almost four years in this field, and it has been a really great learning and challenging opportunity, allowing me to work on various different stacks and to learn and grow extensively.

In my experience, Fivetran is stable with very few instances of downtime or reliability issues, and overall, I am very happy with it.

Fivetran's scalability has been tested effectively, and it has been working well for our organization's growing data needs. However, the performance with SDK connectors could be improved, as it took around 10 to 12 days to conduct historic ingestion for just two years, which I find unsatisfactory.

Customer support from Fivetran is quite good; it's really nice and responsive. I am very happy with my experiences in reaching out to them.

Positive

Before Fivetran, we used Skyvia , which is a no-code ETL tool. I also considered Data Factory as an alternative. However, I found Skyvia  lacking in maturity and Data Factory better suited for enterprise solutions but limited in connectors. Scripting solutions demanded more work and included additional costs, which led to my decision to switch to Fivetran.

My experience with pricing, setup costs, and licensing has been satisfactory; however, I notice that as row numbers increase significantly, it can get a bit more expensive. Overall, it's fairly reasonable.

We see fewer employees needed due to Fivetran's efficiency, which translates to money saved, as employee wages are much higher compared to the costs associated with the software itself.

Fivetran provides time savings, cost reductions, and improvements in data quality. For example, in working with Shopify , we faced numerous errors and challenges with in-house complicated connectors. The time directly correlates to cost; when the time investment decreases, fewer developers are needed, which means significant savings. Generally, we have seen very few instances of data quality issues since implementing Fivetran.

Before selecting Fivetran, I evaluated other options including Stitch , Rivery , Hevo , and Skyvia among others.

My advice for those looking to use Fivetran is to rely on its documentation, which is very straightforward and accurate. Avoid depending too much on AI for Fivetran issues; the dedicated documentation suffices for most inquiries. If problems arise, Fivetran's support team is quick to assist. I rate Fivetran 9 out of 10.