Fivetran Reviews

We mainly use Fivetran for ETL.

There's a very good layer that allows you to connect different data sources and establish a very good ETL workflow. You can manage all of your connectors individually, which gives you a very good ability to trace which one of your ETL processes is running and when. The solution also allows you to schedule each of those runs with different cadences, depending on which plan you are on. At the same time, for those runs, you have the ability to hash, unhash, or map any data that might be sensitive or personally identifiable, and that's pretty robust. It's a feature in the standard plan, so you don't need the enterprise plan for that. One very powerful and cool feature is how you can integrate the solution with other transformation tools, such as dbt, or run the transformations within Fivetran itself to create a transformation layer on top of your ETL layer. That allows you to manage the entire ETL workflow end to end from within the platform without the need for any additional tool.

When it comes to managing ETL pipelines, these features are valuable.

Given that Fivetran is a fully managed third-party solution, the customization could improve because Fivetran gives more thought to people who don't want to manage analytics workflows rather than engineers who want to be able to customize pipelines more thoroughly.

I've had experience with Fivetran for about two-and-a-half to three years.

The scalability will directly relate to the plan you're on. It will be very important to see that the provisioning in compute and the provisioning you have to run workflows and jobs will also be directly related to your use case. Managing scalability in that way is a bit complex, but you get a lot of support from Fivetran.

From a customer support standpoint, Fivetran does an amazing job in establishing and helping you customize while ensuring you have the right plan for your data needs. I have full contact with our account executive, and she's super nice and walks us through many of these use cases that will be important in the future and many of the estimations we might need for the workflow we plan on integrating.

Positive

I have used Splunk before, and it's more technical. Splunk allows you to version things and have a good engineering process around your ETL pipeline. It also has very good logging and documentation tools. But now that Fivetran has a better exposed API, it is much more flexible. But Splunk beats Fivetran regarding things like customization, for example. We switched to Fivetran because we reduced the engineering team. We had fewer people to manage all the different ETL processes, so we had to leverage more automated third-party tools, and Fivetran made onboarding new engineers a lot easier.

Fivetran is very easy to set up. It's a matter of having the connection information of each one of the endpoints you want the ETL through. The solution is very comprehensive and easy to set up. It even gives you assistance on where to find the information that you might need, depending on the data source. It has pre-built connectors for the most common data sources, so it also helps you expedite that process if you are not super tech-savvy.

Fivetran has a pricing model that scales the more data sources you add. When you have a lot of workflows and complex use cases, pricing goes down as you use it more. That's a very good feature of its pricing. Fivetran might be a bit pricey for a very limited or small number of data sources that might be easily collected in other forms. The pricing gets better the bigger you are.

Comparing Fivetran to other tools such as dbt, AWS Glue, or Azure Data Factory, the other tools are more robust and allow us to run custom transformations.

To anyone who uses the product, I say monitor. You have a dashboard to monitor your MAR, which is a row-level metric Fivetran uses to gauge how much you are consuming. If you can monitor your MAR closely, you can get a very good understanding of how much data you're moving, and it will allow you to adjust the cadences you might need to get a better bang for your buck. Watch your numbers and try to plug in as many data sources as you need because that will help you with pricing.

I rate Fivetran a seven or a seven-point five out of ten. It's solid and offers a very simple UI and way to get set up. It might fall short for people who want more advanced use cases or people who have knowledge and tools that they can integrate, and that's where things get a bit more technical. But I feel the leap from just getting started with it to getting very technical with it is very large. There's no middle ground.

We are currently using the private model of this solution. It is CDC and we take data from an OLTP system to populate a data warehouse and to populate a reporting system. 

The ease of usability is the most valuable feature. It's very easy and quick to set up. It also has a central hub as opposed to GoldenGate which is one direct interface. For GoldenGate I would have needed three interfaces whereas with HVR I have a central interface that manages everything. 

We have a number of connections to heterogeneous databases and that could be expanded on. One of the databases is an additional database that should be coming online but there is some difficulty with it. Other than this, in comparison to other products it's competing with, including Oracle, it beats them hands down for everything, including the price.

I would like for them to incorporate additional transformations. A valuable aspect of the product is that it does inflight transformations and that could be expanded. 

Their technical support is excellent, better than Oracle. The online support is very good but equally, their support is immediate and willing. Up until now, they have taken our suggestions and considered incorporating them into the next version. 

The initial setup was straightforward. 

The pricing is considerably less than its competitor, GoldenGate. One of the things that we liked about it and that sold us on it, is that we only licensed production and it then enabled us to install the product on any other non-production. Whereas with Oracle, it's in every environment. In the three environments that I have HVR, I only need to license production. Whereas with Oracle I would've had to license all three environments. 

I would suggest not to discount it because of its low price and small company. Most of the technicians are ex-Oracle employees and they have a passion for data transformations and data support. They are very proactive which is a breath of fresh air when you compare it with Oracle support. 

I would rate it 8.5 out of ten. 

We primarily use the solution for business intelligence.

We use it for data publishing, data logs, and data refreshes These services are all very good.

The interface needs to be more user-friendly.

Data security should be improved in the next release.

The solution is stable. We've never faced any stability issues.

The way we are using it, I think the solution's level of scalability is sufficient for us. We didn't really need anything bigger. For us, it's been fine so far. We have about 30-40 users. There's no need for us to increase usage at this point.

We've never really used technical support, so I can't speak to the level of quality.

The initial setup was complex. We had quite a few data sources, so the database replication was quite in-depth and we had quite a few environments that needed to be refreshed. 

Right now, we have a corporate license. I'm not sure of the term of the license or the cost.

We use the on-premises deployment model.

I would rate the solution eight out of ten.