Your organization is ready to jump into data analytics. You’ve chosen your enterprise data warehouse, you’re getting familiar with business intelligence tools, and now you’re ready to begin replicating real corporate data into your analytics repository. Which ETL tool (or ELT tool) should you choose?
We can save you the time and trouble of reading the rest of this post: Choose Fivetran. But if you have a few minutes and would like a little more context, we can show you why Fivetran makes sense. In this post we hold our platform up against some of our competitors, and show you why we shine.
The Very Model of a Modern Data Pipeline
Nowadays it is easy (well, easier than it used to be) for citizen analysts to build dashboards and reports incorporating data from multiple sources, using cloud data warehouses like Amazon Redshift and Google BigQuery and analytics tools like Tableau and Looker.
Making that data available in a single place ought to be just as easy. If someone needs data for a report from a source that’s not already in your data warehouse, they should be able to seamlessly access that data and copy it to your analytics repository.
Doing so requires a point-and-click graphical interface that walks users through the process. It means having confidence that the data moved will be complete and arrive when you need it. And it demands a platform that offers not just a handful of popular data sources, but a broad selection of connectors.
Let’s start by talking about what Fivetran offers. We support 150 data sources and 10 destinations, including Redshift, BigQuery, Microsoft Azure Synapse, Snowflake, Microsoft SQL Server, PostgreSQL, and MySQL. If you want to use a connector we don’t offer, tell us what you need — we add new connectors based on customer requests.
Compliance? We’re happy to sign a business associate agreement (BAA) with customers who are subject to HIPAA mandates. We’re GDPR-compliant. And we’re annually audited for SOC 2 compliance.
Cost? We offer flexible, transparent, pay-as-you-go pricing based on credits for monthly active rows of data.
Let’s see where some other ETL providers stand in those areas.
Fivetran vs. Informatica
Fivetran is a cloud-native data pipeline. Informatica is not. While the company now offers Informatica Cloud Data Integration, the roots of the product are in software that pushed data to old, on-premises data warehouses. Those on-premises products followed the old ETL model, in which transformations were done in the data pipeline, rather than the modern ELT process, where transformations are done in the data warehouse.
It’s like the difference between a Tesla, conceived from the start as an electric vehicle, and a Honda Civic EV. Which would you feel more confident driving?
Unlike Fivetran, Informatica lacks transparent pricing. If you want to know what it’ll cost you, you have to talk to a sales rep. Informatica does offer a 30-day free trial.
Fivetran vs. Matillion
While Fivetran supports 10 data warehouse destinations, Matillion supports just five: Redshift, BigQuery, Snowflake, Azure Synapse, and Delta Lake on Databricks. The platform has just over 100 connectors (or integrations, as Matillion calls them) but not all of the data sources are available for all of the destinations.
Matillion ETL offers good reliability and has a strong set of compliance certifications for HIPAA, GDPR, SOC 2 Type 2, and more. Pricing is similar to Fivetran, with three tiers based on credits and usage. The company offers a 14-day free trial.
Fivetran vs. Xplenty
Unlike most other platforms, you can’t just enter your email address and begin testing Xplenty. You can get a 14-day free trial only after a demo and a conversation with a sales rep. Xplenty supports almost 100 integrations to data sources, and the widest range of destinations among these platforms, including not only cloud data warehouses and data lakes, but also SaaS and analytics applications. The platform is compliant with HIPAA, GDPR, SOC 2, and other certifications.
Xplenty pricing bucks the usage and consumption model that most ETL products use. Instead, it charges based on the number of connections between each source and destination. It’s a great model if you have very few data sources with a lot of volume from each, but data analytics are typically richer the more data sources you use, and Xplenty’s pricing model disincentivizes adding new data sources.
One key feature Xplenty lacks is change data capture for incremental replication of databases. Instead, it uses the less reliable approach of designating a timestamp column and replicating all changes that happened after the last replication time.
Fivetran vs. Stitch
Once, Stitch was Fivetran’s strongest competitor, but the company was purchased by an old-school data integration company called Talend in November 2018, and since then they’ve kept a low profile. As part of the Talend data platform, Stitch offers more than 130 data connectors — almost as many as Fivetran, including a large number of data sources both companies support.
Stitch is reliable and compliant with HIPAA, GDPR, and SOC 2 guidelines. Like Fivetran, Stitch employs a consumption-based pricing model with several tiers depending on monthly row usage. The company caused a minor uproar last year when it eliminated a free tier it used to offer to low-volume customers, but it still offers a 14-day free trial for new customers.
We’ve taken a pretty high-level look at a sampling of data integration/ETL tools, but the devil is in the details. As you do your own hands-on assessment of each ETL platform, you should test it with your company’s live data, make sure you can replicate the volume of data you need as quickly as you need it, and make sure that it arrives in your destination with the right data types and record structures.
Find out if the documentation is comprehensive and if training services or implementation help are available if you need them.
One other technical point to consider: SaaS software vendors periodically upgrade their platforms, and when they do, they may add new tables, add new fields to existing tables, or merge or delete tables and fields. Because cloud ETL tools are software as a service, they’re responsible for making the changes to their platform so your data analytics stack doesn’t break. But what about the far end of the data pipeline — the data warehouse? It too needs an updated schema, lest the whole process break down.
Fivetran automatically migrates schemas in a data warehouse based on changes in source data. That takes some clever engineering, and not every ETL platform has that capability. However, it’s vital to keep your data flowing. Ask any ETL vendor how they handle data source schema changes at the destination.
CaliberMind Onboards Customer Data With Fivetran
Read how CaliberMind spent more time uncovering data insights and less time worrying about schema migrations with an automated process for onboarding its customers’ data.
Something for Everyone
We’ve covered only a handful of more than 100 platforms that claim to do ETL. If (hard as it may be to conceive) Fivetran or one of the alternatives we talk about here lacks a key feature for you, you can probably find an application that offers what you need. But we’re willing to bet Fivetran can meet your needs. We challenge you to find out — sign up today.