Open-source ETL gets a bad rap in data engineering circles, and sometimes for good reasons. Many vendors never update this type of software, which makes open-source a serious security concern for businesses.
But not all open-source ETL is the same.
Stitch is a cloud-based open-source ETL platform owned by Talend, a cloud integration company with several data transformation tools of its own. So Stitch isn't the type of open-source tool created by a teenager in a dark room somewhere. It's reliable and secure, and it moves data to where you need it to go.
During our tests, though, we struggled with the data transformation element of ETL. OK, so Stitch is free (up to a certain point), and we shouldn't complain, but this type of tool probably won't benefit data-driven organizations that need to pipe data quickly.
However, it's not all doom and gloom. There are things about Stitch that we love, and we're going to tell you about them below. Read on for our Stitch review.
|Trial length||14 days|
|User ratings on Capterra / G2||4.5/5 / 4.7/5|
|Data transformation||ETL, ELT|
|Data source connectors||Fewer than 100|
|Connects to data warehouses? Data lakes?||Yes / Yes|
|Compliance/security certifications||CCPA, GDPR, SSL/TLS encryption and more|
Stitch: What You Need to Know
During our test, Stitch extracted and loaded raw data with no issues. We used several of the connectors to execute this process, but more on that in the next section. The user interface was simple enough, and there was a good amount of documentation on the Stitch website to address the couple of times we got stuck. The entire process required no code, making this tool a wonderful choice for people who don't know how to use Java, Python, C+ or any other complicated programming language. So far, so simple.
The biggest problem we had happened during the transformation stage. That's because Stitch preloaded transformations with basic tasks such as breaking nested structures, which resulted in a few error messages. Transformation on most other platforms is a much simpler process.
The professional interface and relatively easy extract-load protocol impressed us, and Stitch didn't feel like an open-source program — at least not until we realized its data volume limitations. Stitch can't handle large quantities of data, making it unsuitable for most medium- and large-size companies that need to move lots of data to warehouses and other destinations.
We should point out that there's a second version of Stitch — an enterprise version that handles larger volumes of data. We might cover that in a future review, but for simplicity's sake, we'll focus on the open-source version for now.
So exactly how much data can Stitch handle? If you don't use over 5 million data rows a month, you should be fine. Anything more than that, and you'll need to upgrade to the enterprise version. Five million data rows might sound like a lot, but most companies process far more data than this. Again, any company larger than a small business will probably benefit more from another program (or need to upgrade to Stitch's enterprise version).
We loved the security features, something we didn't expect of an open-source tool. Stitch has earned GDPR and CCPA certificates, which provides peace of mind when you're handling customer data and reduces the likelihood of expensive penalties for data governance noncompliance. To our surprise, Stitch is also HIPAA compliant, making it a good choice for (very small!) healthcare organizations.
Finally, Stitch offers ELT capabilities besides ETL so you can transform data into different formats once it's finally reached its destination.
Stitch: Data Source Connectors
For an open-source platform, Stitch offers an unexpected range of connectors (over 130 in total). There are more integrations here than there are in a good chunk of paid-for ETL tools. These connectors help you extract data from sources such as CRMs, ERPs, SaaS apps, relational databases, social media platforms and other data stores. You can then transform the data for analytics and load it to a data warehouse or lake of your choice.
Some of the best integrations include Oracle, Heap, Google Ads, BigCommerce and Amazon S3. For business intelligence, Stitch seems to push users toward Looker, but there are other options. You can find a comprehensive list of integrations on the Stitch website if you're thinking about trying this ETL tool.
Another thing that surprised us about Stitch is its ability to create custom connectors. Again, this is not something you find with most open-source tools, which often come with limited features.
Stitch: Support and Training
Unfortunately, Stitch customer support leaves a lot to be desired. There's no phone or video chat help for users of the open-source platform. So if you're thinking about adding this ETL tool to your tech stack, you'll have to make do with FAQs and tutorials. While this documentation is useful, nothing beats talking to a human being when you're experiencing a problem during the ETL process. If you lack data engineering skills, you'll likely need to reach out to someone at some point, but there's no option to do that here.
If you upgrade to the enterprise version, however, you get far more support options, including phone and video chat support.
Stitch is open-source, which means it's free, but only up to a certain point. You'll pay nothing for up to 5 million data rows. Want to go above that? You'll have to upgrade to the enterprise version for higher data volumes.
In case you are wondering, Stitch's enterprise plan charges users for the amount of data they consume per month. So this is another ETL with a consumption-based model. Choosing an ETL that charges for the number of connectors you use, rather than how much data you consume, could work out to be cheaper.
What Stitch Says
"Stitch rapidly moves data from 130+ sources into a data warehouse so you can get to answers faster, no coding required. Stitch's extensibility features enable support for every data source your team requires."
"Stitch gives you the power to secure, analyze, and govern your data by centralizing it into your data infrastructure."
Stitch Reviews: What Users Think
"Stitch is totally self-serve, there's no relationship with account managers or customer success representatives needed. You can manage your entire ETL system from their UI. For the price and convenience, Stitch is amazing." — Thomas J.
"No info regarding the 'data transfer to destination.' Not a fan of the 'sit and wait and it will eventually happen' part." — An administrator in financial services
"The source library is very limited. There is nothing much we can use compared with Fivetran and DOMO. The product team is not responsive to user requests and takes a long time to build new sources for users. Users are not taken care of well." — Albert M.