Productboard is a customer-driven product management system, which enables companies to leverage customer feedback and data insights to fuel innovation, and ultimately, deliver products that customers will love.
For a few years, the company worked with data consulting agencies, but things weren't working out. Productboard was using Keboola, but they weren't sure how to get the most out of it. If the team was going to tap into the potential of these innovative tools, they knew they needed an in-house expert.
We chatted with Matus Pavliscak, Senior Data Engineer at Productboard, to see how he helped Productboard use Keboola to enhance their business.
Productboard's hunt for their expert ended in late 2018 when they hired Matus as its first data engineer. From the get-go, Matus began rebuilding all the company's data pipelines and reporting systems in Keboola and GoodData.
While he is a data engineer, Matus focuses on the product side of analytics, working hard to understand internal customers’ needs and figure out what data issues they need to solve.
This role wasn't Matus' first encounter with Keboola, as he had used Keboola at his previous job at Mall Group and his side projects: “I've been playing with Keboola on hackathons and used it in some of my pet projects, but I didn't have many business use cases for it. That’s when I realized I needed to find a company that uses Keboola so that I could see it in action,” he explains.
At Mall Group, Matus was a part of the team that worked on the transition to Keboola. This, eventually, removed many bottlenecks and brought data to every employee.
When Matus first worked with Keboola, back in 2015, he knew little about ETL tools, so he had little basis for comparison. However, he was immediately struck by the convenience of the platform.
“It was pretty straightforward for me to onboard. Even back then, the user experience and Keboola’s support team were exceptional,” claims Matus, "Also, the documentation and the APIs and everything was built well. So it was easy for me to start."
Matus found Keboola had an easy learning curve. Before long, he could start onboarding other people and soon began teaching others how to use it.
When asked how he would describe Keboola to newcomers, Matus realizes he can’t pigeon-hole it too quickly.
“I'm not sure if I would define Keboola as an ETL or ELT tool,” he laughs, “I would say it's a Data Ops and automation tool. I wouldn’t use the term ETL because Keboola gives you the flexibility to design your data workflows as you need. It's also a tool for collaboration, as you can share your queries with others, create sandboxes, create data catalogs, and things like this."
At Productboard, Matus has been fortunate to enjoy much more time exploring the capabilities of Keboola. He recognizes the advancements since his first experience and is a big fan of the platform's versatility.
While it’s not a cloud collaboration tool, first and foremost, he acknowledges just how easy it is for people to send assets, such as links to transformations, for example.
Also, Matus explains Keboola is heavy on the extractors and writers as well. Because of that, users don't need to worry about keeping them updated, or creating new ones, as a lot of them are already existing. But even with a missing app, the job can simply be done with a generic extractor and generic writer. Here’s how Matus sees them:
“Even though it’s quite complex to understand generic extractors, once you write one or two of them, you can easily work with any REST APIs to build a simple extractor in under an hour.”
Indeed, with its deep integration capabilities, Keboola is proving to be an efficient tool for Matus and his colleagues at Productboard.
Typically, the Productboard Data team works with multiple sources, such as marketing tools, CRM dashboards, event tracking tools, engineering tools, and analytics software.
Matus explains that one of the most convenient uses of Keboola is the ability to write your components whenever you’re working with a new source or new destination. Being able to customize the pipeline helps the team streamline data analysis.
On the more technical side, Matus asserts that Keboola is excellent when working with APIs. This compatibility is a boon whenever the team needs to perform bulk updates or repetitive tasks, as they can quickly write their scripts.
In effect, Matus and his team developed an entirely new review process in GitLab using the APIs. It was easy to automate the process and transfer code back to Keboola to use as the production and development environment.
Another feature that Matus uses for his work is transformations.
“Transformations are quite convenient with the Snowflake as a backend for SQL queries. Keboola also provides Snowflake as a data warehouse, so if you don't want to spend much time learning and setting it up correctly, Keboola makes it easy to create your environment in a couple of clicks. With that, we didn't have to worry about these things while setting our pipelines at the beginning.”
After several years of working with Keboola, Matus now has the experience and knowledge to put the tool to good use at Productboard. He believes the automation features are one of the best aspects of Keboola.
"If you know where to go and how it works, you can simply authorize the extractors, build transformations, and have data cleaned and modeled. You can build things in hours, or even minutes, rather than days."
Whereas other standard ETL tools are limited in what they can do, Matus and his team find that Keboola has greater functionality, enabling them to do more and perform tasks faster, with no need to talk to developers.
Although the advanced capabilities of Keboola are exciting from a data engineering standpoint, Matus is quick to highlight just how user-friendly the tool is—even if you don’t have a lot of technical know-how.
When he joined Productboard, Matus worked alongside the team's marketing analyst, who had little to no experience working with ETL data tools. However, with Matus there, it didn't take long for the marketer to handle the software.
Matus remarks how Keboola has simplified many tasks for the marketing analyst. He no longer needs to undertake the tedious processes of extracting CSV files from different tools every day and then figure out how to mash them together into a single spreadsheet.
“It was quite easy for him to jump in and understand Keboola,” claims Matus, “Now, he is building his orchestrations and transformations, which saves him plenty of time for other, creative initiatives.”
Productboard’s Data team has grown since Matus joined. They are now a team of 7 data engineers and data analysts who use Keboola daily.
They are always on the lookout for new talent - right now they’re hiring a data engineer.
Most importantly, Matus believes that Keboola is a driving force for business growth at Productboard.
"Keboola helps us to bring data inputs from various sources into a single database, and that can be used in our analytics tools like Looker, where we have company-wide dashboards with goals that can be shared within our whole organization."
With this holistic view and real-time data analytics system, everyone in the company can get quick insights into customer engagement, performance, and whether they are on track to hit their targets.
Productboard also has a "playground initiative" where team members can use cleaned datasets and invite others to work on their data projects. With this company-wide focus, Keboola is at the heart of a growing culture of innovation and data-driven decisions at Productboard.