How Productboard has Increased Data Team Productivity by 80% with Select Star

Productboard is a customer-centric product management platform that helps companies efficiently develop and deploy their products. Over 5,000 brands rely on Productboard to build roadmaps, prioritize features, centralize feedback, validate ideas, and integrate workflows all in one place to get clever products off the drawing board and into the market.

80%
Productivity gain for data team
2x
Faster data analyst onboarding
testim icon

Industry:
B2B SaaS
Company size:
100-500 employees
Integrations:
This is some text inside of a div block.
Integrations:

Challenge

Struggling to find data across different domains

As a company that helps customers make smarter, more streamlined decisions on software product development, Productboard needed to do the same—and becoming data-driven was the first step. To support their growing analytics effort, Productboard needed a centralized data warehouse to store data that was currently siloed throughout sales, marketing, product, and engineering. To that end, Engineering Manager Adrian Toman helped lead the creation of a data warehouse to serve as a single-source-of-truth for everyone at Productboard—but something was missing.  

“Since we’ve grown, the data warehouse has too, and now people don’t always know where the data field is coming from,” says Adrian Toman, Data Engineering Manager. “People simply know their (business) domain, but if they want to pull data from other domains, it’s hard. They need to know where to look and how to access it.” 

Overburdened data engineering team

When users couldn’t find what they wanted, they turned to Adrian and his five-person engineering team for answers. “We were stumped by questions and were spending significant time looking for answers,” explains Adrian.  As the go-to team for data inquiries, they are the primary resource for users with data requests, but they were looking for datasets and reports that weren’t necessarily created by their team. 

Without a way to easily understand data lineage and provenance, they had to painstakingly make the connections themselves. They were fielding as many as 50 questions a week through Slack, and it was becoming a major distraction for Adrian’s small team. The hours spent tracking down ad-hoc questions and context slowed the team down and prevented them from working on other important initiatives.

Opaque data pipeline

To support the company’s increasing data consumption, Productboard was hiring data analysts, and it was up to Adrian and his team to get them up to speed on the data architecture. 

“Training new data analysts and our citizen data analysts throughout the company about our data model was difficult mainly because we have quite a lot of tables (more than 3,000) in our data warehouse. We are combining the data from multiple domains and tools to create our data mart, and explaining this model and process took a lot of time. As a result, data analyst onboarding was taking 2-3 months in most cases.” notes Adrian.

Productboard needed self-service capabilities so that employees could find the information and answers they needed across domains without getting lost or leaning on the engineering team. What they needed was a data discovery tool with a robust and fully-automated data catalog.

testim icon

“The choice was obvious; Select Star was one step ahead in all the important categories compared to other vendor solutions on data cataloging. It fits our needs very well,”

Adrian Toman

Data Engineering Manager, Productboard

Solution

Select Star for data cataloging

As engineers, Adrian’s team wanted to have their documentation linked to their data models to develop their ETL pipelines—and typical documentation tools like Confluence or Google Docs weren’t good options for them. So the next consideration was if they should build their own custom data catalog. But doing so would have been a massive undertaking requiring more time and resources than Adrian’s team had available. 

So they opted to evaluate existing data catalog tools in the market instead - both open source projects and vendor solutions.

It had to meet certain criteria: easy to maintain, priced reasonably, and user friendly. The catalog also had to connect their existing data stacks, including Looker and Snowflake.

Toman’s team went on all-hands-on-deck to research all the available options and compare their features and pricing. Three catalogs made it to the final round of deliberations: two commercial, and one open source. The open source option didn’t include the Looker to Data Warehouse lineage, and the hosting and upgrades that would be required to have data lineage would be a major time sink for the team. When they compared the two remaining commercial products in terms of the features that mattered most to them, they had to make their decision.

“The choice was obvious; Select Star was one step ahead in all the important categories compared to other vendor solutions. It fits our needs very well,” says Adrian.

During the trial period, Select Star automatically cataloged all tables in Productboard’s Snowflake data warehouse and transformed how Adrian and his team work with Productboard data.

Furthermore, as more data consumers at Productboard start getting onboarded to Select Star, the main use cases became the following:

1. Fill in Documentation Gaps

The engineering team first used Select Star’s data catalog to identify where the data warehouse lacked necessary descriptions and metadata, so they could efficiently locate and correct those gaps. Now that their data warehouse is well-documented, it serves as a rich information resource that facilitates data discovery and analysis for everyone at Productboard. Additionally, Select Star has added automation to the documentation management process by pulling in existing documentation and suggesting missing documentation automatically based on lineage.

2. Understand Dashboards with Underlying Data

Users no longer have to wonder where the data in their Looker visualization tool came from. Select Star's automated column level lineage makes it simple for Productboard to track metrics from dashboards back to their source data, providing important context and clarity into what metrics actually mean. Select Star’s lineage also makes it easy to understand what dashboards are downstream of which tables, and do impact analyses for potential changes to the data model.

3. Work From a Single-Source of Truth

With the data catalog serving as the definitive entrypoint into data, Adrian’s team doesn’t have to waste time looking for specific data and sideline other priorities during the search. They have a single-source-of-truth, plus a roadmap to find exactly what they want. 

4. Expedite Data Onboarding

New analysts get up and running significantly faster because Select Star makes the complexities of the data warehouse easy to understand. Productboard is putting fewer resources into onboarding, but seeing new hires ramp up in less time. 

5. Enable Self-Service Analytics

Across the company, users are starting to embrace self-service analytics instead of always relying on Adrian’s team to produce reports.

“When we’re answering data questions on our internal Slack channels, our first thought is ‘Can we use Select Star to answer this question quickly?’ and then point people in the right direction to look for it in Select Star.” says Adrian.

testim icon

Select Star is the best product for data discovery today. We evaluated multiple solutions, including open source and established vendors, and I'm glad we found Select Star.

As Productboard grants data access to more business stakeholders including product managers and operations analysts, Adrian envisions Select Star taking an important role of making it easy to find insights of domain data and act upon them. It also allows his data engineering team to remain at a similar size even as the company (and data warehouse) grows rapidly. 

Result

Engineering team estimates recovering 80% more time with Select Star

Productboard is already seeing so many positive changes with Select Star. The most obvious improvement is how they dramatically improved efficiency from the reduced number of questions in the data team Slack channel. “We got 30-50 questions before, and now it’s close to 10 a week—leaving our data team with more time to do higher value work.” 

Toman projects that his engineers will be more productive than ever now—with up to 80% more time to allocate to other important projects. At the same time, they can now quickly onboard more data analysts to keep data-driven insights flowing to the company’s growing workforce. 

Unlock the full context of your data

Get Started
Ring