F5 Networks modernized its data stack, boosted time to insight, and placed actionable data in the hands of the right decision-makers.
F5 Networks is a Seattle-based application services and application delivery networking company. Because its revenue depends on speed and accuracy, the company is always looking for ways to improve business insights and support data-driven decision-making. Thanks to the data team’s recent migration to a modern, cloud-based data stack, it was able to reduce delays in executing requests from internal stakeholders and improved its internal metric of “code to customer” from weeks to days.
In embracing a modern approach to data strategy, the company had a few ambitious goals:
Improve time to insight by 50x
Move data from on-prem servers to the cloud, with minimal downtime
Boost data culture by getting relevant data and business insights into the hands of more decision-makers
Today, F5’s data team has achieved that and more. At the Modern Data Stack Conference 2020, Lance Hokanson, Sr. Manager of Business Intelligence, outlined the six stages of F5’s modernization journey:
A cloud migration was the first step toward modernizing F5’s business intelligence capabilities. This involved transferring everything on F5’s on-premise servers to a cloud-based platform. The team completed the transfer in a single weekend. “That allowed us to start playing with tools like Fivetran and ask, ‘How do these work?’” Hokanson said. “I highly recommend it because it gives you your own playground space without having to figure it into some big operations model.”
After assessing several options, F5 moved to Snowflake and set up three zones for raw data, curated data and experimental data. All their data is now immediately available for discovery. With its new experimental zone, F5 is encouraging employees to “bring your own data” and explore it in a safe space where insights gained are available to the entire team. This way employees can experiment with and draw insights from all business data relevant to their work efforts, expanding the company’s data culture.
To shorten time to insight, F5 moved to a modern, cloud-based business intelligence tool set. The result: the entire F5 team will have far easier access to better, more informative data. With the BI tool in their legacy setup, Hokanson said, “our on-prem people have to VPN in to use it. It’s difficult to embed content inside tools like Teams or Salesforce — places where people are doing their day-to-day work. We want to make that easier, and moving to a cloud-based BI tool will let us have that experience.” Lastly, because of F5’s modern data stack, data is immediately available for discovery, insights, experiments and data science projects.
This is where the team “flipped our data pipeline on its head” and made the biggest change, Hokanson said. Previously, his analysts would spend a lot of time upfront curating, transforming and enhancing all data before handing it over to decision-makers. They frequently built projects that were rarely used, and they had to re-do lots of work if, say, the sales team realized later that they needed to bring in additional data from Salesforce. Now, F5 uses Fivetran ELT to automatically replicate all data into Snowflake, and analysts only curate data that really needs to be actionable
The company’s new self-service model, using a three-tiered architecture of raw, curated and experimental data zones, means that many requests don’t even need to run through the analytics team anymore, saving time all around.
With the “E” and “L” in “ETL” covered thanks to Fivetran, Hokanson said, “in the next couple of months, we’ll really start to look at the different tools, such as dbt, that focus on code-based transformation and how they can help us modernize in that area.”
The biggest payoff from improving business intelligence at F5 Networks was making data self-service a “way of life.” The company established a data evangelist and customer success program to advance F5’s entire data culture, and Hokanson is even working with HR to add data-related objectives and goals to every job description in the company.
“How do you build data success and analytics success into everything people do?” asks Hokanson. “One of my metrics is how often during meetings do I hear the phrase, ‘Where is your data that supports what you’re telling me?’ When I hear other leaders asking that question, it tells me this world of data is permeating at every level.”