As a trusted maker of software platforms and other services, San Jose-based Calix helps communication services providers (CSPs) streamline and grow their businesses.
But when Pete Watridge, Senior Director of Analytics and Decision Science, first joined the company, he realized Calix’s own data operations process was in desperate need of simplification as well.
“When I came into the organization, I asked, ‘Where can I find data?’” recalls Watridge. “And no one could give me an answer. We had islands of business intelligence sprinkled across the organization.”
At Data Engineer Appreciation Day, Watridge detailed how he’s centralizing that fragmented landscape and building a streamlined DataOps organization that’s capable of providing both technical expertise and an understanding of pressing business issues.
Build the Foundations — Then, Use Them
Before Watridge kicked off Calix’s ongoing centralization effort, business logic was scattered across Excel files, and processes were living within individual employees’ computers or even in their heads. The lack of standardization and transparency bred disorganization and lag time, which in turn hampered executives in making well-informed decisions.
“Everyone would take Power BI, directly connect it to Salesforce, and do very specific reporting,” Watridge said. “People often didn’t know other users were doing similar things, so it caused quite a few problems.”
He and his team of 15 knew they desperately needed to implement common platforms built around a modern data stack — while also keeping up with day-to-day work.
“We didn’t have the luxury of building out our foundations first before we delivered value to the organization,” he said. “We had this legacy mishmash of stuff, but we couldn’t tell the world, ‘Wait stop. We’re not doing that any more, you have to wait until we build these foundations.’”
Focus on Business Outcomes
As the team embarked on this centralization effort, they focused on how their new tools and processes would ultimately deliver greater value to Calix. Or, as Watridge put it, “Data is the raw material, and at the end of the day, the outcome for the business is what we care about.”
Calix implemented a modern data stack using Snowflake, Fivetran and dbt, and the company quickly saw the benefits to keeping all data logic in one place. For example, with Salesforce and Oracle data now joined for the first time, his team could produce richer analyses and rely on one single source of truth. Data is now landed automatically, business logic is centralized, and transformations happen more efficiently within the cloud warehouse. Calix’s new data stack is diagrammed below.
The team is building trust by ensuring that data is up-to-date, consistent and correct, Watridge said. At the same time, users throughout Calix can still explore on their own as needed, which is a big plus for his small team. “We can’t drive all of the data innovation within the business, but we can support it,” he said. “So what we’re trying to do is drive data literacy into the organization and give them the tools and hopefully allow them to do the data innovation. And if they do something cool, we’ll bring it back and curate it in our centralized architecture.”
To that end, his team is also implementing a new foundational data reporting architecture, which results in specific data marts that easily feed into visualization tools:
The team is currently defining a new security and user governance approach to guide this process.
“We want to push data and decision-making as far into the front lines of the business as possible,” Watridge said, “so that when people are doing their work, they can get to the data they need right when they’re doing their process and not have to go somewhere else.”
Reimagine Your Human Capital
Watridge expects Calix’s centralization process to wrap up later this year, and he’s already thinking about what kind of talent his data team needs once they reach a steady state post-upgrade. Analysts are now embedded within functional areas within the business, like sales or marketing, and use shared data engineering resources. Data engineers are focusing on building out the foundational Snowflake platform first with all of the data sources, while analysts are migrating reporting to the Snowflake platform and increasing their focus on critical business issues.
More and more, Watridge said, he’s pushing his analysts to act more like project managers, equipped with equal parts technical ability and a domain knowledge of the functional business area in which they serve.
“They need to be very conversant in the needs of the organizations that they’re supporting, so when they’re working with a senior leader, they can have a conversation,” Watridge said. “They can understand exactly what decisions are being made and then build a quick prototype.”
In fact, one notable new analytics hire came from a business background within a services and success team, and he had picked up technical expertise with SQL out of necessity in a past job.
Calix’s approach of combining new, foundational tools and processes with a fresh perspective on the data team’s role within the organization will help it do what ultimately matters — deliver value for the company.
Or, as Watridge puts it, “The point of leverage that we look to support is the point where someone in the business is making a decision. Really, what matters is for the analytics organization to be a trusted partner to the business owners and the decision makers. Everything that we do is to get to the point where we’re as trusted as we can be.”