arrowCustomer Stories

Fi Money builds smarter features with warehouse-native analytics

Before

PMs and designers wanted experiment results quickly

Data analytics team bandwidth limited scalability and visibility

Rapid growth required better data residency and cost management

After

70% reduction in total cost of ownership

Product teams track features and funnels independently

Reduced time to insight with targeted experiments

Improved compliance and cost with Snowflake integration

About Fi Money

Fi Money (Fi) is a fast-growing neobank in India that helps customers manage their personal finances, from creating savings accounts to tracking spending and investing in mutual funds. Built for mobile-first consumers, Fi’s intuitive product design is powered by advanced analytics behind the scenes. With over 3 billion product events generated every 15 days, data and instrumentation are central to how Fi ships every feature.

Scenario

Product managers (PMs), product marketing managers (PMMs), and designers at Fi constantly experiment with funnels and rely on data teams to track and report on their effectiveness. The previous platform required custom user segments to be tracked for each experiment, and serving these data requests manually required data teams to devote hours of manual work, which strained resources and detracted from more strategic projects.

When evaluating a replacement solution, the team found that many self-serve product analytics solutions were suboptimal for Fi’s needs: many required events and related data transfers out of the data warehouse, compromising control and visibility. FinTech platforms in India have strict compliance and data residency requirements, and Fi needed data to remain within the warehouse to continue to be audit-proof.

The team was also aware that many product analytics tools were prone to runaway compute costs resulting from scaling event volumes and suboptimal compute-cluster configurations related to size, idle times, and other factors. Fi needed a tool that enabled self-serve funnel and experiment tracking while remaining compliant and cost-friendly.

Solution

Fi adopted LaunchDarkly Product Analytics and leveraged its native Snowflake integration. This warehouse-native approach met many of their key requirements, including:

  • Data remains in their Snowflake warehouse, helping Fi achieve its compliance and audit requirements.
  • A single source of truth built on one data copy that serves both analysts and the platform.
  • A fast setup enabling the pilot to go live in under one week.

After the successful pilot, product managers and designers gained a new level of visibility into Fi’s product analytics. Using the self-serve LaunchDarkly Product Analytics interface, non-technical users at Fi began building their own funnels, filtering by app version or user segment, and tracking variations.

/ /

The true value of LaunchDarkly is that our PMs, PMMs, and designers can look at their funnels of interest and their events of interest almost in a self-serve, do-it-yourself manner, without getting blocked by analytics, prioritization, and bandwidth.

Auro Lakshman Tadury

Associate Director, Program Management, Fi Money

Fi also connected experimentation cohorts to its customer engagement platform via shared segments. When a test is launched, the target group can be automatically sent to marketing for differentiated messaging, closing the loop between product delivery and personalized follow-up.

Results

By transitioning from their previous platform to LaunchDarkly, Fi Money reduced its total cost of ownership by 70%. They eliminated runaway analytics costs while giving product teams faster, self-serve access to insights. With compliance assured and compute costs made predictable, Fi Money’s teams now run more experiments and act on insights without delay.

/ /

When a product is launched, or a certain screen variation is released, the product manager immediately wants to see how it’s doing in real time. If a feature goes live today, by 9 AM the next day they’re already looking at that feature: how it’s faring, how many users saw it, and what the conversion looks like. They feel much closer to the data.

Auro Lakshman Tadury

Associate Director, Program Management, Fi Money

One standout example came from the US Stocks team, which manages features for investing in the U.S. stock market. A designer noticed that users were spending far more time on a swipe-through screen than expected. That insight led to follow-up research calls and design changes that addressed user concerns directly on the page. Without granular, self-serve analytics, this issue might have gone unnoticed.

Proximity to data has driven more momentum across the product teams. PMs are running more experiments, tweaking flows, and checking results independently. The analytics team has shifted its focus from reporting to strategic analyses.

/ /

What the analysts can do now is spend more time thinking about the nature of experiments and the exact hypotheses to test, instead of just trying to produce numbers.

Auro Lakshman Tadury

Associate Director, Program Management, Fi Money

More case studies