Snowflake Ventures Invests in Eppo to Bring Experimentation to the AI Data Cloud

Experimentation tools like A/B tests, Geolift incrementality tests and AI model evaluations have become indispensable for product and marketing teams seeking to optimize their initiatives and drive better business outcomes. By systematically comparing two versions of a product feature, marketing asset or user experience, companies can make data-driven decisions that eliminate the guesswork and, ultimately, the risk of costly mistakes. As a result, teams are able to shift from relying on intuition and assumptions to making decisions rooted in concrete evidence — transforming conversations from "we think" to "we know."

The impact of A/B testing extends across various aspects of product development, generative AI use cases and marketing strategies. For product development, it enables teams to evaluate new features, user interfaces and customer journeys to enhance user engagement and satisfaction. A/B testing can also help with emerging generative AI solutions. With gen AI, models can generate complete sentences, images or videos. A/B testing against business-specific metrics from your data platform helps ensure top quality. In marketing, A/B tests can be applied to email campaigns, landing pages, advertisements and social media posts to improve conversion, click-through and overall campaign performance. Marketing leaders can also use quasi-experimental methods, like Geolift tests, to prove the incrementality and efficiency of their advertising spend in contexts where A/B tests are impossible. Companies can achieve compounding improvements over time by continuously running experiments and iterating based on results, incrementally moving closer to their key metrics and business goals.

At Snowflake, we strive to help businesses become as data-driven as possible. Our investment in Eppo, an end-to-end experimentation platform, will help provide users with an A/B testing capability where their data already resides. Eppo was co-founded by CEO Chetan Sharma, one of the first data scientists at Airbnb, where he established a culture of experimentation at a crucial time for the company. At Eppo, he has created a company committed to being trustworthy while helping companies power sustainable decision-making in an ever-changing tech environment plagued by escalating costs.

Following this investment, Snowflake customers can access Eppo’s experimentation platform directly from their data in Snowflake. Experiments — from scenario-planning and feature-flagging to AI segment detection and post-product launch analytics — can be analyzed against the same metrics and data used for business reporting. By providing these capabilities, Eppo aims to help companies tie product development efforts more directly to business metrics and revenue, while enabling a data-driven experimentation culture across product, growth and ML teams.

By integrating directly with Snowflake, Eppo's platform includes built-in data governance, with no data movement, to run high-quality, end-to-end A/B tests on AI-powered features, product growth and lifecycle marketing tactics. Furthermore, Snowflake Cortex will power Eppo’s cutting-edge AI features, such as Contextual Bandits for predictive personalization and Eppo’s CUPED++ statistical engine for reducing experiment variance and runtime. As Snowflake continues to bring experimentation for marketing initiatives to the AI Data Cloud, Eppo will be an integral partner.