The quality we expect from the software we use continues to soar to new heights. While design will always be a critical input, personalization is equally important. A personalized experience signals respect to the customer, and helps earn their loyalty and trust. This is why customer data is the lifeblood of the modern business. The more of it that can be collected and processed, the more opportunities emerge to personalize and improve the experience.
However, the infrastructure that is necessary to derive value from customer data is non-trivial to build and operate. Much has already been said about these challenges, and the resulting explosion of new data infrastructure products over the past decade that has culminated in the industry’s coalescence around the Modern Data Stack. A key implication of the MDS architecture is the central role of the cloud data warehouse. This has forced adjacent layers of the stack to conform to this paradigm, and given rise to a vibrant ecosystem of high-growth startups.
Surprisingly, one of the most critical components of the stack has yet to be reinvented with this paradigm in mind - the customer data platform (CDP). The role of the CDP is to collect, route, and process all customer data into profiles that can be consumed by downstream systems such as marketing automation, CRM, and product analytics. The current generation of CDPs are “closed” systems that run pipelines and features in a black box. This is acceptable for the marketing teams they were originally conceived for, but becomes problematic for developers. As customer data infrastructure is now so foundational to any company, developers have become more responsible for it.
This is where RudderStack comes in. RudderStack is an open-source customer data platform, built for developers. Instead of storing data and applying proprietary models as current CDPs do, RudderStack conforms to the MDS architecture by running directly on top of the data warehouse. This provides unlimited flexibility to pursue new customer data use cases, and the ability to work with popular tools like DBT for modeling. RudderStack also integrates with DevOps tools like Grafana and PagerDuty, and can be managed within existing CI/CD workflow as developers now expect.
RudderStack also builds upon the traditional role of the CDP by bundling the ability to replicate data from SaaS applications and databases (ELT), as well as the ability to expose data warehouse tables to downstream systems (Reverse ETL) into its product. This compound approach feels much needed in the increasingly fragmented data infrastructure landscape of today.
During our diligence, the customer love for RudderStack was impossible to ignore. Just two years since founding, they have seen meaningful bottoms-up adoption of their open-source that has translated to a growing list of paid customers including multiple publicly-traded companies and well-known consumer brands. This is a remarkable achievement in such a short time period, and speaks to the merit of their ambitious vision.
RudderStack was founded in ‘19 by Soumyadeb Mitra, who most recently led machine learning at 8x8. His time there left him all too familiar with the lack of resources engineering teams have to deliver on customer data projects, and motivated him to found the company to address this.
I am excited to announce that we are leading RudderStack’s Series A, and that I have joined their board. Kleiner Perkins is proud to be partnering with Soumyadeb and team on RudderStack’s journey to provide developers with the open, flexible tools they need to deliver value to the business with customer data.
ー Bucky