Customer Story
Background
Apollo.io, a leading B2B sales intelligence platform, faced significant challenges with their search functionality, particularly when dealing with their two core data models: contacts and accounts. The platform maintained a synchronization system where account information was replicated to contact records to enable faster querying. However, this approach began to show its limitations, especially when handling accounts with large numbers of contacts and frequently changing fields.
Challenge
The primary challenge emerged when dealing with accounts that had approximately 150,000 associated contacts. Fields like ‘account owner,’ which could change frequently, created a significant burden on the system. The platform couldn’t efficiently queue millions of Elasticsearch reindex operations, leading to incorrect results, missing data and underestimated Total Addressable Market (TAM) calculations for their enterprise customers. This issue generated approximately 30 support tickets monthly and became a substantial roadblock for Apollo’s internal sales team and enterprise customers.
Solution Implementation
Apollo’s engineering team partnered with Siren Federate to address these challenges. They worked closely with Renaud Delbru, Siren’s founder and the architect of Siren Federate, who provided extensive support in analyzing Apollo’s infrastructure and developing a bespoke solution. The implementation process, while met with initial skepticism about making foundational changes to Apollo’s search architecture, proved to be transformative.
“This partnership has been both a performance milestone and an inspiration for us,” added Renaud Delbru, Co-Founder and Chief Scientific Officer at Siren. “Siren Federate was designed to handle mission critical, large-scale searches and the Apollo performance validates our approach.”
Technical Innovation
Siren Federate developed a new type of aggregation that enabled custom drill-down views based on their joint technology. The solution was particularly noteworthy due to its unprecedented scale – handling an Elasticsearch cluster of 350 nodes, pushing the boundaries of what was previously thought possible in information retrieval. The implementation required maintaining stability and reliability across a complex system involving massive concurrent user data and numerous machines.
Results and Impact
“The impact on our customers was immediate,” shared Matt Curl, COO at Apollo. “Our products are now faster. Customers can use Apollo products in even more precise ways to filter and target, allowing sales and marketing teams to have unparalleled results.”
The implementation of Siren Federate delivered remarkable improvements across multiple metrics:
- Search Performance: Average search time reduced to 1.2 seconds, down from the initial 5-7 seconds when first implementing the solution.
- Comprehensive Results: Federated searches now return 50% more results, approximately 400,000 additional contacts per search, providing users with more complete data.
- Support Issues: Support tickets related to search problems dropped from 30 per month to zero, eliminating the need for technical issue explanations to customers.
- Platform Coverage: Despite initial conservative goals of implementing the solution for only their 50 largest customers, Apollo successfully transitioned 100% of their joint traffic to the Siren Federate solution.
Fresh from Griffin Brodman’s internal deck
Business Benefits
The implementation provided Apollo’s users with the flexibility to choose specific data views, offering predictive indicators for both churn and buyer intent. This capability directly impacted the bottom line for both Apollo and their enterprise customers. The solution also allowed Apollo to clean up their codebase by removing the old “fake join” solution, leading to a more maintainable system.
Future Outlook
With the search platform issues resolved, Apollo can now focus on moving their product forward and deepening their integration with Siren. The successful implementation has not only solved immediate challenges but also set new standards in investigative search capabilities, scalability and performance. The collaboration between Apollo and Siren continues to evolve, promising further innovations in enterprise search technology.
