The retailer was struggling to analyze ~160 billion rows on BigQuery. They had unsuccessfully tried numerous expensive techniques to tune the environment, as well as the queries themselves. BigQuery was also the latest disappointment in a 10-year series of unsuccessful attempts to solve this use case.

For years, they relied on Essbase to support their analytics. As data volumes, dimensional complexity and refresh requirements increased, the underlying architecture struggled to meet the performance demands of both analysts and executives. Trying to run Excel on Google Cloud and BigQuery modernized parts of the stack, but fundamental limitations remained.

All previously attempted solutions were severely challenged by the retailer’s dataset. The best-case scenario was day + 2 batch reporting when analyzing 13 months’ worth of data at the weekly level.

Slow queries limited merchants’ visibility into product sales at a store, channel and item level. They needed to interactively analyze sales performance at every level of data granularity. In lieu of this, they were, at best, missing opportunities to maximize revenue and profit margins. In the worst case, they were actually encouraging customers to shop elsewhere.

Kyvos exceeded all stakeholder expectations in just 2 weeks. Kyvos automatically enabled interactive analysis of the retailer’s dataset comprising ~160 billion fact rows, at every level of granularity. Those analyses took place with 48 months’ data at the daily level across Store, Product, Geography and Time dimensions. Kyvos succeeded where no other solution could.