As Databricks grew, so did its support volume. Its customer support centers grappled with over 14,000 cases each year, but the company was still relying on reactive approaches to customer support analytics: namely, survey analyses like CSAT. According to data from a TSIA webinar, 81% of companies have the technology for CX analytics, but 38% of these are only using basic survey tools. For such companies—including Databricks—customer support is thus a backward-looking endeavor.
“By the time you wait for CSAT to be the deterministic factor to understand what the customer experience was like, it’s too late,” says Tanvir Kherada, Senior Director of Technical Solutions at Databricks. “The damage has been done. But if you look at the ticket’s lifecycle…there’s a lot more you can do to salvage the situation and provide the best-desired outcome for the customer by intervening at the right time.”
Databricks resolved to adjust its approach to customer support in order to do just that: intervene at the right time and thus improve customer satisfaction and outcomes by addressing concerns more quickly. The company initially built an in-house sentiment analysis mechanism, which it ultimately abandoned. “It was essentially sub-optimal,” explains Tanvir. “It rendered a lot of false positives.”
Databricks wanted an AI-based tool that could analyze and process both structured and unstructured data—i.e., ticket metadata combined with customer messages, customer comments, and ticket updates—in order to identify customers’ most urgent situations earlier in the support process.
In addition, Databricks needed a more comprehensive understanding of the issues its customers frequently faced. “[We were looking for] something that would identify trends,” says Tanvir. “Are there consistent tickets coming from the last couple of days, where they are having specific issues that may be something we broke within the product? We want to identify that and put together a plan to solve that quickly.”
Databricks also wanted to optimize time zone alignment in order to make its global team of engineers as effective as possible—and solve customer problems in real-time, no matter where they occurred.
Its solution? SupportLogic SX.