Startup Measures Success By Metrics Its Own System Generated

Cascade Analytics, a two-year-old customer success platform based in San Francisco, announced Tuesday that it had achieved a ninety-seven percent customer satisfaction rating using a measurement system the company had engineered from the ground up to detect satisfaction. The startup's logging infrastructure, deployed across its own customer accounts last quarter, captures behavioral signals that its dashboard interprets as contentment, then surfaces those interpretations as charts to be reviewed by Cascade's own leadership team each morning.

We wanted to know if customers were happy, so we built a system that could tell us.

The method works by identifying moments when users perform actions the system was configured to classify as positive, then tallying those moments against thresholds the company established after consulting its own records about what customers were already doing. "We wanted to know if customers were happy, so we built a system that could tell us," said Marcus Webb, Cascade's chief product officer and the engineer who wrote the initial detection rules. "That's science."

The approach has become standard across the customer success technology industry, where platforms routinely measure their own efficacy by instrumenting their own infrastructure and then analyzing the data they collected. Competitors including Gainsight and Totango have built similar closed-loop systems, creating a market in which the primary customer success metric is how successfully a customer success platform can convince its operator that customer success is occurring. Venture investors have backed this model enthusiastically, funding forty-three such companies in the past eighteen months.

At press time, Cascade had begun building a second system to measure whether the first system was generating accurate measurements, and had already configured it to report that it was.