There's a particular anxiety that keeps SaaS executives up at night: the slow-motion train wreck of customer churn. You see the signs—fewer logins, declining usage, radio silence from a once-enthusiastic champion—but usually too late to salvage the relationship. What if you could see it coming a year out?
QuadSci, a New York outfit founded just three years ago, thinks it's cracked the code. The company announced Tuesday it has closed an $8 million Series A round, led by Crosslink Capital with backing from Alumni Ventures and Correlation Ventures. A handful of notable angels joined as well, including Rob Eberle, who ran payments company Bottomline Technologies, and Tom Roloff, a veteran of EMC's enterprise machine.
The pitch is deceptively straightforward: feed us your product usage data—every click, every login, every feature touched—and we'll tell you which customers are headed for the exit, which ones are primed to spend more, and which are quietly sliding toward downgrades. Co-founder Sean Murray frames it as a signal extraction problem, not a data problem. "Revenue teams don't have a data problem," he said in the funding announcement. "They have a signal problem."
Perhaps. But the claims are bold enough to raise eyebrows. QuadSci says it can predict churn twelve months in advance with 94% accuracy, forecast expansion opportunities at 90%, and spot pending contractions six months ahead with 80% precision. Those are the kinds of numbers that either transform how software companies manage their customer base—or prove too good to be true in messier real-world conditions.
Reading the Tea Leaves in Telemetry
Murray and his co-founder Dan Harmeson, who launched QuadSci in 2023, bet everything on behavioral telemetry over traditional signals. Forget CRM notes. Ignore satisfaction surveys. The real predictive power, Harmeson argues, lives in the raw exhaust of product usage. "Telemetry data contains 80% of the predictive power of whether a software customer will churn or grow," he said.
The platform hoovers up what it describes as trillions of telemetry events across its customer base, pulling from more than 50 data sources—Salesforce, Segment, Mixpanel, Amplitude, Datadog, and the usual suspects in the B2B stack. Clari, the revenue operations platform and QuadSci's marquee customer, has piped in over six billion usage signals. Other clients include data management firm Reltio, sales engagement platform Salesloft, integration specialist Boomi, and—perhaps tellingly—Gainsight, itself a major player in customer success software. J.P. Morgan has signed on in an advisory capacity.
The technical approach is less about building another analytics dashboard and more about pattern recognition at scale. QuadSci's two main products, Growth AI and Cohorts AI, automatically segment customers based on behavior and surface which accounts warrant immediate attention. The company is also working on what it calls "revenue agents"—AI systems designed to embed directly into go-to-market workflows, presumably nudging sales and customer success teams toward action before it's too late.
A Market Under Pressure
The timing feels deliberate. SaaS retention metrics have been sliding. QuadSci cites research it conducted alongside SBI Growth Advisory showing that average net revenue retention—a key health indicator for subscription businesses—has dipped from 110% to 107%. More worrying: 58% of SaaS companies report their retention is weaker now than it was two years ago.
That's the backdrop. Public SaaS multiples have compressed. Growth-at-all-costs strategies have given way to profitability mandates. And in that environment, keeping existing customers happy matters more than it has in a decade. Which creates an opening for a company like QuadSci—assuming revenue teams are ready to trust algorithmic forecasts over the gut instincts and relationship intuition that have long governed account management.
There's also the question of competition. Gainsight, already a customer, has been aggressively adding AI features. Product analytics vendors like Amplitude and Mixpanel occupy adjacent territory. Customer data platforms are expanding their predictive capabilities. QuadSci's differentiation hinges on being really good at the prediction piece, and on processing usage data at a scale most competitors haven't prioritized.
What's Ahead
The fresh capital will go toward product development and headcount. QuadSci currently lists between 11 and 50 employees—a vague range, admittedly—and is recruiting engineering talent in Mexico City, a signal the company sees opportunity beyond U.S. borders. The firm previously raised $2.1 million in August 2023, according to SEC filings, and claims it grew revenue fivefold year-over-year. It also collected a "Machine Learning Company of the Year" award from AI Breakthrough last June, though such accolades can sometimes mean more internally than in the market.
The real test will come in execution. Can QuadSci's models hold up across different industries, product types, and customer segments? Will revenue teams actually change their behavior based on algorithmic recommendations, or will the platform become another source of noise competing for attention? And can a relatively small, young company move fast enough to stay ahead of incumbents with far deeper pockets and larger customer bases?
For now, the company is betting that behavioral telemetry unlocks something CRM data and customer surveys can't. Whether that thesis survives contact with a skeptical market—and whether 94% accuracy translates to renewed contracts and saved deals—remains to be seen. But if QuadSci is even half right about what's lurking in usage logs, a lot of customer success teams are about to rethink how they see trouble coming.
