From the March 2014 issue of Entrepreneur

Most healthcare facilities do their staff scheduling on paper. OnShift, a software-as-a-service (SaaS) startup that automates the scheduling process for nursing homes and other healthcare facilities, is trying to change that.

Even with sales growing 90 percent annually since 2008, Mark Woodka, CEO of the Cleveland-based company, was obsessed with the thought of customers abandoning his SaaS scheduling program. The management team discussed churn "at every single meeting," fearful that they lacked the manpower to retain their ballooning customer base.

Woodka considered hiring additional customer-service managers.

Then he found Gainsight, a Mountain View, Calif.-based customer-success management company tailor-made to address customer retention for SaaS companies like his.

Churn police: Gainsight's Mark Woodka.
Churn police: OnShift's Mark Woodka.
Photo © Ian Londin

The Fix
Woodka brought in Gainsight to determine OnShift's unique leading indicators of churn, such as a customer not using the product for a period of time or a rash of service tickets. Once these metrics were set, OnShift linked Gainsight's software into its systems, and whenever a customer triggered a churn metric, OnShift's "customer success" team received an e-mail alert.

The Results
OnShift's churn rate dropped from 4 to 2 percent, and Woodka estimates that Gainsight saved the company $325,000 in lost business.

In addition, his managers now have a 360-degree view of every customer within Gainsight's user interface, which enables them to respond to client calls with more complete information. Woodka claims the feature will allow his customer-success managers to double the number of clients they can competently handle to 400 each. He expects the company to reach that scenario within the year.

A Second Opinion
Peter Fader, professor of marketing at the Wharton School of the University of Pennsylvania and co-director of the Wharton Customer Analytics Initiative, says the statistical models used by Gainsight are good at identifying whether a customer will continue to engage with a service in a given period. But forecasting precisely when a customer might leave is much more difficult.

"Those models are good for the which, bad for the when," Fader says.

So Gainsight's clients should be prepared to act quickly on any alerts. It's a reminder that gathering analytics will get you only so far--a firm must commit to applying those analytics intelligently.