Tucked into the tiny Cleveland, Ohio suburb of Solon, Insurance.com does a brisk business. Since their humble beginnings in 2000, the independent agency has built its online insurance business, today representing 15 top carriers in 46 states, and counts 700,000 policyholders in its ranks. Today, the company processes 1 million comparison quotes annually.
The direct-to-consumer business model is working, and consumers can purchase insurance online or speak with licensed agents.
This success did not occur in a vacuum; the company has focused on strategic improvements in top-line growth fueled largely by developing expert customer service. With 150 agents in the sales center, 80 of whom are staffed on a 24/7 basis, the call center is considered paramount to the company's success.
But Insurance.com's aptitude for customer service did not arise without its share of difficulties. Until recently, one of the internal systems used to monitor the company's call center systems was comprised of applications written in C-Sharp.
"The system was expensive to maintain, and actually started to fail," says Scott Noerr, director of IT systems at Insurance.com. "There were no alerts, so the end user had to watch to see if there were problems. When you look at our various business requirements that relate to time to market, we realized we needed to revisit this application."
The need to review its outdated call center systems led the company to start thinking about creating ways to leverage their pending investment into an even larger return. "As our systems continued to grow, so did the complexity, with multiple data sources, and we knew we needed to turn that into actionable information," Noerr says.
BOLSTERING CUSTOMER SERVICE
When the opportunity to revisit their call center technology presented itself, the Insurance.com IT team began focusing on how to use business intelligence (BI) to bolster customer service levels. The group then created a punch list of elements they felt necessary to accomplish those goals.
Continuous monitoring and feedback were high on the list, says Noerr. "In any contact system (sales center), call volume will be dynamic. We knew it was difficult to optimize the agent workflows or agent utilization, so by putting a system in place that lets us monitor various heterogeneous data streams concurrently, we've been able to optimize agent utilization."
The end result, Noerr adds, is that instead of peaks and valleys, we've level-set the agent workflow.
SETTING THE DIAL
A long-time Cognos (now IBM Cognos) client, Insurance.com had always embraced business intelligence (BI), reports Craig Bedell, director of Global Performance Management Insurance Services at IBM Cognos, and had been using the Cognos 8 platform to obtain and process routine business information. What came next meant changes to hardware and software.
A little more than a year ago, IBM Cognos NOW! was loaded into an appliance (a piece of hardware dealing with real-time data), reports Bedell. The BI solution is designed to deliver operational dashboards to continuously monitor key performance indicators (KPIs) and operational metrics from disparate data sources.
Tightly integrated with IBM Cognos 8 BI, the new system provides up-to-date, actionable intelligence for rapid, fact-based decision-making, especially key for Insurance.com's line operations, Bedell adds.
"They use that appliance for two purposes," he says. "They've hooked it into their infrastructure to follow the quoting process, page activity relative to their Web site and load times. They also use it to review and monitor errors that come from not just their system, but from any of the 15 companies for which they provide services."
The software times the various feeds coming in from quoting engines for VIN validation, etc., and the system holds set tolerances to create an alert that feeds information to a cube. Traditional business intelligence acts on that information.
Noerr explains that there are a variety of data feeds (from CRM, telephony systems, interviews, etc.) that can trigger an agent to take a call. Some of those feeds the insurer has control over; for example, how long an agent sits idle before the call center manager pushes them a lead, or if, for some reason, they were not able to finish the quote and have been requested to call the lead back.
DIAL IT UP
"Our managers will get alerts in real time, notifying them that certain agents are either overloaded or underutilized, so we will contact them through the Cognos appliance, telling them to up or down the dial," Noerr says. "When we see volume getting lower, instead of waiting 60 seconds, we will start turning the dial down to 45 seconds, so we continue to dial that down ... if volume picks up, it automatically ups the time to 90 seconds, and so on."
Future plans call for a Web service to be created to automate this functionality. For now, however, Insurance.com is capitalizing on an intelligent "pacing" of lead flow in real time. Noerr says the system has helped maximize the impact of their online marketing campaigns (responding to a local TV ad or banner campaign, for example) and streamline agent utilization to keep costs in check.
"All along, we've wanted to be more proactive in the data center," says Noerr, "and we are on the right track. When we have volume online, and we know what percentage of inbound calls end up in our sales center, we can staff up. We can change the phone load of an agent. If we know there will be a spike in call volume in California, for example, and our agents are in "chat" mode, we can change that to "inbound."
Although privately held and therefore, unwilling to reveal revenue figures, Insurance.com's Noerr confirms that both soft and hard returns on the system have been positive. One such return is in cost savings.
"With 15 different variables currently going into the agent utilization metric, we are continuing to turn the dials up and down based on tolerance," he says. "This has already helped us reduce costs."
Another return helps Insurance.com retain its competitive advantage, as external business partners obtain real-time alerts on issues such as degradation or errors from the insurer that affect the agent's ability to obtain information and close the sale.
(c) 2009 Insurance Networking News and SourceMedia, Inc. All Rights Reserved.
