Business
This unique insurance company is the largest global provider of wireless handset insurance and wireless roadside assistance programs, serving 70+ million consumers worldwide. It also offers data-backup and credit protection services through wireless carriers in North America, as well as credit protection services. The company’s largest customers include ALLTEL, AT&T, Rogers Wireless, Sprint, T-Mobile, US Cellular and Verizon Wireless. It also partners with MVNO and other regional wireless providers.
A key program in the company’s Network Operations Center is a Customer Relationship Management (CRM) system. A prime example of CRM functionality is the filling of warranty orders, in which a customer calls an 800 number and follows the spoken instructions, in order to have a new unit or parts sent out overnight. Such a process is actually quite complicated because it involves multiple interactions between different applications including validation, inventory control, distribution, etc. which translate into complex transaction flows across multiple tiers
The CRM system, which operates via call centers distributed throughout the US, is deployed on 36 clustered servers and 7 middle-tier (.net app) servers all connected to an SQL server database in one location.
Challenge:
Assuring website availability and customer access
The company introduced a new concept to the world – insuring a single cellular device in any hands, as opposed to insuring a particular person or vehicle. As the idea took hold, the company took off. Due to its success in the US, the company is establishing mirror organizations in Japan and Europe. As the number of consumers grew, so did the amount of CRM calls for warranty replacement units and parts. The IT department was tasked with reducing downtime and speeding up the time to isolate and resolve bottleneck and availability problems.
Before BTM was implemented, problems were addressed by holding “all-hands” calls:
1. IT would receive a system alert or a complaint from people calling in.
2. Since IT could not “see” or isolate the problem, they could not verify it.
3. Various experts would be invited to participate from all areas of IT.
4. The experts would go back to analyze the data generated by existing OS monitoring tools and Java thread dumps, in order to try to isolate and resolve the problem.
Besides often leading to inconclusive or no results, this process typically took days and weeks. The IT Director later remarked, “Receiving ‘hints’ about busiest processes just wasn’t enough. We needed to ‘see’ the real flow of transactions.”
Solution:
Automated application transaction monitoring and problem isolation
The IT department began evaluating available monitoring solutions. After a two-month comparative evaluation period, BTM softwar was installed across the CRM application, delivering cross-tier visibility in real time.
The cause for this quick decision came from an unexpected quarter. Another vendor had invited the insurance company to visit a top-10 commercial bank to see its APM component monitoring solution in action. The head of IT at the bank asked what kind of tools the insurance company was looking for and suggested that they also take a look at the technology. He explained that “the [BTM] solution has become the bank’s primary monitoring tool and first line of defense against bottlenecks and outages.”
No sooner had he spoken, than a major outage occurred in the insurance company’s CRM system in production. Taking the bank’s advice, the IT Director decided to deploy BTM straight into a production environment, while still being piloted. To make a long story short – the way IT likes it! – the problem was isolated and resolved in record time. The rest is history.