Challenges in the Global Delivery Model Aliza Heching, Abhijit Bose IBM TJ Watson Research Center Wharton Call Center Forum February 28-29, 2008
Managed Systems What is the Global Delivery Model? Servers, networks, monitored business processes, resources provider manages on behalf of clients 24x7 basis; provider or customer owned data centers Managing Systems/SDCs Tools, systems, processes used to deliver services to clients Receives work requests from clients Interacts with clients about delivery Remote Sites/GDCs Processes workflows from SDC These three components can be established in different geographies 2 Wharton Call Center Forum February 28-29, 2008
Benefits of Global Delivery Model Ability to provide round-the-clock services Business continuity and resiliency in case of disaster Scalability Enforces worldwide best practices Ability to leverage local skills and cost structure 3 Wharton Call Center Forum February 28-29, 2008
Outline What is the global delivery model? Outsourcing model Service delivery model Organization of delivery personnel How are the contracts structured? Key challenges Examples of our approach 4 Wharton Call Center Forum February 28-29, 2008
Outsourcing (Managed Systems) Model Outsourcing Provider Owned Can host multiple client infrastructure Easier to scale More economical and better operational efficiency can be achieved Customer Owned Customer Benefits: maintain legacy systems, security is in-house Service Provider: Increased complexity Potential higher cost of delivery Integration challenges 5 Wharton Call Center Forum February 28-29, 2008
Service Provider Delivery Models Service Delivery organized by Service Towers (UNIX, SAP, ) Gain specialized skills in each service tower Delivery teams are specialized in 1-2 skills Services are coordinated across multiple service towers Clients purchase service components from providers Service Delivery organized by Industry Sector (Banking, Retail, ) All of a customer s needs are handled by a single service delivery team Customers in the same industry sector often have similar needs Providers can develop specialized solutions by industry sector Delivery teams are cross-trained in multiple skills 6 Wharton Call Center Forum February 28-29, 2008
Organization of Delivery Personnel What is each site used for? Onsite Deskside support System architecture, design & consultation Project Management Offsite Application development System support Call center and help desk operations Offshore Backend operations Most of offsite operations 7 Wharton Call Center Forum February 28-29, 2008
Components of Contract Description of Service line Length of contract Cost Personnel location Customer Contracts Customer owned versus provider owned Cost savings target Sometimes paid upfront SLA agreement penalty per violation x% of tickets over a given period must be solved within the SLA 8 Wharton Call Center Forum February 28-29, 2008
Challenges Global delivery environment: offshore-onshore delivery model local regulatory environment, cultural differences, time zone differences Integration of resources to provide single (unified) service Tight margins and high SLA penalties Non-standardized contracts Lack of industry-wide standardization Processes, materials, configurations Lack of automation, high variability in tools, service times, nonstandardized workflow management 9 Wharton Call Center Forum February 28-29, 2008
Approach Take methods that have traditionally been applied in a manufacturing setting and transfer them to this service delivery setting Metrics to measure performance of delivery system Statistical process control Skills requirement (workforce) planning Account health management Correlation of financial and operational metrics Skills-based routing 10 Wharton Call Center Forum February 28-29, 2008
1. Metrics to measure performance of delivery system Metrics to measure process performance: Manufacturing: yield, scrap Service delivery: # of tickets closed, mean service response time Standardized metrics allow for: Cross comparison across different IT infrastructure Benchmarking Identify areas for improvement Means to measure impact of process improvement initiatives SLA NOT MET 3 2 1 0 10/12/2007 10/19/2007 10/26/2007 11/2/2007 11/16/2007 11/9/2007 Metric measurement before/after process improvement 11 Wharton Call Center Forum February 28-29, 2008 Identify nonrandom patterns
2. Statistical process control Developed by Shewhart and popularized by Deming (1920s) Long used in manufacturing to ensure acceptable system variation Monitor process performance according to the standardized metrics Determine if Voice of Process is aligned with Voice of Customer Insight into process capability provides guidance in contracting phase With this example, the process is not physically capable of consistently meeting the customer s needs. Individuals Chart Percent Within Criteria Individuals Chart Percent Within Criteria 105% 100% 95% 90% 85% 80% 75% 70% 65% 60% 12/1 12/3 12/5 12/7 12/9 12/11 12/13 12/15 12/17 12/19 12/21 12/23 12/25 12/27 12/29 12/31 1/2 1/4 VOC UNPL Average LNPL 105% 100% 95% 90% 85% 80% 75% 70% 65% 60% 12/1 12/3 12/5 12/7 12/9 12/11 12/13 12/15 12/17 12/19 12/21 12/23 12/25 12/27 12/29 12/31 1/2 1/4 Target UNPL Average LNPL 12 Wharton Call Center Forum February 28-29, 2008
3. Skills Requirement Planning Problem: Lack a complete and reliable understanding of the resources (skills) needed for different tasks. Result: Inappropriate resources being assigned to projects Added cost for incremental capacity Poor account performance/lost revenue Resulting effort: Analyze utilization data by requirement 13 Wharton Call Center Forum February 28-29, 2008
3. Skills Requirement Planning: Identify Activities Requiring High-Band Skills Task A Task B Task C Task D Task E Task F Task A mostly used Band 2-4 skills, where as Task C required higher Band skills 14 Wharton Call Center Forum February 28-29, 2008
3. Skills Requirement Planning: Identify Major Trends in High- Band Skills Utilization Task C1 Task C2 Task C3 Task C4 15 Wharton Call Center Forum February 28-29, 2008
Key Take-aways Outsourcing is a challenging business Global delivery model exacerbates the problem We are applying lessons learned from the manufacturing environment to improve performance in our service delivery business 16 Wharton Call Center Forum February 28-29, 2008