Lion Financial Services Case Study

Lion Financial Services provides investment management services, mutual funds &
retirement plans to corporations, institutions and individual customers. In July of 2000, LFS
had 350,000 individual and institutional customers and 7.2 billion $ of assets under
management. It served its customers via 3 channels: face-to-face service in branch offices;
phone service through call centers; and online service through website. 50% of these
transactions occurred through the company’s call centers. The company had 3 call centers
located in Chicago, Boston and New Jersey, with Chicago being the largest by a huge margin.
The Chicago call center handled 20,000 calls per week on average, compared to 5000 each
for Boston and New Jersey. The LFS call center system handled 170 agents in total, each
agent being productive (“in the seat”) for 75% of his 40-hour work week. The staffing of an
agent pool was managed to keep 30% buffer capacity to protect against unforeseen variations
in call arrivals and durations. Incoming calls could be divided into

 “Quickline” calls: basic calls, with an average duration of two minutes, that did not
require any follow-up. These were divided into 40% customer and 60% broker calls.
They were all routed to the Chicago call center irrespective of call source location.
 “Non-Quickline” calls, requiring an average of five minutes of customer interaction
and three minutes of follow up work. These calls were routed to the call center
responsible for their region of origin.
All incoming calls were in effect split into 8 streams: 2 each in Boston and New Jersey
(Broker and Customer Service), and 4 in Chicago (Broker & Customer Service, Quickline
Broker & Customer Service). Each of these streams had to maintain an agent pool of 10-30
agents, headed by supervisors. This disaggregated arrangement led to hiring more agents than
necessary to maintain a good agent strength in each of the three location. There was
inefficient utilization of agents too, as there could simultaneously be a situation where an
agent in New Jersey is sitting idle while the Boston center is flooded with non-quickline calls
that cannot be transferred to another center. This led to redundant overstaffing, and hence an
inflation of the overall staff size that did not translate to increased productivity, reflected in
the long waiting times (ASA 46 seconds) that caused ~15% of callers to hang up in
frustration. With increased utilization and proper performance evaluation, each agent could
be motivated to be productive for more than 75% of his time, further reducing the number of
agents required. Thus, the staff size is not appropriate.

Andy’s proposal of reducing the number of physical locations to one, Chicago, instead of
three, facilitated direct reduction of the infrastructure and operational costs, further
ensuring the feasibility of reducing the agent pools from eight to three, which when supported
by the queueing and staffing techniques led to both cost reduction and customer service
level improvement by enabling the following steps:

 By combining call centers, economies of scale could be leveraged to eliminate

redundant workforce among agents and supervisors, thus reducing cost of wages.
  The simplified overflow scheme facilitated by the new queueing system, ensured that
the call was passed to only a qualified agent even during high congestion times, thus
ensuring fewer escalations, process cost reduction & customer satisfaction by
proper queueing system.

 The two full-time call routing and scheduling staff could be eliminated due to the
simplification and automation of overflow routing schemes.

 The average waiting time reduced for Non-Quickline customers due to consolidation
of calls to one physical location.
Andy's philosophy of work system design was highly process-oriented. He focused on
doing the work economically by standardizing the process, automation and making use of
economies of scale. He was looking to optimize the process to increase productivity, better
the service level and for optimizing the use of less experienced resources to reduce the
average wage rate. He wanted to set up a production-line mentality and to objectify the
expectations and priorities.
These convictions are reflected in the following proposals to LFS management:

 In the measurement for call center performance, Andy aimed to move every
interaction into the “satisfied category”. Although the prime focus was on
minimization of disastrous calls, he also discouraged any heroic efforts which will
cause variability or inconsistency in the service process.

 He wanted to introduce follow up questionnaires to ensure that the handle times

were reduced and the follow-ups are taken offline.

 He wanted to introduce an objective feedback system to ensure clarity and

consistency in understanding.

 The agent pools had to be reduced to three from eight, which would ensure
economies of scale to reduce the costs and ensure lower waiting times for the
customers.

 He proposed carefully designed blueprints and rule of thumb to be followed and to

categorize calls into clear categories, to emphasize on commonalities striving to
standardize and streamline processes.

 He suggested that it is not necessary to develop personal ties with the clients since
this will add on to their account management costs.

 He also quoted Henry Ford while advocating to make the processes simple and robust
enough to exist without the need of highly skilled workers.

Andy’s proposal had several attractive features. Chief among them was the reduction in
operating cost by 1,50,000$ annually. Using the new training and performance measurement
schemes, he intended to make the customer service predictable and economical. The call
center’s performance was measured in quantifiable terms like Average speed of answer,
Average talk time & handle time. Lower handle time was a clear indicator of high customer
satisfaction. Along with Agent utilization and Error rate, the call monitoring scores were used
to ensure compliance with the new SOPs. The scores were also standardized in a score sheet
format to ensure greater parity in evaluation. Mystery shopping comparisons was also an
effective tool to conduct random, surprise checks to get qualitative feedback on customer
service. By consolidating the call center locations, one could simplify the overflow routing
rules. The implementation of automation in call routing and customer service reduced
workload on manpower, thus facilitating the reduction in the number of agent pools and the
probability of hold up in calls. To the agent’s delight, the obligation to answer email
messages and other peripheral tasks was removed. By focusing on training through SOPs and
blueprints, Andy’s recommendation had reduced the dependency on the agent’s individual
expertise in finance and investment matters, thus making possible the hiring of less educated
and experienced agents who could be turned into professionals with 6 months training,
resulting in reduction of hourly wage rate. Individual performance evaluation was also
facilitated through the Automatic Call Distributor, making it more objective and transparent.
Group bonuses were modified to be based on a standard set and encompassing each of the
three groups entirely, thus reducing variability in performance criteria.

The major concern was the possibility of high turnover resulting from loss of morale due
the agent’s loss of freedom and autonomy. The loss of morale and resulting indifference
might also grossly affect customer service quality. All this is supposed to be mitigated due to
the strength of the new SOPs, blueprint and training modules, but if these protocols fail to
live up to the expectations, the results could be disastrous. The changes suggested also
seemed too radical to be implemented all at once without causing mass uproar in the
workforce. His modifications to the existing training program also overlooked the fact that
the current training program was very successful as a recruitment tool because of the perks it
offered and the new program might not be as successful. Andy’s casual dismissal of the
attrition of knowledgeable workers could be a sign that he does not appreciate the gravity of
the problem and the replacement costs associated with said attrition.

Similar to TPS/JIT principles, Andy’s recommendation was also centered around the
importance of “process development” to improve productivity and service levels. It called for
a production line mentality, with standardization of operating procedure for agents through
workflows, blueprints and rules of thumb. This led to a consistent and error-proof customer
experience, which itself is a key value that the customer will appreciate over personalized,
but idiosyncratic, service. It emphasized on trainings and performance measurement in clear,
quantifiable terms. Similar to the Toyota framework, Andy’s recommendation calls for
agents operating in a tightly defined and managed work system, with ‘best practices’ captured
and documented in the form of manuals and rules of thumb; workers generate and test
incremental adjustments to said process (similar to kaizen- continuous improvement) to keep
ironing out the kinks and improving productivity that won’t be affected by worker attrition.
With regards to employee involvement, Andy believed that treating people with dignity and
respect in a process-oriented environment is very much possible, especially through open and
honest dialogue about expectations and priorities. Individual initiative should be encouraged,
but within a disciplined framework and only in service of the needs of the system.