Capacity and Demand Management

1
Capacity and Demand Management

• MD254
• Service Operations
• Professor Joy Field

2
Strategic Role of Capacity Decisions in Services

• A capacity expansion strategy can be used
  proactively to
• Create demand through supply (e.g. JetBlue,
  Dunkin Donuts)
• Lock out competitors, especially where the market
  is too small for two competitors (e.g. WalMart)
• Get down the learning curve to reduce costs (e.g.
  Southwest Airlines)
• Support fast delivery and flexibility (e.g.
  Mandarin Oriental)
• A lack of short-term capacity can generate
  customers for the competition (e.g. restaurant
  staffing)
• Capacity decisions balance costs of lost sales if
  capacity is inadequate against operating losses
  if demand does not reach expectations.
• Strategy of building ahead of demand is often
  taken to avoid losing customers.

3
Capacity Planning Challenges in Services

• Inability to create a steady flow of demand to
  fully utilize capacity
• Enforced idle capacity if no customers are in the
  service system
• Customers are participants in the service and the
  level of congestion impacts perceived quality.
• Customer arrivals fluctuate and service demands
  also vary.
• Capacity is typically measured in terms of
  (bottleneck) resources rather than outputs (e.g.
  number of airplane seats available per day rather
  than number of passengers flown per day).

4
Customer-Induced Demand and Service Time
Variability

• Arrival customer arrivals are independent
  decisions not evenly spaced.
• Capability the level of customer knowledge and
  skills and their service needs vary
• Request uneven service times result from unique
  demands.
• Effort level of commitment to coproduction or
  self-service varies.
• Subjective Preference personal preferences
  introduce unpredictability.

5
Modeling Service Delivery Systems Using Queuing
Models

• Customer population
• The source of input to the service system
• Whether the input source is finite or infinite
• Whether the customers are patient or impatient
• The service system
• Number of lines - single vs. multiple lines
• Arrangement of service facilities servers,
  channels, and phases
• Arrival and service patterns e.g. for many
  service processes, interarrival and service times
  are exponentially distributed (arrival and
  service rates are Poisson distributed)
• Priority rule (queue discipline)
• Static
• First-come, first-served (FCFS) discipline
• Dynamic
• Individual customer characteristics e.g.
  earliest due date (EDD), shortest processing time
  (SPT), priority, preemptive
• Status of the queue, e.g. number of customers
  waiting, round robin

6
Queue Configurations and Service Performance
Multiple Queue
Single queue
Take a Number


Enter
7
Arrangement of Service FacilitiesChannels and
Phases
Service facility
Server arrangement Parking lot
Self-serve Cafeteria
Servers in series Toll booths
Servers in parallel Supermarket
Self-serve, first stage parallel servers, second
stage Hospital Many service
centers in parallel and series, not all used by
each patient
8
Distribution of Patient Interarrival Times for a
Health Clinic
Patient interarrival times approximate an
exponential distribution.
9
Temporal Variation in Arrival Rates
Ambulance Calls by Hour of Day
Physician Arrivals by Day of Week
10
Queue Discipline
11
Single-Server, Exponential Interarrivaland
Service Times (M/M/1) Model

• Assumptions
• Number of servers 1
• Number of phases 1
• Input source infinite, no balking or reneging
• Arrivals mean arrival rate mean
  interarrival time
• Service mean service rate mean service
  time
• Waiting line single line unlimited length
• Priority discipline FCFS

12
Single-Server Operating Characteristics

• Average utilization
• Probability that n customers are in the system
• Probability of less than n customers in the
  system
• Average number of customers in the system
• Average number of customers in line
• Average time spent in the system
• Average time spent in line

13
Multiple-Server (M/M/c) Model

• Assumptions
• Number of servers M
• Number of phases 1
• Input source infinite, no balking or reneging
• Arrivals mean arrival rate mean
  interarrival time
• Service mean service rate mean service
  time
• Waiting line single line unlimited length
• Priority discipline FCFS

14
Multiple-Server Operating Characteristics

• Average utilization
• Probability that zero customers are in the
  system
• Probability that n customers are in the system
  
• Average number of customers in line
• Average time spent in line/system
• Average number of customers in the system
• Average waiting time for an arrival not
  immediately served
• Prob. that an arrival will have to wait for
  service

15
Capacity Utilization and Capacity Squeeze

• A capacity squeeze is the breakdown in the
  ability of the operating system to serve
  customers in a timely manner as the capacity
  utilization approaches 100. As the variability
  in arrival and service rates increases, a
  capacity squeeze occurs at a lower capacity
  utilization.

100 10 8 6 4 2 0
With
Then
System line length
0 0 0.2 0.25 0.5 1 0.8
4 0.9 9 0.99 99
0
1.0
Capacity utilization
16
Service System Cost TradeoffTotal Cost of Service
Let Cw Hourly cost of waiting customer
Cs Hourly cost per server C Number
of servers Total cost/hour Hourly service cost
Hourly customer waiting
cost Total cost/hour Cs C Cw Ls

• The total cost of service reflects both the
  firms capacity cost as well as the customers
  cost of waiting. Service processes should be
  designed to minimize the sum of these two costs.
• How can the economic cost of customer waiting be
  determined?

17
Queuing Model Takeaways

• Variability in arrivals and service times
  contribute equally to congestion as measured by
  Lq.
• Even though servers will be idle some of the
  time, there will be customer lines and waits, on
  average. These lines/waits will get very long
  very quickly as capacity utilization approaches
  100.
• Given the potential for a capacity squeeze as
  capacity utilization approaches 100, service
  firms typically design their processes with a
  capacity cushion (i.e., the amount of capacity
  above the average expected demand). The greater
  the variability in arrival/service rates, the
  larger the capacity cushion needed for a given
  service level.
• To improve system performance (waits and line
  lengths)
• A single queue vs. multiple queues with multiple
  channels.
• More servers can be added (reducing capacity
  utilization but at a higher operating cost).
• A fast single server is preferred to
  multiple-servers with the same overall service
  rate.

18
Managing Waiting Lines
In a lifetime, the average person will spend
SIX MONTHS Waiting at stoplights
EIGHT MONTHS Opening junk mail
ONE YEAR Looking for misplaced objects

TWO YEARS Reading E-mail

FOUR YEARS Doing housework

FIVE YEARS Waiting
in line

SIX YEARS Eating
19
The Psychology of Waiting

• People dislike empty time Fill this time in a
  positive way.
• Service-related diversions convey a sense that
  the service has started (e.g. handing out menus).
• Waiting can induce anxiety in some customers
  Reduce anxiety by providing information to the
  customer (e.g. expected wait times).
• Customers want to be treated fairly while
  waiting First-come-first-served (FCFS) queuing
  discipline or logical prioritization process
  (e.g. triage)

20
Managing the Customer Waiting Experience

• Conceal the queue from the customer.
• Engage the customer in co-production tasks during
  the wait.
• Provide diversions during the wait.
• Serve priority customers or customers who are
  willing to plan ahead faster.
• Automate standard services to enable
  self-service.
• Manage waiting time perceptions under promise,
  over deliver.

21
Managing Demand and Capacity to Reduce Lines and
Waiting Times
Yield management
22
Managing Demand

• Segmenting demand (e.g. random vs. scheduled
  arrivals)
• Offering price incentives (e.g. lower matinee
  pricing at movie theaters)
• Promoting off-peak demand (e.g. use of a resort
  hotel during the off-season for business or
  professional groups)
• Developing complementary services (e.g. HVAC)
• Reservation systems and overbooking (tradeoff
  between opportunity cost of unused capacity and
  costs of not honoring an overbooked reservation)

23
Managing Capacity

• Increasing customer participation (e.g.
  e-commerce)
• Scheduling work shifts (based on historical
  demand patterns and desired service level)
• Creating adjustable capacity (e.g. Tesco online
  grocery fulfillment)
• Using part-time employees (e.g. during tax
  season)
• Cross-training employees (to increase workforce
  flexibility and leverage capacity to provide
  additional value-added services)
• Sharing capacity (e.g. gate-sharing arrangements)

24
Flow Management
Three stage service process, average service
rates

• Flow management focuses on relieving bottlenecks
  so that customers can move more smoothly and
  quickly through the service process.
• How can the flow of this service process be
  improved?
• Resource-side
• Demand-side

25
Maximizing Utilization vs. Flow Management

• Compare and contrast the process performance with
  a maximizing utilization vs. flow management
  approach.
• Why does flow management usually improve capacity
  utilization, but maximizing utilization often
  results in poor flow?

26
Yield Management

• Yield management attempts to dynamically allocate
  fixed capacity to match the potential demand in
  various market segments to maximize revenues and
  profits.
• Although airlines were the first to develop
  yield-management, other capacity-constrained
  service industries (e.g. hotels, car rental
  firms, cruises) also use yield management.
• Possible ethical issues associated with yield
  management? (http//en.wikipedia.org/wiki/Yield_ma
  nagement)

27
Ideal Characteristics for Yield Management

• Relatively fixed capacity
• Ability to segment markets (i.e., discount
  allocation)
• Perishable inventory (i.e., potential for
  spoilage)
• Product sold in advance
• Fluctuating demand
• Low marginal fulfillment costs and high marginal
  capacity change costs