Services Processes and Waiting Line Analysis

1
Services Processes andWaiting Line Analysis

• Selected Slides from Jacobs et al, 9th Edition
• Operations and Supply Management
• Chapter 8 and 8A
• Edited, Annotated and Supplemented by
• Peter Jurkat

2
Service Businesses
8-2
A service business is the management of
organizations whose primary business requires
interaction with the customer to produce the
service

• Customer is the entire focus of attention a
  common definition of quality is satisfaction of
  the customer (more on quality later)
• Facilities-based services Where the customer
  must go to the service facility
• Field-based services Where the production and
  consumption of the service takes place in the
  customers environment

3
Characteristics of Workers, Operations, and
Innovations Relative to the Degree of
Customer/Service Contact
8-3
4
8-4
Service Blueprint, Failure Anticipation, and
Poka-Yokes
Complete blueprint (p262-3) identifies 16 failure
opportunities
5
Three Contrasting Service Designs
8-5

• The production line approach (ex. McDonalds)
• The self-service approach (ex. automatic teller
  machines)
• The personal attention approach (ex. Ritz-Carlton
  Hotel Company)

6
Well Designed Services

• 1. Each element of the service system is
  consistent with the operating focus of the firm
• 2. It is user-friendly
• 3. It is robust (avoid failures, poka-yokes)
• 4. It is structured so that consistent
  performance by its people and systems is easily
  maintained

• 5. It provides effective links between the back
  office and the front office so that nothing falls
  betweensic the cracks
• 6. It manages the evidence of service quality in
  such a way that customers see the value of the
  service provided
• 7. It is cost-effective

Lets consider Problem 8.4
7
Behavior and Guarantees

• Recent research suggests
• Any guarantee is better than no guarantee
• Involve the customer as well as employees in the
  design
• Avoid complexity or legalistic language
• Do not quibble or wriggle when a customer invokes
  a guarantee
• Make it clear that you are happy for customers to
  invoke the guarantee

• The front-end and back-end of the encounter are
  not created equal
• Segment the pleasure, combine the pain
• Let the customer control the process
• Pay attention to norms and rituals
• People are easier to blame than systems
• Let the punishment fit the crime in service
  recovery (task error vs. treatment error vs.

8
Waiting Lines

• Almost all services can have waiting lines, even
  along manufacturing line
• Waiting lines involve both layout and service
  management
• Can be the most damaging of service failures
  since customer never gets to experience the
  service
• Waiting lines also called queues (first in, first
  out)
• Trade-off more service (cost) vs. longer waits
  (customer dissatisfaction)

9
Managing Queues

• 1. Determine an acceptable waiting time for your
  customers
• 2. Try to divert your customers attention when
  waiting
• 3. Inform your customers of what to expect
• 4. Keep employees not serving the customers out
  of sight
• 5. Segment customers

• 6. Train your servers to be friendly
• 7. Encourage customers to come during the slack
  periods
• 8. Take a long-term perspective toward getting
  rid of the queues

10
Components of the Queuing System
Queue or
11
Customer Service Population Sources
Population Source
Example Number of machines needing repair when a
company only has three machines.
Example The number of people who could wait in a
line for gasoline.
Arrival Processes (usually measured by time
between arrivals) Constant (e.g., assembly
line) Deterministic (e.g., based on occurrence of
another event) Random/Stochastic (e.g.,
Exponential, Erlang) Batched (e.g., elevator, bus
load at rest stop) Depends on number in system
(e.g., machine repair)
12
Service Pattern
Service Pattern
Example Items coming down an automated assembly
line.
Example People spending time shopping.
Same classification as arrival process
13
The Queuing System
Single Q, single S Single Q, multiple S Multiple
Qs, multiple Ss, w/ Q switching
Queuing System
First in, first out (FIFO) First in, last out
(LIFO) Various priorities
Constant inter-arrival times Random Event
dependent
14
Examples of Line Structures
Single Phase
Multiphase (Sequential Servers)
Single Channel
Multichannel
15
Degree of Patience
No Way!
No Way!
16
Examples

• Service Systems
• Traffic on Networks messages to/from computers,
  cars on roads/rails, airplanes to/from
  airports/gates, ships to/from harbors/piers,
  elevators
• Retail/Service stores selling goods,
  service/repair shops
• Manufacturing Systems
• Primarily job shops, piece work, mass
  customization
• Appliances, Automobiles/Trucks, Toys, Clothing
• Logistics/inventory/distribution/MRP

17
Notation

• Many combinations of arrival and service
  processes, queue disciplines, populations, etc.
• Standard notation A/S/c/N/K/Qdiscipline
• A Arrival Process e.g., C for constant, M for
  Markov (exponential), Ek for Erlang, G for
  arbitrary
• S Server Process e.g., C for constant, M for
  Markov (exponential), Ek for Erlang, G for
  arbitrary
• c Number of Servers
• N System Capacity both queues and server
  stations
• K Size of Calling Population
• Queue Discipline FIFO, LIFO, various priorities
• M/M/1///FIFO default, shown as M/M/1
• Various A/S distributions possible most frequent
  are constant, exponential, Gamma, empirical

18
Poisson Process

• Inter-arrival time is exponentially distributed
• Completely determined by average time between
  arrivals
• Easy to specify (count arrivals and divide by
  time period)

• Equivalent to exponential inter-arrival time
• Provides probability of a given number of
  arrivals in unit time

19
Notation Infinite Queuing Models 1-3
See Exhibit 8A.8, p26
20
Infinite Queuing Models 1-3 (Continued)
21
Utilization

• Notice how sharply the average length of the
  queue grows with increasing average utilization
• For average r gt .7 short term increases in
  arrival (l) and/or service (m) can make queues so
  long that recovery is very long or may never
  happen

22
Calculating Performance

• Different models and conditions will generally
  dictate different equations for each performance
  measure
• Most situations fall into one of four models (all
  assume FIFO)
• Single server, single queue (SSQ M/M/1)
• Multiple servers, single queue (M/M/c) call
  center
• Finite system capacity (M/M/c/N)
• Finite population (M/M/c/K/K) maintenance crew
  of c for K machines
• Most included in available tables and software
  and approximations - see Ch08A_Queue.xlsx,
  Ch08A_Queue_Models.xlsx