User Needs and User Profiling

1
User Needs and User Profiling
2
User Population
Target Expert Level
Target Novice Level
High Performance
Low Performance
3
Single View

• Any single view of the User is incomplete,
  whether it is the novice, average, or expert
• Each view only captures a limited number of users
  
• The implemented design may actually hurt the
  performance of the other non-represented users.
• Representing all of the Users individually is not
  cost effective.
• A reasonable number of views (3 to 5) has high
  leverage.
• Small enough to be definable, but large enough
  for coverage.

4
Behaviors - Strategies

• Basic Call Up
  Selling Total Revenue
• Rep 1
  
• Rep 2
  

5
User Models

• Modeling Approach
• The first step is to categorize the service
  representatives behaviors into clusters.
• The second step is to objectively and
  quantitatively describe the service
  representatives behaviors and strategies.
• The third step is to create a model for each of
  those clustered behaviors.
• Simulate the models to determine impact.

6
Grouped Agents
7
Performance by Group
High
Blue
Yellow
Monthly Sales
Orange
Purple
Low
Few
Many
Offers per Call
8
Qualitative User Models

• Blue Group
• Very few cross sell attempts
• Efficient call takers, shortest call duration
• Most likely do only what the customer requests
• High Revenue
• Yellow Group
• Moderate number of cross sell attempts
• Moderate call duration
• High Revenue

9
Comments

• Large, diverse user-populations can be
  categorized into distinctive groups.
• Top performers may use different strategies and
  those behaviors need to be understood.
• Models reflect factors that are important in
  achieving business goals.
• Models can be constructed of those behaviors.
• Models should be integrated with interface design
  and operational improvement.

10

• Customer-Centric Interface Design

11
Customer-Centric

• Customer-Centric approach takes the customers
  view or perspective in accomplishing tasks.
• Frequently, organizations focus on the business,
  department, or process perspective.
  Organizations concentrate on the processes,
  people, and technologies that must interact to
  address the customers request and ultimately
  accomplish the customers task.
• The goal is to have the Customer routed to the
  most appropriate location (agent) where the
  customer can on their own, or with our
  assistance, accomplish their desired task with
  the highest level of satisfaction at the least
  amount of cost.

12
Business-Centric Interface Customer has Mapping
Burden
Customers Contacts
Customers Expressions
Task Completion
Interface
Want Caller ID
MENU For Service Order Center, press 1.
For Billing Inquiry Center, press 2.
For Revenue Management Center, press 3.
SOC
Need a 2nd line
BIC
Dont understand my bill
?
RMC
Reconnect phone
How many? 1,000,000s
How many? 10
How many? 100,000s
13
Customer-Centric Interface System has Mapping
Burden
Task Categories
Customers Contacts
Customers Expressions
Task Completion
Interface
Want Caller ID
MENU To order service, such as caller id
or a second line, press 1. To get billing
information, press 2. To reconnect your
services, press 3.
Order Service
SOC
Need a 2nd line
Get Bill Information
BIC
Dont understand my bill
Reconnect Services
Reconnect phone
RMC
How many? 1,000,000s
How many? 100
How many? 10
How many? 100,000s
14
Customer-Centric Design

• Customer-Centric approach uses the customer task
  as the primary influence on interface design.
• Menu items would directly match tasks that
  customers are trying to accomplish.
• Menu items would be grouped and ordered by
  frequency of customer tasks.
• Menu items would be worded in the language of the
  customer.

15
Performance Realities

• All interface technologies have limitations. A
  given interface technology (e.g. IVR systems) can
  achieve a certain level of customer satisfaction
  and a certain level of performance (e.g.
  accomplish their task). All interface
  technologies will have less than perfect customer
  satisfaction and performance.
• Customer-Centric approach will specify a design
  that maximizes the performance of a given
  interface technology. In other words,
  Customer-Centric approach delivers the best
  interface possible for a given interface
  technology.

16
5 Step Design Approach
1. Customer Task Frequency Table 2. Customer
Perceptions for Interface Structure 3.
Interface Design 4. Customer Usability Test 5.
Refine Design for Implementation

• Understand customer goals

• Understand customer preferences

• Design from users perspective

• Lab tests field evaluations

• Reduce misdirects
• Increase customer satisfaction

17
Step 1 Customer Tasks

• Why do customers contact us ?
• Customers want to accomplish a task.
• What are those tasks ?
• Gather data.
• Customers opening statement provides an
  excellent snapshot of their (first, primary)
  task.
• Gather sufficient number of tasks to be
  statistically valid. (e.g. 2,500)
• Categorize tasks into Customer-Centric groupings.
  
• Tabulate results and construct Frequency Table.

18
Customer Task Frequency Table
19
Task Frequencies
20
Class Discussion Action-Objects
21
Step 2 Customer Perceptions

• Card Sort Method
• Customers perception of how customer tasks
  should be grouped into menu items. They are
  given the most frequent customer tasks and asked
  to categorize them into logical groups.
• Procedure
• Sort (numbered) task cards into groups based on
  the similarity of tasks.
• Rank groups in order of importance.
• The more tasks are linked, the more they should
  be in the same menu structure.

22
Customer Frequency of Use

• How often do the customers visit the interface?
  
• Frequent daily use.
• Infrequent walk up and use.
• Key behavior - do the customers remember the
  interface from one use to the next?

23
Customer Perception of Structure
24
Customer Perception of Structure
25
2. Add Optional Service 3. New Service/Open
Account 4. How to Use 5. Disconnect Optional
Service 7. Change Optional Service 8. Disconnect
Services/Close Account 12. Move Service 22.
Reconnect Service 24. Schedule a Move
26
1. Get Info about Bill 10. Schedule a Payment 13.
Information on Account 14. Discuss Bill 15. Get
Information on Payment 17. Where to Make a
Payment 18. Give Information on Payment 19. Make
a Payment
27
(No Transcript)
28
(No Transcript)
29
Step 3 Design

• Customer-Centric approach
• Customer Task Frequency Table identifies what
  topics should be included on the menu.
• Card Sort identifies how to group menu items and
  how customers would structure the interface.
• Customer Opening Statements identify how to word
  or what language to use in the menu items.

30
Customer-Centric vs Business-Centric
31
Action-Object in Design

• Rank order of design styles (best to worst)
• Action-Specific Object. (e.g. to order
  CallNotes)
• Specific Object. (e.g. for CallNotes)
• Action-General Object. (to order a service)
• General Object. (e.g. for all other questions)

32
Customer-Centric Interfaces

• Examples of Business-Centric menu items
• For orders, press 1.
• For repair, press 2.
• Examples of Customer-Centric menu items
• To get information about your account, or our
  services and prices, press 1.
• To get new service, move, change, or disconnect
  your service, press 2.
• To report a problem with your service, press 3.

33
Pseudo-Problem
34
Cut-Throughs
CRT Score
Announcement
Ann. A
Ann. B
2 2 4
User Selection
1
3
Time in seconds
5
10
15
20
1
Announcement
Ann. A
Ann. A
CRT Score
2 10 12
User Selection
9
3
Time in seconds
5
10
15
20
1
Ann. B
Announcement
Ann. A
CRT Score
- 3 (-1) -4
User Selection
1
3
Time in seconds
5
10
15
20
1
35
Cumulative Relative Time - CRT
36
Step 4 Customer Usability Testing

• Testing reduces risk. No matter how well the
  interface has worked in the past or on other
  applications, testing is a relatively quick
  inexpensive approach to reducing the risk.
• We set up a controlled experiment to ensure that
  implementation surprises are minimized.
• Approximately 100 customers (actually 96) perform
  tasks using the interface in order to achieve
  statistically valid results.
• Normally, more than one design (usually four) is
  tested and evaluated.

37
Experimental Design
38
Customer Satisfaction Comparison
39
Customer-Centric Results
40
Step 5 Refine Design for Implementation

• Performance Matrix
• Provides a view of two important performance
  indicators
  (i.e. customer satisfaction and correct call
  routing).
• Used to evaluate the performance of specific
  customer tasks.

41
Business-Centric Performance Matrix
42
Customer-Centric Performance Matrix
43
Post-Implementation Data
44
Comments

• Assume reported data is wrong. Or, at minimum,
  the data is not what it appears. Lesson always
  validate and verify data before performing
  analysis.
• Key Step Compare predicted to observed.