WordCruncher

1
WordCruncher

• Verbatim Analysis

2
Search Manager

• Searching Comments
• Products
• Services
• Sales
• Support
• Searching Area or BU Type

3
Linking and Notes

• Linking to Glossaries, Terms, and Acronyms
• Creating Notes

4
Classification Lexicons

• Words, Complex Phrases, and Wild Cards
• _at_Positive vs. _at_Negative
• _at_Optimism vs. _at_Pessimism

5
Classification Lexicon (cont.)

• Product Lexicon

6
Foreign Language Lexicons

• _at_RU-Excellent
• (Russian)
• _at_AR-Excellent
• (Arabic)
• _at_CH-Excellent
• (Chinese)
• 12 Languages Counts without Fluency

7
Counts to Applications

• Applications
• Action Items
• Associations and Comparisons
• Trends
• Predictive Modeling
• Consistency

• Counts
• Words
• Phrases
• Word Parts
• Related Words
• Collocated Terms

8
Output and Results

• Flagging Action Items for Product or Sales
  Managers
• Classification Coding for Statistical Analysis
  (Numeric Variables, Dummy Variables, Flags)
• Pasting variables into MS Excel, MS Access, SAS,
  SPSS

9
Flagging Action Items

• Product Development -- Improvements
• Sales Equipment Update Sales
• Business Development Testimonials (Positive
  Responses)
• Support Improved Documentation
• Competition Strategy mention of competitors
  products

10
Output for Statistical Analysis

• Association Correlation and Regression
• Trend Time Series
• Component/Cluster and Discriminate Analysis

• BU vs. Product Type (Hubs, Routers)
• Negative Comments on Routers
• Area/Theater vs. Support or Service

11
Digging Deep and Discovering Relationships

• Making the Verbatim data useful to your client
  and his product and service managers
• Discovering what you dont know?
• Help managers discover what they dont know?

12
Making the Verbatim Data Useful

• Action Items
• Product/Service Notification Items
• Management Notification Items

13
Distributions
Education Level Respondent Care Effectiveness of
Survey
14
Automated Language Grouping

• English 1937
• Japanese 387
• German 216
• Chinese 185
• Spanish 164
• French 76
• Unknown 71
• Korean 69
• Russian 41
• Portuguese 39
• Italian 29
• No Comment 9
• Unknown languages are predominantly Polish,
  Hungarian, and Czech

15
Flags and Markers
Language Marker
Flag
16
Action Items

• Contact Requests
• (_at_contact lexicon call me, contact me, email
  me, . . . )
• Suggestions Questions (_at_suggestion lexicon
  I suggest, can you, would you, . . . )

17
Prod/Service Notification Items

• Routers (_at_router lexicon)
• 1700 Series
• Switches (_at_switch lexicon)
• 6500 Series
• Web Documentation (_at_documentation lexicon)
• TAC (_at_assistance lexicon)
• Firewall (_at_security lexicon)
• Certification (_at_ certification lexicon)

18
Management Notification Items

• Price (_at_price lexicon)
• Performance (_at_performance lexicon)
• Sales (_at_sales lexicon)
• Service (_at_service lexicon)
• Competitors (_at_competitor lexicon)

Action Items and Notification Items can be
broken down by Product,/Service, Area/Region, and
Market Segment.
19
Sorting Combinations

• Action Items and Routers

20
Sorting Combinations (Cont.)

• Routers and Competition

21
Sorting and Modeling Relationships

• Simple Sorts
• Routers by level of dissatisfaction
• Routers by suggestions
• Routers by competitor or products mentioned
• Routers by Documentation Dissats
• Multivariate
• Cluster Routers by dissat (price, performance,
  service, documentation, . . . )
• Finding the principal variables which contribute
  to router dissatisfaction (Area, Market Segment,
  price, performance, service, documentation, . . .
  )

22
Building Foreign Language Lexicons

• Foreign Language Product Lexicons
• Precision of Foreign Language Dissat Levels

23
Example of Foreign Language Analysis
Language Selection
24
Searching Collocated Terms
Discovering words that do not fit?
25
Summary

• Building precision into Lexicons by searching
  contextual material and exploring collocated
  words and phrases.
• Paste classification markers built by the
  classification lexical searches.
• Discover relationships by sorting the data
• Examine the relationships for underlying
  relationships
• Analyze the data using Multivariate Methods
  (Cluster Analysis, Principal Components, and
  Discriminant Analysis)