PG 1. PG 2 Improving Speech Applications with Usability Surveys How does Nortel measure the...
-
Upload
sheila-bradley -
Category
Documents
-
view
217 -
download
1
Transcript of PG 1. PG 2 Improving Speech Applications with Usability Surveys How does Nortel measure the...
PG 1
PG 2
Improving Speech Applications with Usability Surveys
How does Nortel measure the ‘Usability Pulse’ of Self
Service?
Judith SherwoodSales Engineer
Nortel Self Service Solutions
PG 3
Many Usability Test Methods
Data from Each Participant
Num
ber
of
Par
ticip
ants Live Pilot
EmployeeTest Calls
Follow-Up(call-back)Surveys
Focus Groups
Usability Survey
PG 4
What is a Usability Survey?
• Usability testing is an evaluation of a customer touch-point from a user perspective
• Typically conducted using small focus groups (12-20 subjects) in a controlled studio environment
PG 5
What is a Usability Survey?
• Methodology: Usability Survey using 200 to 500+ panelists to reveal problems affecting fewer callers o Chosen from a large pool of over
80,000 panelists with known demographics
PG 6
• Traditional Usability Testing Methods have sample size limitations
• If a problem affects only 5% of the users:
o 10-call sample has a 40% chance of finding it
o 100-call sample has a 99% chance of finding it
• If a problem affects only 1% of the users:
o 100-call sample has a 63% chance of finding it
o 500-call sample has a 99% chance of finding it
Size Matters !
PG 7
Methodology
Panelist places call, completes a task, and then fills out a questionnaire on an internet website
Entire call conversation is recorded for analysis
Survey ties the individual caller experience and questionnaire response to the call recording
Call-in Platform
Target Service
Post-Questionnaire
Panelist Pool
Design Tasks &Questionnaire
Analysis &Recommen- dationsReport
PG 8
Methodology
More efficient, less expensive, faster execution, broader feedback Panelists recruitment
and call-in campaign is outsourced
Analysis provides recommendations for service improvement Listen to problem calls
to suggest ways to fix
Call-in Platform
Target Service
Post-Questionnaire
Panelist Pool
Design Tasks &Questionnaire
Analysis &Recommen- dationsReport
PG 9
• Percentile-based letter grading system to compare against other speech applications
• Raw Scores are based on:
o Caller Satisfaction (% very satisfied - % dissatisfied - % very dissatisfied)
o Task Completion (% who finish task in one call)
o Consistency (variability in call length)
Usability Survey Grading Process
PG 10
Lessons Learned: A Case Study
• “Acme”: A regional Managed-Care Health Insurance Company. Customer Service available for Members and Healthcare Providers
• Available Tasks in Self-Service:o Members:
• Check Co-Pay amount and Physician Name• Order a replacement ID card
o Providers: • Check Claim Status• Verify Member Status and Co-Pay
• Initial Usability Survey, then Tuning, followed by 2nd Survey
PG 11
Initial Results1.0 How easy was it for you to accomplish your objective in this call?
Very easy 16% (85 Panelists) Easy 27% (146 Panelists)
Neither easy nor difficult 15% (79 Panelists) Difficult 21% (110 Panelists)
Very difficult 4% (22 Panelists) I did not accomplish my
objective in this call. 17% (93 Panelists)
2.0 How satisfied were you with your overall experience?
Very satisfied 12% (66 Panelists) Satisfied 33% (174 Panelists)
Neither satisfied nor dissatisfied
14% (73 Panelists)
Dissatisfied 29% (155 Panelists) Very dissatisfied 12% (66 Panelists)
No Response 0.2% (1 Panelist)
• Satisfaction Grade = D• Satisfaction Score= %VS – %D – %VD• Satisfaction Score= +12 – 29 – 12 = – 29%
• Call Completion Grade = D• Call Completion Score= % accomplished objective in one-call only• Call Completion Score= (100–17) x (0.83) = 69%
PG 12
After Tuning
1. How easy was it for you to accomplish your objective in this call?
Very easy 52% (285 Panelists) Easy 25% (136 Panelists)
Neither easy nor difficult 8% (44 Panelists) Difficult 6% (35 Panelists)
Very difficult 1% (6 Panelists) I did not accomplish my
objective in this call. 7% (37 Panelists)
No Response 1% (3 Panelists)
2. How satisfied were you with your overall experience?
Very satisfied 50% (275 Panelists) Satisfied 30% (165 Panelists)
Neither satisfied nor dissatisfied
8% (46 Panelists)
Dissatisfied 8% (46 Panelists) Very dissatisfied 2% (13 Panelists)
No Response 0.2% (1 Panelist)
Call Completion Grade = CCall Completion Score= (100–7)x(0.86) = 80%
Satisfaction Grade = ASatisfaction Score= +50 – 8 – 2 = +40%
PG 13
UI Overall Improvement After Tuning
Very satisfied 12% (66 Panelists) Satisfied 33% (174 Panelists)
Neither satisfied nor dissatisfied
14% (73 Panelists)
Dissatisfied 29% (155 Panelists) Very dissatisfied 12% (66 Panelists)
No Response 0.2% (1 Panelist)
Very easy 16% (85 Panelists) Easy 27% (146 Panelists)
Neither easy nor difficult
15% (79 Panelists)
Difficult 21% (110 Panelists) Very difficult 4% (22 Panelists)
I did not accomplish my objective in this call.
17% (93 Panelists)
Very easy 52% (285 Panelists) Easy 25% (136 Panelists)
Neither easy nor difficult
8% (44 Panelists)
Difficult 6% (35 Panelists) Very difficult 1% (6 Panelists)
I did not accomplish my objective in this call.
7% (37 Panelists)
No Response 1% (3 Panelists)
• How easy was it for you to accomplish your objective in this call?
• How satisfied were you with your overall experience?
Very satisfied 50% (275 Panelists) Satisfied 30% (165 Panelists)
Neither satisfied nor dissatisfied
8% (46 Panelists)
Dissatisfied 8% (46 Panelists) Very dissatisfied 2% (13 Panelists)
No Response 0.2% (1 Panelist)
Initial Results
After Tuning80% Call Completion
69% Call Completion
+ 40% Satis-faction Score
– 29% Satis-faction Score
PG 14
• Look for the Red Flags of Voice Recognitiono Low satisfaction and call-completion scoreso Low voice-recognition rating scoreso Check Complaints in free responseso Observe if “can’t hear” problems for low-
score panelists
• Work with Developer and IT supporto Is speech level strong enough in the IVR?
Check switch gain levelso Then check speech detector parameterso Then check recognition confidence
thresholds and grammars
Best Practice 1:Catch Recognition Problems Quickly
PG 15
Voice Recognition Issues• How well or poorly did the system recognize your responses
when you spoke the answer to questions?
• Fix: Increased digital gain from host switch
Very well 47% (257 Panelists) Well 31% (167 Panelists)
Neither well nor poorly 8% (46 Panelists) Poorly 10% (57 Panelists)
Very poorly 3% (16 Panelists) No Response 1% (3 Panelists)
Very well 11% (59 Panelists) Well 21% (110 Panelists)
Neither well nor poorly 12% (66 Panelists) Poorly 34% (183 Panelists)
Very poorly 22% (116 Panelists) No Response 0.2% (1 Panelist)
Initial Results
After Tuning
• How quickly or slowly did the system respond to your spoken answers?
Very quickly 46% (249 Panelists) Quickly 41% (225 Panelists)
Neither quickly nor slowly
8% (46 Panelists)
Slowly 3% (16 Panelists) Very slowly 1% (8 Panelists)
No Response 0.4% (2 Panelists)
Very quickly 16% (88 Panelists) Quickly 44% (238 Panelists)
Neither quickly nor slowly
21% (113 Panelists)
Slowly 13% (72 Panelists) Very slowly 4% (23 Panelists)
No Response 0.2% (1 Panelist)
PG 16
• Look for the Red Flags of Prompt Confusiono Low satisfaction and call-completion scoreso Low ‘What-to-Speak’ scoreso Listen for caller hesitations for low-score
panelists
• Work with Dialog Designero Let callers know ahead of time that they can
speako Reword prompts; Callers appreciate clear
choiceso Give Touch-Tone options when repromptingo Coach your voice actor
Best Practice 2:Spot Prompt Clarity Confusions
PG 17
Voice-Prompt Issues
• Was it clear what you needed to select or say at each step of the call?
Clear for all steps 58% (315 Panelists) Clear for almost all
steps 29% (160 Panelists)
Clear for some steps 8% (45 Panelists) Clear for only a few
steps 4% (21 Panelists)
Clear for no steps 1% (4 Panelists) No Response 0.2% (1 Panelist)
Clear for all steps 27% (142 Panelists) Clear for almost all
steps 27% (147 Panelists)
Clear for some steps 24% (129 Panelists) Clear for only a few
steps 16% (88 Panelists)
Clear for no steps 5% (28 Panelists) No Response 0.2% (1 Panelist)
• How appropriate or inappropriate was the speaking style and voice of this service?
Very appropriate 58% (319 Panelists) Appropriate 36% (198 Panelists)
Neither appropriate nor inappropriate
4% (20 Panelists)
Inappropriate 1% (5 Panelists) Very Inappropriate 0.2% (1 Panelist)
No Response 1% (3 Panelists)
Very appropriate 32% (172 Panelists) Appropriate 52% (276 Panelists)
Neither appropriate nor inappropriate
11% (61 Panelists)
Inappropriate 3% (17 Panelists) Very Inappropriate 1% (7 Panelists)
No Response 0.4% (2 Panelists)
Initial Results
After Tuning
PG 18
Voice-Prompt Issues
• Fix: Clarify prompt choice wordingo Initial: “Are you a member, provider, or a
non-member looking for information?”o After Tuning: “First, tell me who you are:
a member, a provider, or a non-member. ”
• Fix: It’s not just what you say, but how you say it.o Coach your voice talent for proper
inflections.
PG 19
• Look for the Red Flags of Call-Flow Frustrationo Low satisfaction and call-completion scoreso Low Enough-Choices scoreso Check complaints about inflexible systemso Listen for caller frustrations for low-score panelists
• Work with Dialog Designero Look for easy ways to complete repetitive taskso Provide an easy exit strategyo Leverage the spoken language instinct
Best Practice 3: Spot Call Flow Frustrations
PG 20
Call-Flow Issues
• When you were given menu choices, were you given too many, just enough, or too few?
Too many choices 10% (55 Panelists) Just enough choices 85% (466 Panelists)
Too few choices 4% (22 Panelists) No Response 1% (3 Panelists)
Too many choices 11% (61 Panelists) Just enough choices 74% (395 Panelists)
Too few choices 15% (78 Panelists) No Response 0.2% (1 Panelist)
After Tuning
Initial Results
PG 21
Call-Flow Issues
• Fix: To clarify options, Provide Anchor, Split Main Menu choices, and Add Grammar Synonymso Initial: o “Alright, I’m going to tell you the things I can
help you with. When you hear the right one, just say it… Verify member status and office co-pays. Get status of a claim. Get member ID cards. Change PCP. Or, order forms and literature.
o After Tuning: o “Alright, Main Menu. Please say one of the
following options at any time … Verify member status. Check PCP co-pay. Get claim status. Order ID cards. Change PCP. Order forms. Or, order literature …”
PG 22
• Look for the Red Flags of Task Differenceso Low satisfaction and call-completion
scores for Specific Taskso Check complaints about complicated taskso Listen for caller confusion in low-scoring
tasks
• Work with Dialog Designero Look for easy ways to complete repetitive
taskso Make sure task instructions are clearo Provide an easy operator exit strategy
Best Practice 4: Spot Task Differences
PG 23
Task Differences
• Relative Task Satisfaction can change after tuningo Can’t Hear problem can swamp
other UI task issueso Claim Status Sat. low for business
reasons (no shortcut for multiple claims)
PG 24
Task Differences• Fixes
o Improve recognition to shift all scores up and reveal other UI issues
o Future: Give clearer exit menu optionso Future: For Claim Status, offer stream-
lined repeat-claim function
-75 -25 25 75
Office CoPay
ID Cards
Verify Status
Claim Status
Task
Satisfaction Score
Satisfaction AFTER
Satisfaction BEFORE
PG 25
• The Usability Survey method is very effective for tuning applications prior to pilot production
o Collect hundreds of calls and analyze results efficiently
• Recognition/Prompt/Call-Flow issues are revealed quickly in Usability Survey
o Listen to calls for which various ratings are low
• Task differences show the effect of Task Consistency, Complexity, and Business Rules on Usability Quality
o Longer tasks require more call-flow efficiency and friendly hand-holding.
o Transfer to Agent for business reasons can sometimes lower satisfaction if callers must first spend a long time in Self-Serve, or if transfer takes place with no explanation.
Conclusions
PG 26
Thank You!