2. Usability on the run
• Championing usability
• Click analysis of HR services
• How much can we actually improve a website
• An easier way to get users to test
• Designing with the users, not for them
• Capturing user intentions
3. Championing usability
4 UX specialists @ NorwegianTax Administration
• Project work
• Marketing usability
• Usability testing
• Interaction design
• Ad hoc usability advice
• Lecturing
4. Living the usability lifestyle
• Communicating the UX message to people who
can change it
• Empowering non-experts
• Downsizing the effort to small nibbles
• Helping improve the usability within the limits of
project budgets
• Liason with other govt. agencies to share
usability competence and resources
• National UX standardization - ELMER
6. Click-analysis of HR services
• Click-analysis carried out before and after
redesign
• Redesign of information architecture and start-
page for HR Services (Norw.: Ansattguiden)
• Measured no. of clicks and time on task
• Reduced time on task significantly
• Was a success
• Increased use of services after redesign
• The term became current – people started asking for
click analysis
7. Comparing time on task
Tasks for Ansattguiden
Before
redesign
After
redesign
Simplest task 61 sek 17 sek
Most difficult task 173 sek 62 sek
Average for all tasks 140 sek 32 sek
Total cost per year if each employee has
2 queries per month
2 175 049,- 496 269,-
8. HOW MUCH CAN WE ACTUALLY IMPROVE?
Raising awareness
9. How much can we actually improve?
• Can we reduce the number of calls to Tax
Information Services by improving
communication via our web-site?
• Survey at tail-end of 105 phone calls to helpdesk
• ”Why didn’t you use the website?”
• We could pinpoint particular issues that need
attention immediately
10. Other answers
4 %
Can't be resolved on-
line
15 %
PIN problem
14 %
Don't know
15 %
Poor usability
21 %
Internet not an option
31 %
Why didn’t you use the website?
Caller didn’t know that
she could find it on the
website
•Prefer talking to a person
•Haven’t got internet access
•Haven’t got web-access right now.
•Haven’t got PIN codes
•Have lost them
•Can’t remember my password
Caller wants to do things that
aren’t posible via the website
•Found the subject, but not the
answer to my specific query.
•Found the subject, but didn’t
understand. Too difficult
11. How much can we actually improve?
• Improving services on our website will potentially
help 1/3 of the callers
• Isolated the single most used function that users
had problems with (PIN code)
• Improved navigation and content
• Changed 800 80 000 (phone no) to
www.skatteetaten.no (URL) on paper tax-card
• Increased awareness for usability in the
helpdesk organisation
• Recognition that they can contribute to improve
systems
12. AN EASIER WAY TO GET USERS TO TEST
Recruiting test subjects
14. We went to IKEA!
• Obtained permission from IKEA
• 2 emails back and forth
• Card table in IKEA exit lobby
• Pulled them in with poster
• Recorded with Mac (Silverback)
• No time / expense recruiting
(saved 3 weeks 30 000,-)
• Self-selected test subjects
• Tested 20 subjects
• 10 each at IKEA East Oslo and IKEA West Oslo
17. Turboprototyping
• Prototyping method with user participation
• Challenges the notion that ’Design by committee’
doesn’t work
• Straight from high-level requirements to rich
wireframes
• Presented at Interact’99 and NordiCHi 2000
• used since the dotcom era
18.
19. Designing with the users – real time
• Whiteboard design – empowering users
• Using rich wire frames – detail is of the essence
• Allowing them to err – not forcing the usability
message down their throats
• From requirements to design in 2 workshops (+
a bit of work in between them)
• Disbelief at opportunity to design ’our’ application
• WYSIWYG
21. Capturing user intentions
• Information about feedback at login
• Text box for usability feedback available in all
screens with caption Difficult? Illogical? Errors?
• ’Tell us what you were trying to do’ embedded in
all error messages
• Collected feedback approx one week.
• Over 100 test users
• Feedback was timestamped, with user name for
follow-up
24. Capturing user intentions
• Really low cost evaluation
• User reactions, not our opinion
• Getting effort focussed
• Used similar approach on Skatteetaten.no
• Only displayed on content page found via search
engine
• 1% feedback 70% negative, 30% positive