 |
|
|
|
Domains & Technologies
|
|
Since its inception, HFI has done extensive human factors work in almost
every domain and technology, such as:
|
|
|
 |
|
|
|
Ease of use pays off financially. Callers really hate to suffer through
unclear IVR menus. And when they hit the "O" button for a personal conversation,
it costs you money. (Your customer service reps don't work for free.)
Neat techniques exist to simplify complicated stuff. We use them.
|
| Benefits to count on:
|
- Short menus
- Easy comprehension
- Consistent terminology
- Well-organized menus
- Categories that make sense
- Reasonable number of steps
|
| |
|
|
| 1. |
Do you want to cut your CSR costs?
|
Are your CSRs (Customer Service Representatives) doing low-level
"busy work" like telling a customer his or her current account balance?
At $25 to $40 per hour loaded labor rate, it's expensive. A voice
response unit (VRU) can handle easy tasks with pre-recorded messages
or message components. Is your organization customer-sensitive?
Research shows if you keep the menu short and sweet, callers accept
it. Long or complex, you turn them away. See our article on IVR.
|
|
| 2. |
Perform task analysis and user characterization.
|
We find out the frequency of use for different activities. We learn:
who calls you? Are your customers older (and techno-phobic) or Gen-X?
(Research shows that attitudes vary strongly by age!) We...
- Rank task options by frequency of access
- Check for shortcuts and task flow re-engineering
- Don't make caller "dig" for the menu option if the
task demands a CSR
- Keep most-used options at top
- Consider "expert mode" for frequent callers (it happens!)
|
|
| 3. |
Design the "high-level architecture" (HLA) and test.
|
We don't use hi-fi prototyping too early! We test your high-level
script (first two menu levels) using the "Wizard of Oz" method.
We play the Wizard and read out loud from your script. The subject
listens as though we were the VRU. We take the action the subject
suggests and read the next option. Remember: we tell the subject
we are testing your script, not them! (This relieves performance
anxiety.)
|
|
| 4. |
Based on test results, fix navigation and scripts. Test again.
|
We protect your investment in the script. We test it again, after
revisions. We listen to subject comments. They tell you something.
However, we also beware of "experimenter demand" effects
in which a small percentage of subjects tell us they love the system
– even though they failed to choose the correct menu options!
|
|
| 5. |
Consider your VRU standard call flow structures.
|
We avoid "intuitive design". Much VRU research shows
optimum designs for different situations. We can advise you on a
VRU standard or on your existing call structures. Here's some call
structures we use:
- Pick One
- Scan and Pick One
- Pick Many
- Search for Selection
- Segmented List Playback
- Nonsegmented Playback
- Control/Trap Menu
- Greeting Menu
|
|
| 6. |
Test detailed design and iterate.
|
Now we're coming into the home stretch. We pick three or four or
more full task sequences that represent challenges to the caller.
We design the entire menu sequence for those. We test to verify
that your standard call structures work for your user group. We
revise and test before continuing with the other task sequences.
See our case study on testing.
|
|
| 7. |
Check out your "talent."
|
The audio business calls your voice provider "the talent."
We help pick the right talent. The talent's voice communicates what
your organization is all about. We survey representative subjects
with sample talent voices. See our case study on testing. Research shows callers want
to hear a voice that is...
- Practical
- Intelligent
- Courteous
- Efficient
- Straightforward
- Methodical
|
|
|
 |
|
