2013 Korean Cognitive Science Conference
3rd Call for Participation
Sogang University, Chung Ha-sang (J) Hall,
http://CogEng.net/CogSci2013
May 25, Saturday
Korean Cognitive Science Conference (2/3)
USER MODELING — Empathizing With and Understanding User Behavior
Song Young-il
Evolving into a form like a development language
People are, by nature, blind to everything outside the area they care about (the gorilla experiment)
We need to reflect the uneven, irregular nature of people
Alan Cooper (using personas to ground things in real people)
1. Personas
A persona is a "fictional character (user model)" that represents a type of "varied user behavior pattern" within a specific group that might use a particular product or service.
(The opinion of someone higher-up, "my wife says so"
— helpful for focusing decision-making conversations; lets a project steer in a fun direction and lets people have different conversations on the same footing.)
people inspire empathy, bullet points don't
Humans are drawn to emotional projection, empathy, and immersion.
User modeling — understanding — > personas
Modeling (: simplification, generalization, reference-point role, process of typifying...)
We model in order to apply large amounts of data to design. Model things so they're easier to see and understand without losing detail.
So isn't it better to make them feel like real people?
Key elements for a convincing illusion
attitude
behavior
context
details
excerpt: a signature line they'd say
foto: a face
goal
(pain point)
-> In the end, it's about solving a problem.
Process
- Organize the data
- Analyze the data
- Profile the data (like a real person)
designers are not users. but
A designer is just a user.
what is good personas?
Something you can emotionally project onto and empathize with
Comparing differences
Drawing out connections
Focus and selection
1. Comparing differences
Compare behavior between users and pay attention to the differences. Special value emerges from small differences. (One person alone can't have personality. You need at least two, so that comparison becomes possible — that's where individuality comes from.): Relatively doesn't clean. Relatively tidies diligently.
(Relative comparison)
Individuality emerges from the relationships between different behaviors that are all respected. Don't jump to conclusions about someone's other traits based on one characteristic
: Eats a lot -> is diligent
— could be the case, could not.
The reason
Humans don't always behave consistently
-> We need to grab the distinctive points
-> We need to capture the individual characteristics one by one
Additional tool usage
critical characteristics
: the number/count should come across as a person
2. Drawing out connections (inference through grouping)
No matter how much research you do, you'll end up interpreting it your own way, so classify the patterns
(Reorganize while re-examining the research and determine priorities)
: Infer causal relationships between factors that have no apparent causal link — like laundry type, fabric sensitivity, and degree of staining!
: Pay attention to the minority of users.
: Obvious personas fail. (People who eat a lot are fat — X)
The three sheets below contain the same information, only expressed differently. (Each one profiled like a real person)
* A persona is not a profile.
3. Selection and focus
Various customer personas -> Strategic choice is required.
What ends up explaining a product/service better is not who you target, but who you left out.
