Mental models workshop – Eric Fain

We cross our fingers that the design wil be successful. We worry we might have missed some part of the ‘whole user experience’. Mental model is not specific to a tool.

Is about talking with your customer, is cheap to talk with people. As you start to talk, the cost of changing your mind increases.

Evolution of the UX discipline:

Cumulative impact; 1970 utility, 1990 usability, 2000 profitability 2010 we are looking fwd strategy.

See pictures used in this workshop

When to use mental models:

  • Usually after product is out there -> redesign
  • Often during incubation for start-ups

Mental models are not used for:

  • Evaluating usability. is to try to understend what (how) our customers are thinking
  • Showing cyclic relationships
  • Depicting detailed task flow
  • Creating art (composing…)

Exercice: Sketch a mental model of going to the movies. Start evaluating how people think about problems. Describe behaviours of persons.

thumbnail image for mental model example

Other mental model examples at

Choose your audience wisely

Examples of different behaviors and philosophies:

  • Explore cultures and places (/ act like a local, like the character, flirt with living this way myselff, gain new understanding
  • Challenge my skills (eg. ski trips, climb a mountain, river rafting… / ensure equpment functions, train for event…)
  • Experience sth rare  (/ be envied feel awe, feel the spirit)
  • distract & entertain myself (cruise ship, destination resort / dull my senses, relax for a while..)

Once you define people, yo choose to support some of the groups of people. Can we solve for both audiences with one application? Or do we need two applications?

(For obtaining this, we talk with some stakeholders.)

Example with one business that sells computer products:

  • Pre-sales don’t own Sybase
    Research products
    Become familiar with Sybase
  • Pre-sales own Sybase
    Research products
  • Post-sales support
    Upgrade products
    Thought leader
    Stay current with tech.

We can sort those characteristics into: Evaluator, xxx, xxx

Audience segments are not the same as defining Personas.

We define: demographic, geographic, psychographic, behavioral. VALS

How many users do we need to interview?

  • Perhaps the biggest debate in user research practice
  • Remember: this isn’t science or mechanical engineering
  • My experience: 4 to 6 people per audience segment. If you tal, to more people you’ll start to get similar results that you already have.

Also look for..

  • Find people who are currently engaged in the task
  • Pick the people who will give you the best response
    • Will talk a lot
    • actually uses or do the thing we are studying (dont force people to remember)

Eg with original moviegoer segments:

  • social moviegoer (loves companuy at theater)
  • movie buff (make sacrifice to get good seats)
  • big fan (gets the models and action figures)
  • film purist (wants pure experience of the film)
  • make believe artists (wants to be in the story)
  • facilitator (takes someone else to movie of their choice)

There’s a story (passing interest, enjoyment, learn/see sth new, excitement, re-creation) and a craft. (low awareness, appreciation, aficionado, technical knowledge), and companionship (alone, social, enabler) [IMAGE]

In this range, you decide in which kind of customer are you going to focus.

Scoping your project

Ask stakeholders about the project, history function. We often try to study everything in other thing is we find a specific idea that stakeholders didnt even know that that happens. ??

Make a Table with group of tasks, and give an idea to those groups of tasks.

Select a level of scope to explore

  1. Home depot store – como leroy merlin pero con electrodomesticos- (electicians, plumbers, homeowners)
  2. Home depot homeowners (appliance, bathroom rug, irrigation, bathroom renovation)
  3. Home depot bathroom renovation

State your scope (examples of three different organizations)

  1. How do prospective customers make a purchase decisión?
  2. How do employees maintain their contractual relationship with the company?
  3. How do patients deal with a diesease they have just been diagnosed with?

Prepare PROMPTS to remind you of the workflow to explore


  • Stakeholder objectives transcripts
  • Task brainstorm from task-based audience segmentation
  • customer support reports
  • customer visit reports
  • requirements reports

eg. Prompts for moviegoer interviews -note there are not questionmarks-:

  • tell me about movies you see
  • before theater
  • at theater
  • beyond theater
  • others with you
  • relationship to movies

Review and interview

Hear stories -> feel empathy -> know how to support people

Dont think about the goal of the research.

Don’t introduce the vocabulary.

Get behind the emotions of the person. Sometimes youn need to go around a topic to obtain the real feels and behaviours.

Non directed interviews

Six rules:

  1. behaviors, beliefs, reactions (not preferences)
    Focus on exploring the behaviors, philosophies, feelings
    This is not preference/opinion research
    the key verbs are ‘do’, ‘believe’, ‘feel’, ‘think’. Avoid: widh, love hate, want.
  2. Open questions only
    why? what? when? where? how?
    closed questions = did? will? have?
  3. no words of your own
    refer toc concepts using hte participants own words
  4. follow the conversation
    similar to a dinner party conversation
    reminders,/prompts, no list of questions
    allow the interviewee to direct the flow of conversation (opposite of most radio and tv interviews)
  5. not about products
    dont investigate the application or service, or other specific tools
    if they metion how they use a tool, ask them why
    this is note evaluative reseracch, which covers how well a tool supports what someone wants to do 
  6. immediate experience
    concentrate on what the person had done recently

The hallway test (…)

Phone vs face-to-face. Definitely, is better face to face.

Practice, practice, practice:

  • How do I avoid leading the interview when I know the goal of the research?
  • How do I improve my ability to capture the right level of detail?

Roles conducting an interview (one person can play a few roles):

  • interviewer
  • transcriber
  • task analyst
  • sales representative
  • translator

Exercice: conduct an interview:

  1. form a group of 3 (or 4, but not 2)
  2. designate an interviewer, a participant and a coach
  3. coach gets paaper, pen and timer ready
  4. conduct an interview about ______ for 5 minutes
  5. coach writes…[]

Make the interviews in a natural place, not in a laboratory.

Analyze the Stories

Once you have the transcript, go through them, and give id to parts.

What to look for?

  • Bheavior – direct
  • Behavior – implied (when the person doesnt tell sth directly: The dogs need to go for a awalk at least once a day
  • Behavior – third party (my daughter walks the dogs before school in the mornings)
  • Philosophy – direct (I think dogs should be walked twice a day)
  • Philosophy – implied (I read that dailiy exercise is required for dogs)
  • Philosophy – third party (my neighbour believes lazy…)
  • Feeling – direct (i get grumpy when I …)
  • Feeling – implied (walking the dogs is relaxing)
  • Feeling – third party (my son rolls…)

How do I sumarize these?

verb + noun

Here are some examples of how to sumarize:

We are very interested in following developments in this work. It came to the attention of several of us in the department way before the business office would even consider it. -> Follow Developments in My  Industry (direct behavior)

Being a gadget guy, I went with the camera phone… it was only another $29. Spread that out over 36 months, thats not much. -> Justify cost of camera phone (implied behavior)

I don’t think it’s a very comfortable experience for either person. I mean actually trying to hook up two people who are single. -> Believe Hooking Up two people is uncomfortable (direct philosophy)
– dont use negative tenses for obtaining the idea –

I don’t like walking the dogs on Oak street – > Dislike walking we should have asked why
preference – why?
Because there is a dog who lives there that barks…. -> and then we obtain the idea

I hope to take the dogs to the dog park when it opens
desire – why?
I want my dogs to have ore social interaciton, plus i wnatn to wear them out a little more -> believe dogs do better when theyrs social

I hope everyone will clean up after their dogs …
Expectation – why?

those orders are printed out form the srpeadsheet roughtly about three times a shift -> print emails to include with orders roughly three timesa as day – print emails to include with orders (direct behavior)

Vague verbs: be, get/receive, have, manage, use, know. Is important to use the same verbs that the person has used, buto sometimes is better to concise.

Compund task: Draw picture of curtomer and their usage environment

Agile principles and ux design

  1. synchronize ux with development
  2. separate modeling from design
  3. increase designliteracy
  4. design on the whiteboard with developers, stakeholders
  5. strip documentation down to 
  6. […]
There are some ways to apply mental models research to the project: generative research + generative research + ideas & design = sprintA / generative research + ideas & design + generative research = sprint 
You can use other ways to collect information; write them down and use it for your researches. Do it when talking with your customers, try to understand them, and write down tasks and behaviors. Keep a log of how customers are using things, their feedback, etc.
Prioritize the possibilidties: put features into a ranking and priorize them by complex and hardness, and plot.

Indy Young’s at Google Techtalk:


Deja un comentario