10. Run sessions in English
Will richer, better
educated, younger
respondents give you
good enough feedback?
11. Run sessions in English
Will richer, better
educated, younger
respondents give you
good enough feedback?
India, Philippines, Hong
Kong
Germany, Netherlands,
Sweden
15. Sharing analysis with your local team
Close analysis while
you’re still in the field
16. Sharing analysis with your local team
Close analysis while
you’re still in the field
M T W T F
Research
sessions
Research
sessions
Analysis
workshop
Brief local
team
Pilot
sessions
Debrief Debrief Debrief
18. "I am worried about the environment. So I take
my 4WD off-roading in Tibet to enjoy the
environment before it is ruined."
19. Suraj monitors phone resale price analogous to used cars
Suraj keeps his phones for about 8 – 10 months.
When he buys a new phone, he sets a target price that he
wants at resale
He bought a Blackberry Torch for 21K Rs and will sell it
when its resale price is below 13K Rs. (The day we met
him, it was worth about 18K Rs)
Previously, he bought a BB Curve for 14K, but within a
month its price dropped to 10K so he sold it immediately
because “the rate was coming down very fast”
21. Bridging the Language Barrier
Engage with product
Small research that is strategic
Share analysis responsibilities
Eliminate exotic explanations
http://www.andrewharder.com
@thevagrant
Notes de l'éditeur
A few core ideas about how you can approach research analysis In international contexts
What you might want to get out of it
And how you can approach it
Not unifying, instead craft
At Nokia, I was head of research for emerging markets software
That meant I was responsible for making sure Usable, Desirable and Competitive
To emerging market consumers
Learned a lot from applying ethnographic approaches in my work
If that sounds glamorous, I guess some parts of it were fun
But a lot of it was like this:
You’re jetlagged, sitting all day in a dark room
Struggling to follow what’s going on - Pixellation isn’t intentional but it reflects the fuzzy feeling
List of problems
The translator says crazy things like “I don’t like it because this music player isn’t free.” and you see the moderator nod and move on
WTF, the music player isn’t free?
All these puzzles sit in your head
You want to absorb everything that you can
But it feels like strange behaviour is just pulling on a piece of thread
And understand the total difference that people have
There is so much that you don’t know
So – how do we analyse this?
Pinned Butterly fly approach
The natural ending place of the naturalistic approach to understanding users in their own language and context.
Informed by anthropology, we kind of want to write a book about what we’ve found
Example: Design a new Hindi keyboard. Step 1: Understand Indic languages
User behaviour captured in all its specificity
Richness of context celebrated
We keep pulling on the thread of a single user thing they’ve said until we get someplace
I’ve done so many of these, and I’ve seen my friends do so many of these too.
Had to work so hard to get here
Every one thinks you just go on junkets
Pressure to get results
When you come back with these pinned butterlies in your baggage and you get into meetings, and try to convey the importance of them you can get glazed eyes
If you really want to see results, you can get pushier and demand what changes you’re going to see as a result
----- Meeting Notes (01/07/2013 18:14) -----
Set up anthropology earlier
Resulting in the dreaded shrug
This sparks fear in client-side researchers like you wouldn’t believe
The problem is that pinned butterflies belong in a museum and that product development looks more like this
Why is our internal decision making so important in international research?
If our job is to capture how our users see the world, why should internal chaos matter?
What are the implications for analysis of this?
No research gods talk about this stuff
Malinowski didn’t, Jan Chipchase doesn’t
Make this decision conscious
My contention here is that the lens of product making is essential to delivering research in a design environment
It increases your relevance to the decision makers
And it gets you out in the field again
Now I’ve done plenty of pinned butterfly pieces of research
There are probably contexts where they are still valuable
But the story of my years in Nokia is moving away from the museum and into the battlefield
In international UX work, we have inherited an set of approaches from anthropology
I want to rethink how some of these are applied
So how can we structure our analysis in this context?
Lower our expectations about NOVEL or COMPLEX research approach and analysis techniques, in order to have a bigger impact
And get to the field more often
The greater impact you will get is incredible
If you have stakeholders in the field, this is the single biggest thing you can do to have a better impact
Whether you can do it depends on the answer to a specific question:
For many products, the answer will be “it’s good enough”
For many products, this is the exact group that people target
These countries all have populations greater than 10% who speak English
Two reasons to go to international markets
There is a specific market that is so important to us
We want to make sure we have an international perspective
If Poland is your focus, then you have to go to Poland. But if you want to just explore Europe a bit, then Germany will do.
WHEN NO
When we were recruiting tech-loving 20yos, in India we recruited English speakers. In China we didn't.
CHINA WAS SUCH AN IMPORTANT MARKET that we didn't want to substituite it
But the extra work meant that we went there less
When important parts of the stimulus will be in local language
Or testing final translations
Important thoughts on responsibility
Your local contacts have privileged information
They already understand much of the local context
So we naturally rely on them
But sometimes we can get our approach wrong
An informant is a broader cultural perspective
Change from a provider of comments and numbers to a true informant about the product
They are now part of the team to improve the design
ACCOUNTABLE
Not to hand over a lightly-edited transcript or stay locked up in an ivory tower
We have to BRIEF THEM ON THE DESIGN
make sure they understand the design intent and conceptual model
For complex areas like when we tested dual-SIM functionality, this took a long time to explain
QQ Music example
“The music player isn’t free.” WTF
We identified a lot of usability issues in that session, so the temptation was just to ignore it and move on
But after discussing with our moderator, we learned that “Free music stores” are included in local music players on Android
Nobody uses the native player
One approach is to convert the WTF moments into questions for them
Learned I needed TO INTERVIEW THE MODERATOR after every session
Especially for the desirability and competitiveness concerns
This takes a long time, and you only begin to scratch the surface
The thing about cultural insight is that it is never-ending
We don’t need findings, WE NEED ARGUMENTS
We need arguments from the field that we can use
And arguments only come from closing the analysis
Debrief every day.
Usability issues will tend to appear from the behaviour
Competitiveness and Desirability issues will only appear from discussions with the moderators
Workshop at the end that I led, where I interviewed them about what we found, screen by screen.
At the end of the workshop we have a list of all the usability issues, all the competitiveness issues and all the desirability issues
We have gotten to the bottom of the WTF pile
The WTF moments were translated into issues, or contextual needs
Final idea
A fact of life with research in general, and research in other cultures in particular is that people say crazy things
The temptation is to deliver an explanation of this that rests on
“Chinese people don’t understand the environment.”
Example – When are we going to tell them what environmentalism is really about
This is the equivalent of giving up
Add China link
The basic question we ask is: Is this behaviour exotic?
Give your partners enough information so that they could predict their own behaviour in that situation