Presumptive Design: Cutting the Looking Glass Cake
1. The Art of Prototy ping
Presumptive Design,
or Cutting the Looking-Glass Cake
Leo Frishberg > Tektronix, Inc. > leofrish@acm.org
SPECIAL SECTION
In the generally accepted approach to User exposing your assumptions to your
Centered Design (UCD), the designer/inves- end users and having them react
tigator researches the needs of the target • The faster you go, the sooner you
population, analyzes and transforms the know—do all of these things quickly
raw data, and then synthesizes a solution to learn how wrong you are sooner
ultimately reviewed by users who deter- • Iterate, iterate, iterate—do this
mine its fitness. Reducing the “gap” process as long as your stamina and
between the analysis and synthesis steps is budget provide
key to improving the outcome of design Each of these principles establishes a
activities [2]. frame of mind about the proposed solu-
In Lewis Carroll’s Through the Looking tions that keeps the designer open to the
Glass Alice faces a puzzling and paradoxical needs of the user. The process is especially
approach to serving cake [1]. A process I effective with small cross-functional teams
call Presumptive Design turns the tradition- facing limited time frames because it elicits
al practice of UCD upside down in a simi- two responses simultaneously: reaction to
larly paradoxical way. The designer, with the specific design solution and expression
minimum input about the requirements for of user requirements.
the project, creates a set of solutions, puts Rapid prototyping is fundamental to pre-
them in front of the target user, collects sumptive design as it provides an effective
data about the fitness of the designs, and means of quickly expressing the design
then performs analysis that feeds into an team’s assumptions and intentions. Users’
iterative cycle. reaction to artifacts simultaneously uncov-
Presumptive Design is based on five prin- ers their requirements while exposing the
ciples: team’s biases.
• Design for failure—expect your solu-
tions to always be off-target Risks
• Create, discover, analyze—create your The method is not without risks. Listed
ideas first, let users show you your below are some of the common risks that,
errors, analyze for the next round while not unique to presumptive design,
• Make assumptions explicit—embrace are more likely to occur because of priori-
your human egocentricism by explicitly tizing synthesis over analysis.
: / 18 i n t e r a c t i o n s / j a n u a r y + f e b r u a r y 2 0 0 6
2. “You don’t know how to manage
Looking-glass cakes,” the Unicorn
remarked. Hand it round first,
and cut it afterwards.”
Believing the solution is correct. questions that the design-
Accepting a solution because it “feels er had not anticipated. Messy prototypes
right,” being blinded by your creation, are likely to distract the user from your Forum brought a model for
and/or defending the solution to the cus- objectives; they are equally likely to provide users to react to.
tomer, sabotages the process. In presump- opportunities for conversation. Having a • Industrial designers (IDs) are more like-
tive design, you always believe you’re cross-functional team with experienced ly than interaction designers (IxDs)
wrong; remembering this principle reduces designers and usability professionals is a (and non-designers) to engage in pre-
this risk. key component of the presumptive-design sumptive-design behaviors. During the
Focus on the wrong solution. Because process. “analysis” phase, IDs sketched, proto-
this approach is “inductive”—you create an typed, and played around, while IxDs
example of an idea—you can’t know from Testing the Process made lists, drew charts, categorized,
the outset whether you are on the right Twice recently—at the CHI2004|ICSID and analyzed.
track. Rapid prototyping generally creates Forum and SEC05 (see the following article • Interaction designers and usability pro-
unattractive artifacts, creating a natural by Nancy Frishberg)—I had the opportuni- fessionals (regardless of years of expe-
feedback loop that helps reduce the team’s ty to explore whether presumptive design rience) prefer to present rough proto-
faith in the design. was idiosyncratic to my practice or one types to users rather than allow users
Lack of solution coverage. Design shared among professionals. In the first to engage directly with the artifact. At
resources are a precious commodity; using event, we presumed that the experience SEC05, none of the teams provided
them to develop one solution reduces their level of the participants would lead some users with prototypes, opting instead
availability for investigating alternative solu- teams to use presumptive design; in the to present, explain, demo, or defend
tions. Here, too, rapid prototyping quickly second event, presumptive design was their designs.
creates widely variant solutions, improving explicitly required. • Inexperienced designers, and those
coverage of the “solution space.” from related disciplines, expressed fear
Loss of User Focus. Users may react to Lessons Learned about exposing their assumptions so
elements in the solution that were either We learned several lessons from these directly to users.
unintended or not yet developed. An object two events: • Participants (at all levels of experience)
implies much more than the minimum • Even designers with more than ten were able to rapidly create prototypes
inputs used to create it. This risk can only years’ experience don’t use presump- that expressed the key assumptions
be mitigated through experienced facilita- tive design approaches. We were sur- underlying their designs. All teams
tion, often requiring a multidisciplinary prised to see that only one individual successfully created prototypes in time
team. Low-resolution prototypes raise user from one team at the CHI2004|ICSID for their presentations to the group.
i n t e r a c t i o n s / j a n u a r y + f e b r u a r y 2 0 0 6 : / 19
3. The Art of Prototy ping
Conclusions that revealed their assumptions, whether or Presumptive design is not for the
Presumptive design, or serving the cake not they used the artifacts to elicit user faint of heart. Putting your assumptions in
before you slice it, is a powerful way for feedback. Teams with less-experienced front of users can be a frightening
designers to make their assumptions explic- designers had greater difficulty shifting prospect. In addition, the presumptive
it while providing customers opportunities from analysis to synthesis, suggesting that designer must be comfortable synthesizing
for rich engagement and feedback. The design experience is a key component in quickly while simultaneously remaining
SPECIAL SECTION
approach can accelerate the data-gathering building artifacts. uninvested in the result.
process at the same time it builds multidis- Facilitating user interaction with Presumptive design, by its reliance on
ciplinary teams. Some conclusions drawn junk prototypes is necessary and chal- rapid prototyping, is a fast way to find out
from the two workshops helped broaden lenging. Users, especially the active seniors how wrong you are and what users really
my appreciation for the pros and cons of in our case, are comfortable responding want. The process permits multidisciplinary
the approach. negatively to proposed solutions. Low-reso- teams to work quickly and collaboratively. It
Users respond more intensely to lution prototypes challenged teams to elicit exposes designers’ assumptions for review
objects than they do to questions. At constructive user feedback; real products by target populations while allowing ana-
the CHI2004|ICSID Forum, users demon- were much easier for users to discuss posi- lysts opportunities to capture users’ reac-
strated the advantages and disadvantages tively. Teams with experienced usability pro- tions and implicit requirements.
of their own artifacts. At SEC05, users sur- fessionals skilled in eliciting user feedback
prised the teams with their (generally will be more successful than those without.
Prototypes must be designed. With its REFERENCES 1. Carroll, Lewis, Through the
intensely negative) responses to the proto-
Looking-Glass and What Alice Found There, The
types. reliance on low-cost materials and rapid MacMilllan Company, New York, London, 1899,
p152 2. Wood, L. A., User Interface Design, Bridging
Building junk prototypes helps illus- turn-around, presumptive design requires
the Gap from User Requirements to Design, CRC
trate assumptions quickly. Both groups experienced design- Press, 1997.
created objects ers. The presumptive ABOUT THE AUTHOR Leo
designer must care- Frishberg is a user experience
fully consider the architect for Tektronix, Inc.’s
desired attributes of Logic Analyzer Advanced
Development Group. For over 20 years,
product, the specif-
Frishberg’s professional life has focused on
ic characteristics of
enhancing the user experience with architectural,
the p ro t o t y p e software, hardware, and Web projects. For the
object, the objec- past five years, Frishberg has served as both pro-
tives of the ses- gram chair and executive director for CHIFOO
sion, and the (Computer Human Interaction Forum of Oregon),
the Portland-based local chapter of the SIGCHI.
anticipated
Frishberg is a licensed architect, with an M.Arch
interactions
from SCI-ARC and a BA in Environmental
during the Planning from the University of California, Santa
feedback ses- Cruz. Nancy Frishberg is his sister.
sions.
: / 20 i n t e r a c t i o n s / j a n u a r y + f e b r u a r y 2 0 0 6