Timothy Prestero

Better by Design
How Empathy Can Lead to More Successful
Technologies and Services for the Poor

Discussion of Design Case Narratives:
Rickshaw Bank
Solar-Powered Tuki
FGN Pump

Professional training becomes a lens through which we see the world. You can
imagine a dentist who finds it impossible to concentrate on an opera because the
singer has bad teeth. The writer Annie Dillard imagined that the happiest people
are constantly surrounded by opportunities to apply their expertise—for example,
those who study rocks or clouds. Designers are a pretty happy bunch because
opportunities to improve the human condition through design are everywhere. 它
becomes an obsession. 例如, 后 10,000 years of post-Ice Age develop-
ment as a tool-using species, I find it amazing that human beings are still creating
uncomfortable chairs (such as the succession of airplane seats where I wrote this
文章).

My goal in this article is to discuss the accompanying case studies: 这
Rickshaw Bank of Pradip Sarmah, the Solar Tuki of Anil Chitrakar and Babu Raj
Shrestha, and the FGN pump of Gustavo Gennuso. I look at them in the context
of design for developing countries, and from my perspective as the founder of
Design that Matters (DtM), a nonprofit design consultancy for social enterprise.
“Design” is a term used so broadly as to be almost meaningless. 在本文中, I will

Timothy Prestero is the founder and CEO of Design that Matters (DtM), a nonprofit
based in Cambridge, 马萨诸塞州. DtM collaborates with leading social entrepre-
neurs and hundreds of volunteers to design new products and services for the poor in
developing countries. A former Peace Corps volunteer and MIT graduate, Tim has
worked in West Africa, 拉美, and Asia. He is a Martin Fellow at the MIT
Laboratory for Energy and the Environment, a Draper Richards Fellow, and was
named an Ashoka Affiliate in 2004. His awards include the 2007 Social Venture
Network Innovation Award and the 2009 World Technology Award.

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023

Timothy Prestero

数字 1. Product Development Processes.

discuss design in terms of a problem-solving methodology that uses empathy to
identify and contextualize needs, and then translates those needs into a set of spec-
ifications to guide the development of new products and services.

DESIGN AND INVENTION: TWO APPROACHES TO SOLVING PROBLEMS

Every problem-solving effort begins by defining a general problem and a general
approach to solving that problem. In the context of the developing world, the gen-
eral problem could be summarized as “One billion people need X,” where X could
be defined as access to clean water, 电, 教育, and health care. 一
example of a general problem-solving approach is the development of new tech-
科学. In my work, I see two distinct approaches to problem-solving with techni-
cal innovation: the invention approach and the design approach. 数字 1 summa-
rizes and compares the two.

In the invention-centric approach, the inventor begins by specifying the technolo-
gy that they think will solve the problem. The inventor then attempts to fit the
technology to the problem through an iterative series of design refinements.
最后, having tweaked and changed the product into what the inventor hopes will
be a useful tool, they then go in search of a specific user group or market segment
for which the product is a match. This approach is also described as a technology
“push.”

It is not uncommon for product development efforts targeting the poor to
begin with the technology, for example by setting the goal of adapting light-emit-
ting diodes (LEDs) to rural lighting, or vapor-phase water purification to provid-
ing clean water in rural communities. With the technology approach in hand, 这
tinkering and refinement process begins—often without the inventor ever having
visited the places where the final product is intended to be used.

Our experience and the history of international development suggest that
invention—specifying the technology before identifying the market—is a very
risky approach to product development. An analogy is the tailor who makes a clos-

创新 / winter 2010

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023

Better by Design

et full of lime-green size 56 swallow-tail tuxedos. When customers with that pref-
erence walk in the door the tailor is going to make money—but what if they never
到达?

To anyone surveying the landscape of new products and services, “survivor
bias”can create the impression that product development is simply a matter of
having a great idea.1 It is possible to look at the success of the Solar Tuki and con-
clude that there is a tremendous market for LED-based, solar-powered lanterns.
The reality is that there have been many solar lantern projects, even in Nepal, 和
most have failed to achieve significant scale. 相似地, consider the many unsuc-
cessful attempts to scale technologies like the solar cooker and the improved wood-
burning stove.

Despite enormous budgets for R&D and marketing, most new products and
new technologies created for the commercial market in the industrialized world
fail to achieve scale.2 In poor countries, the statistics are even worse: we find very
few examples of successful appropriate technologies.

The design approach is an alternative to invention. The most basic difference
is that where invention often leads to a technology in search of a user (or a solu-
tion in search of someone who has that problem), design starts with the user and
then goes in search of the technology. In design, specifying the user involves con-
ducting direct and indirect research to define who the user is and what they
want—sometimes described as “consumer pull.” A combination of the user’s state-
ment of need and the constraints imposed by their environment creates the prod-
uct requirements, statements, and metrics that define the “victory conditions” for
product features and performance. Developing the product also involves design
迭代, but the direction of the refinements is dictated by the user require-
评论.

Tolstoy wrote that “happy families are all alike; every unhappy family is unhap-
py in its own way.”3 Successful products are all alike in that they represent a happy
marriage between product function and the user’s needs and circumstances.
Products that fail do so for a million different reasons.

A common failure of defining the technology before we have defined the user
is that the approach puts the burden of adaptation on the user. Who is to blame if
they refuse?

In the case of the Rickshaw Bank, the initial production run of the improved
rickshaw had a problem with the bicycle chain repeatedly falling off the cog. 这
team could have satisfied themselves by blaming the problem on user behavior—
perhaps the rickshaw-pullers were riding over too many bumps and that was caus-
ing the rickshaw frame to flex and the chain to fall off. They could have argued that
the rickshaw pullers simply needed to adapt to the technology: ride more slowly,
take fewer passengers, or avoid bumps. 反而, the project succeeded because the
team listened to the user and adapted their design.

Another common failure is that many inventors attempt to create a “one-size-
fits-all” solution to a general global problem: a new product that will solve the
problem in any context. 第一的, the “all-in-one” technology often requires that every

创新 / winter 2010

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023

Timothy Prestero

single user pay for features they don’t need and will never use. 第二, given that
context is so important, “one-size-fits-all” rarely works. In the case of the FGN
pump, based on their experience in Patagonia, the team assumed that a wooden
frame was a universal solution to the problem. In Patagonia, this was certainly
true—wood parts are easy to find and inexpensive. But when the team installed the
pump in La Puna, where wood is scarce and the local building material is made
from cactus, their assumptions did not hold.

Product development is a relatively straightforward process if the inventor is
the only customer. This helps to explain the success of the open source software
movement and the Linux operating system. Every volunteer programmer is also a
system user, so in a sense they are all simply solving their own problems. The slow
adoption of Linux outside of the developer community is in part a consequence of
non-programmers finding the system to be anything but intuitive.4

EMPATHY IN DESIGN,
OR WHY KIDS HATE GETTING CLOTHES FOR THEIR BIRTHDAY

The principal difference between the examples of design and invention described
above is the demand for empathy: the ability to imagine the world from someone
else’s perspective. If we want to make a child happy at their birthday party, 这
approach may not be to give them what we think they need (a new sweater), 但
rather what they think they want (a new toy).

The first component of empathy is the understanding that there are no “dumb
用户,” only dumb products. 例如, my cell phone, which was clearly devel-
oped by a bunch of engineers, contains dozens of amazing features that after two
years I have yet to figure out. Hearing this, the cell phone engineer might reply that
I am merely lazy—that all of the clever features buried in multiple sub-menus and
behind cryptic key combinations would be intuitive if only I would bother to read
and memorize the 45-page product manual. In great design, the burden is on the
innovator, not the user, to justify every quality and feature of a product.

The second component of empathy in design is the appreciation of context.
再次, an engineer might complain when a user shorts out his cell phone in the
rain, arguing that the device was only intended for use in dry weather (然而
absurd that claim). The qualities that define a product or service become either
virtues or liabilities as a function of context. Many products are developed with
embedded cultural assumptions that prove to be crippling liabilities in the context
of a developing country. Examples include general assumptions about the avail-
ability of spare parts and trained maintenance, or very specific assumptions about
a user’s familiarity with the standard iconography of consumer electronics.

The Rickshaw Bank case presents a series of excellent examples of applied
设计. Although a veterinarian, Pradip Sarmah was not moved by the plight of
rickshaw-pullers to offer them free veterinarian services. His approach was dictat-
ed by empathy rather than the resources he had at hand. The bank’s business
model is based on the insight that the rickshaw-puller’s lack of access to capital

创新 / winter 2010

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023

Better by Design

kept him from owning the tools of his trade. Several of the bank’s practices—the
daily loan payments that match the puller’s cash flow, the provision of a uniform
to increase dignity and brand identification, and the identity card that allows bank
customers to secure gas connections to their homes—are based on direct observa-
tions of, and empathy for, the rickshaw-puller’s point of view.

The Rickshaw Bank also illustrates how good design includes non-technolog-
ical considerations, such as matching loan payments to cash flow. 同样的道理
for the Solar Tuki. The Solar Tuki may be technologically superior to kerosene
lanterns, but Chitrakar and Shrestha understood that the kerosene lantern was a
better match for consumer budgets. A key innovation was the design of a financial
产品: a loan for the Solar Tuki whose payment schedule and terms were rough-
ly equivalent to the intended user’s daily expense of kerosene for home lighting. 经过
including a five-year warranty for after-sales service on the two-year loan, 他们
could effectively market the Solar Tuki as offering three years of “free” lighting.

DESIGN CAN’T SAVE THE WORLD (ON ITS OWN)

Like financial planning and weight loss, the principles of good design are much
easier to describe than to follow. 此外, having illustrated the virtues of the
design approach, I must now report the bad news: slavish attention to the details
of user and context does not guarantee success. Even in the industrialized market,
most new products fail. What makes good design so difficult?

事实上, a bewildering number of variables affect product success. The cost of
developing an understanding of user behavior and context follows the 80/20 规则:5
although it is possible to uncover critical insights very early on in the research
过程, full knowledge of the constraints imposed by the user, 环境, 产品
supply chain, and product lifecycle would require an infinite and therefore impos-
sible investment. From the perspective of design research, the hard part of the
80/20 rule is that we can’t control the 20 percent of the insights that we miss. 在里面
worst case, these missing insights, a combination of what Donald Rumsfeld called
the “known unknowns” and “unknown unknowns,” are the most important ones.
In my experience, the following four user and context constraints have proven

particularly difficult to characterize.

Alignment of Incentives

One of DtM’s most important and painful lessons is that needs do not necessarily
equal markets. We have learned the hard way that altruism is an excellent motiva-
tion for someone to do something once. What allows a program to scale is its abil-
ity to repeat success over and over again; if that is to happen, all the stakeholders
must be motivated as much by self-interest as by the desire to do good. It must be
clear from the beginning who will make money selling and/or maintaining the
干涉, and how everyone else involved in development and implementation
will benefit from the product’s success.

创新 / winter 2010

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023

Timothy Prestero

数字 2. Stakeholders in Product Development and Implementation.

此外, user enthusiasm for a great design does not necessarily guaran-
tee that it will be possible to solve the problems of product financing, 生产,
and distribution. As Michael Free at the global health nonprofit PATH says, 这是
necessary to identify, as early as possible, who will “choose, 使用, and pay the dues.”
数字 2 shows all of the stakeholders involved in the development, distribu-
的, and use of a medical device targeting infant and maternal health. Notice that
the designer is the stakeholder in the top-left corner. Given the huge scope of prod-
uct development, 生产, and implementation, design may be the “least
hard” part of the process. The challenge is that any one stakeholder in the process
can say no and effectively kill the project. 换句话说, every stakeholder is nec-
essary; no single stakeholder is sufficient.

In the Rickshaw Bank case, after many unsuccessful attempts to raise project
financing through bank loans, the team hit on the innovation of raising capital by
selling advertising space on the backs of their improved rickshaws. 相似地, 这
bank realized that the individual rickshaw puller’s financial success depended on
their ability to attract more business, which led to key innovations like the rick-
shaw puller’s uniform and the redesigned rickshaw’s improved ergonomics, safety,
and aesthetics.

The Solar Tuki program faced stiff competition from the existing supply chain

创新 / winter 2010

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023

Better by Design

for kerosene lanterns and fuel. Rather than competing with kerosene, the team
convinced existing kerosene vendors to sell the Solar Tuki as well. This gave ven-
dors a home lighting product to sell during the frequent interruptions of the coun-
try’s kerosene supply. Because the Solar Tuki diversified and strengthened their
商业, kerosene vendors—potentially their biggest competitors—became their
marketing ally.

Another excellent example of design elements informed by stakeholder context
is the way Solar Tuki is distributed to students through the program developed by
Ashoka Fellow Mahabir Pun. The team changed the distribution pattern so that
each student received a light of their own, but the school owned the collective
solar-powered charging station. This created a natural alignment of incentives
between the goal of Solar Tuki, to distribute lights, and the goal of the school, 到
increase student attendance. The key was that students had to return to school
every day to charge their lights for the evening.

In the FGN case, Gennuso describes the challenge of partnering with NGOs to
distribute his products in regions far from the FGN production centers. 虽然
the partnership had many advantages on paper, providing water was not central to
the mission of these partner NGOs. This led to disappointing performance in
terms of product marketing, sales, and distribution.

Cultural Expectations: We See a Problem, but Do They?

A product or intervention intended to improve a user’s quality of life will not nec-
essarily change a culture’s view of what’s “normal.” For example, people in poor
countries have adapted in many ways to high infant mortality rates: hospital staff
in rural Indonesia have developed low expectations for the survival rates of at-risk
newborns, and many parents do not name their infants until their first birthday. 它
is not that doctors and parents in Indonesia care less about newborns, 反而
that within the constraints of the local context they perceive few opportunities to
improve infant survival rates. In the FGN case, Gennuso describes the risk of
focusing on the extraction of clean drinking water for the home when their cus-
tomers are more concerned about irrigating their crops.

Sometimes “No Change” Looks Pretty Good

A common misconception in appropriate technology development is that any
intervention will automatically be an improvement on existing conditions. 在
一些案例, products intended to generate better health outcomes in clinical care
require already overworked hospital staff to take on more responsibilities without
any improvement in resources or rewards. 相似地, elaborate water-purification
processes can be a hassle when it’s hot and we’re thirsty—especially if we don’t
have a clear understanding of the connection between dirty water and disease. 在
the case of the FGN pump, compelling technical arguments favored a metal frame
for the device, but conservative user expectations forced FGN to use a wooden
frame.

创新 / winter 2010

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023

Timothy Prestero

The Challenge of Understanding Real Costs: What to do When “Free” Isn’t

There is an important distinction between a product’s “cost to buy” and “cost to
own.” If I won the Formula One Ferrari in the Changi Airport raffle (free car!), I’d
probably have one absolutely wonderful day tearing around the streets of
Singapore—but then, when it came time to pay the thousands of dollars necessary
to refill the gas tank and replace the worn tires, joy would turn to sadness and the
car would go up on blocks. To extend the metaphor, after the raffle committee
delivered the Ferrari to my house, I might not even be able to figure out how to
start the motor, let alone drive it safely. International development efforts are full
of free “Ferrari giveaways.”

Technology donations are rarely accompanied by grants for ongoing product
maintenance and effective user training. 因此, after the ribbon-cutting cere-
金钱, the donor, in Schopenhauer’s words, “hastens to let the curtain fall.”6 All too
经常, organizations misinterpret their delivery of free equipment as the finish line,
rather than the starting point, of an aid program.

Meulaboh, a city of 120,000 in West Aceh, 印度尼西亚, was among the areas
hardest hit by the 2004 Indian Ocean tsunami. By November 2005, the local refer-
ral hospital had received eight high-tech infant incubators, donated by various
international donors. By the time I visited this hospital in late 2008, all eight incu-
bators were broken—primarily for want of simple maintenance and inexpensive
spare parts, and in part as a consequence of Meulaboh’s high humidity, 经常的
power surges, and limited support staff. During the same visit, I found in the hos-
pital maintenance office a brand new, unopened incubator diagnostic and calibra-
tion tool. The staff explained that they had never received training on how to use
the repair device, nor could they read the English-language manual. This situation
is common—according to a study conducted by the Engineering World Health
group at Duke University, up to 98 percent of donated medical equipment in
developing countries is broken within five years.7

Lack of maintenance and training is not the only high cost of “free.” The devel-
opers of Solar Tuki and the FGN pump both found that competition with donat-
ed equipment led to market distortions. “Free” products created unrealistic expec-
tations for product pricing and undermined incentives for change. 在很多情况下,
customers in a region receiving donations have no interest in credit schemes, pre-
ferring instead to wait for the next round of handouts.

DESIGN LESSONS LEARNED

At DtM we have learned that in product design, unlike academia, there is no such
thing as partial credit. To paraphrase Paul Hudnut, social entrepreneurship isn’t
Olympic gymnastics—there are no points for difficulty.8 Regardless of whether the
project we’ve chosen is extremely challenging, like addressing the 1.8 百万
annual preventable newborn deaths from hypothermia, or relatively “easy,” like
maintaining the rural cold chain for tuberculosis drugs in a single country—in the
结尾, the social impact is the only thing that matters. A good designer is an existen-

创新 / winter 2010

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023

Better by Design

Ashoka fellow Souleymane Sarr (L), the author, DtM volunteer Martin Tolliver and
literacy teacher Amadou Traore during early user testing with the Kinkajou Microfilm
Projector. The beta prototype is in pieces on the table.

Photo by Design that Matters.

tialist: either the design is a match for the user need and people use it, or they don’t,
and the design effort has had no positive effect.

The good news is that our experience with the 80/20 rule tells us that even a
relatively small up-front investment in researching user needs and product context
can pay enormous dividends in the development of appropriate product require-
ments and the alignment of stakeholder incentives. Below are some lessons learned
from the three cases and from DtM’s work.

Marry in Haste, Repent at Your Leisure

“Marrying” the wrong set of assumptions early in the product design process can
lead to expensive course-corrections at later stages. Assumptions are risky and
don’t travel well between countries and cultures. New data has a tendency to
demolish old assumptions, but only if we’re paying attention. We have learned to
ignore our fear of asking dumb questions in the pursuit of a solid understanding
of user needs.

在 2004, we were in the prototyping stage of developing the Kinkajou
Microfilm Projector, a portable tool for nighttime adult literacy education in rural
非洲. After we received the client’s approval of the basic design, our engineering

创新 / winter 2010

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023

Timothy Prestero

DtM Fellow Matt Eckelman interviewing NICU nurses at the Kanti Children’s
Hospital, Kathmandu. Having already conducted research at American hospitals, 我们的
goal here was to compare standards and practices in the context of extreme financial
constraints..

Photo by Design that Matters.

team produced a concept for a beautiful, rounded projector housing that used an
aluminum extrusion. At a design review, MIT Professor Woodie Flowers was
unimpressed. “Why don’t you just put the Kinkajou in a lunch box?” he asked.
Imagining life from the perspective of a poorly-equipped teacher in rural Africa,
he assumed that aesthetic appeal was far less important than product cost—and
subsequent user testing proved him right. The final design is very much like a
lunchbox, at a significant cost savings in production.

Alternatives to the Direct Approach

The process of collecting user feedback more resembles psychotherapy than
administering a multiple-choice exam. We have learned that the direct approach,
asking specific questions based on a survey form, is rarely reliable. 作为回应, 我们
have developed techniques to more fully understand the user and the context.

例如, in our observations and interviews, we have learned to “trust but
verify.” We listen carefully to what people say and watch very carefully what they
做. We ask the same question many times, in different ways and in different con-
文本, and with different people at different levels within the organization. We look
for variance, disagreement, inconsistency. 例如, during early user inter-

创新 / winter 2010

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023

Better by Design

views on our infant incubator project, Nepali nurses in a Kathmandu NICU
claimed that they weighed the newborns every 12 到 24 小时, exactly according to
协议. When we asked one nurse to demonstrate the weighing procedure with
a doll, she struggled for a couple minutes with an apparatus that resembled a gro-
cery scale combined with a baby hammock before admitting, “Actually, this scale is
a hassle and we never use it.” Until this conversation, we never considered adding
a scale to the mattress of the incubator.

We have also established some general rules in the process of conducting stake-

holder interviews:
• Ask open-ended questions, nothing that could be answered with a yes or no.
例如, “Tell me what you think about X.” Rather than interrupting to
clarify a question, we will wait and let the interview subject finish. Often peo-
ple who veer off-topic are answering the question they wish we asked—a great
source of unexpected insights.

• People in general are polite, and no one wants to insult us or our product. 我们
try to frame interview questions so that there are no obvious “right” answers.
• Rather than asking directly if a user is having a good experience, we prefer to

take an indirect approach: “Could your friends figure this out?” The individual
user is, 当然, a genius, but most are willing to detail the failings of people
一般来说: “I wouldn’t have any trouble using this, but the other nurses here
would never figure it out.”

• Create opportunities for serendipity. We’ve used questions like, “What’s the

coolest thing in this hospital?” or “What did I forget to ask?” The answers may
provide unexpected insights into local values and constraints.

These are only a few methods for generating insights into the user’s needs and con-
文本. The commercial design firm IDEO has developed a Human-Centered Design
Toolkit, available on-line,which provides a framework for applying design specifi-
cally to the needs of social enterprise.9

The Customer Isn’t Always Right

When asked directly what they want, users will typically suggest incremental
improvements to what they already have: more of this, less of that. This kind of
evolutionary approach to the development of new products and services often rep-
resents a failure of imagination.

We have found that in order to create revolutionary product ideas, we have to
carefully interpret the user feedback and identify the underlying needs. If a user
says they wish a button were red, they could be saying that they wish it was more
visible, or that the current color is somehow offensive or inappropriate. In initial
interviews for our Kinkajou Microfilm Projector, our client focused on the limita-
tions of their program financing. Although fundraising is not a DtM service, 在
observing the nighttime adult literacy classrooms in rural Mali we realized that
there were significant opportunities to improve the effectiveness of the client’s use
of funds. The two most significant obstacles to education in the rural classrooms

创新 / winter 2010

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023

Timothy Prestero

DtM Fellow Tom Weis user-testing our infant incubator prototype with Childrens
Hospital NICU nurses Maria Davila and Lori Boa. Seeing them accidentally launch
a baby doll out of the bassinet and onto the floor gave us the idea to incorporate an
infant seatbelt for hand-transport.

Photo by Design that Matters.

were access to books and lighting. Limiting ourselves to only these two constraints
during brainstorming, we came up with the idea for the portable, battery-powered
microfilm projector, a design that overcame the problems of a shortage of books
and a lack of classroom lighting in a single product.

Avoiding “Mission: Impossible”

The best products are designed to do one or two things very well, rather than many
things poorly. Given that the cost of maximization—finding the optimal compro-
mise between the various requirements—is high in both time and capital, we are
typically forced to employ “satisficing,” an approach that attempts to identify the
best fit in the least time.10

同时, in early user interviews, it is easy to generate a list of impos-
sible requirements: works perfectly, costs nothing, and lasts forever. Attempting to
satisfy every user demand easily leads to “feature creep,” and it becomes exponen-
tially more difficult to optimize the product. This is the real reason for my uncom-
fortable airplane seat: the airline demands a seat that weighs as little as possible and
allows them to pack many passengers into a small space, while the designer aims

创新 / winter 2010

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023

Better by Design

to provide a comfortable chair. All three of the cases here serve as good examples
of first doing one thing well—providing pumps or solar lights or credit to rickshaw
peddlers—before adding complexity.

Failure Is, in Fact, an Option

At DtM, our goal in every product development process is the contrarian mantra:
“Fail as fast as possible.” Product development is an expensive process and fraught
with risk. It is critical to get products in front of users at the earliest possible stages
设计的. As Diego Rodriguez wrote, “Prototype as if you are right. Listen as if you
are wrong.”11 Frequent user course corrections increase the efficiency of the prod-
uct optimization process.

I have already mentioned the Rickshaw Bank’s problem with the falling bicycle
链. The solution to the problem eluded the bank staff and their partners at IIT
Guwahati, until a perceptive rickshaw puller suggested that they replace a section
of the rickshaw chassis with a solid iron bar to keep the frame from bending.
幸运的是, the bank could quickly implement this change on the first 100 rick-
shaws in its pilot program. But imagine their difficulties if they had waited to start
user testing until they had built thousands of rickshaws using the flawed design.

DtM’s strength as a product design company is not that we have amazing ideas
for products. 实际上, most of our ideas are terrible. Our strength comes through
our collaboration with hundreds of volunteers who provide us with lots and lots
of ideas, and the efficient process we have created for vetting our ideas with users
and experts to find and develop the winners.

Doing good is no reason to run a bad business.

We have learned to be wary of any project assumptions that depend on the “spe-
cial case” of social benefit. Product design is an expensive and risky process, 和
should be considered an intervention of last resort. 到底, the most important
question to ask in our due diligence research on new projects is not “Can we do it?”
but “Should we?”

One common example of special-case thinking relates to donor-subsidized
research and development. The social sector lacks a universal performance metric
to serve as the equivalent to profit in the commercial sector. Many social enterprise
programs are small, and it is possible that the most cost-effective solution to a spe-
cific problem involves the use of existing, off-the-shelf technologies, even if they
aren’t perfectly suited to the context. In evaluating a recent program to develop
vaccine coolers for a cold chain application in Cambodia, we realized that we could
buy each of the 40 clinics a thermo-electric cooler from Wal-Mart, along with a
Chinese gas-powered electric generator and a supply of fuel for less than it would
cost us to design a prototype cooler that was perfectly suited to the client’s needs,
语境, and budget.

Another example is the choice of production strategy. As in all design efforts,
the trade-off is typically between doing one thing well and a few things poorly. 在

创新 / winter 2010

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023

Timothy Prestero

也就是说, local production is often assumed to be the best manufacturing strat-
egy for appropriate technologies, in that it creates local jobs—but this approach
also introduces significant constraints on product cost and performance, and ulti-
mately limits the potential social impact. Had local production or user assembly
been required initially, I doubt that cell phones would have ever become a global
技术. The challenge for each product is to identify the appropriate point on
the continuum between the high efficiency of a capital-intensive mass production
process and the ultimate local production methodology of having every user build
for themselves what Gennuso, thinking about pumps, very appropriately describes
as a prototype.

结论

An invention approach to problem-solving is, and will remain, popular among
those developing new technologies for the poor. The challenge is that good design
is neither necessary nor sufficient for a product to succeed. The oft-cited example
is the enormous difference in usability between the Microsoft and Apple
Macintosh operating systems of the 1990s. Apple’s superior user interface design
was not enough to guarantee commercial success in the face of competition from
Microsoft’s marketing savvy and existing user base.

Saying that we don’t need great design insights to launch a successful product
or service is similar to saying that we don’t need great market insights to make
money trading on the stock market. Great success requires luck, and in advocating
for the design methodology, I side with those like the authors of the three cases,
who clearly define luck as “opportunity meets preparation.”

Neil Postman wrote that the early success of “Sesame Street,” which is now a
global phenomenon, was based on children’s familiarity with television commer-
西尔斯, “which they intuitively knew were the most carefully crafted entertainments
on television.”12 The same observation applies to product design. We live in a world
where luxury goods and consumer electronics represent some of the most careful-
ly crafted products available. It is exciting to have the opportunity to address real
human needs as opposed to simply creating consumer desire, and to take the same
skills and insights that make a product like the iPod so irresistible and apply them
to the creation of new and useful tools for the poor.

Acknowledgements

Thanks to Ben Spiess, Elizabeth Johansen, Elizabeth Bruce, and the editors for
shaping up a crummy first draft.

1. Survivor bias: the tendency to draw conclusions from a sample set that includes only successes.
Although the oldest woman in the world reportedly eats sweets and drinks a glass of sherry every
天, it does not follow that those same dietary habits will lead to the same results for everyone.
2. “Among history-making innovations, those based on new knowledge—whether scientific, techni-
卡尔, or social—rank high. . . . Knowledge-based innovations differ from all others in the time they
拿, in their casualty rates, and in their predictability, as well as in the challenges they pose to

创新 / winter 2010

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023

Better by Design

企业家. Like most superstars, they can be temperamental, capricious, and hard to direct.
他们有, 例如, the longest lead time of all innovations.”from Drucker, P。, “The Discipline
of Innovation,” Peter Drucker on the Profession of Management, 波士顿, 嘛: Harvard Business
School Publishing, 2003. 60.

3. Tolstoy, L。, Anna Karenina, Constance Garnett trans., 纽约: P.F. 科利尔 & 儿子, 1917. 1.
4. A popular joke in the early days of the open source movement was, “Why do in ten minutes with
Windows what you can do in four hours with Linux?” Ubuntu Linux’s success outside the devel-
oper community has much to do with their focus on user experience.

5. Also called the Pareto principle, this a rule of thumb that states that, with a given initiative, 这是
possible to achieve 80% of the benefits with the first 20% of the investment, 和, 反过来, 那
realizing the last 20% of the benefits requires the remaining 80% of the effort.

6. Schopenhauer, A。, The World as Will and Idea, R.B. Haldane trans., 波士顿: Ticknor and Co., 1888,

1:413.

7. Malkin, R.A, “Technologies for clinically relevant physiological measurements in developing

countries.” Physiological Measurement 28 (2007): R57-R63.

8. Hudnut, P。, “Radical, Transformative, Unreasonable, Extreme, Leap Frog, Super-Bad Social
Entrepreneur.”
2008.
.

Billion

下一个

六月

blog,

9. .
10. A portmanteau of “satisfy” and “suffice” coined by economist and psychologist Herbert Simon.
See also Bordley R. and Kirkwood C., “Multiattribute Preference Analysis with Multiattribute
Performance Targets.” Operations Research 52 (6), (2004): 823–835.

11. 加西亚, D ., “Prototype as if you are right, listen as if you are wrong.” Metacool blog, 四月 14,
2009. .

12. Postman, N。, Amusing Ourselves to Death, 纽约: 企鹅, 1985. 142.

创新 / winter 2010

从http下载的://direct.mit.edu/itgg/article-pdf/5/1/79/1838402/itgg.2010.5.1.79.pdf by guest on 08 九月 2023 Timothy Prestero image

下载pdf