16 Default Thinking: Why consumer products fail
Short Message Service (SMS), or texting, is a typical killer application. It is not only popular but
profitable, bringing in significant revenue to network operators. There is even a strong after
market selling RingTones, info alerts and crude interactive games. A great technological irony
is that such a successful product is so under appreciated. For all of the frenzied SMS marketing
discussion, the product has hardly changed over the last few years. Given its success, you
would think the industry would put more effort into understanding the value SMS offers to
consumers and then produce new services that extend this value.
Instead, the next services trotted out by the industry were WAP, Instant Messaging, and
SyncML. These services, for all of their hype, are but a pale shadow of their original
expectations. None of these services are building on the success of SMS. It's as if the industry
turned its back on text messaging.
You may think I've forgotten about Multi Media Messaging or MMS. This is intended to be a
substantial upgrade to SMS, allowing the sending of rich text, photos and even sound clips
within the same message. However, MMS under appreciates SMS as well. It does this in two
ways. First, its new application interaction is far more complex; it effectively ignores the
powerful simplicity that made SMS so popular. The second is that MMS doesn’t add any
substantial value for consumers. I will explain both of these points in more detail later in the
However, MMS is just an example of a broader trend. The true culprit here is the approach to
product design that sets up these products and considers them valuable. It seems all too easy to
create products that misinterpret the true needs of the consumer and engender little
enthusiasm. The marketing failures of both WAP and Instant Messaging should be an
enourmous cautionary tale to us all.
I call this shallow approach to product design "Default Thinking". I have seen this throughout
my career in product discussions within many companies. By identifying the components of
Default Thinking and showing how it affects product design, I'm attempting to create a better
vocabulary: a starting point in how to discuss product design issues in more detail. I'll use
MMS as an example, showing how it contains many instances of Default Thinking. This is not
to prove that MMS is doomed but rather to show how this new vocabulary can be used as a
communications tool. The goal is to encourage a more insightful discussion of mobile service
The analysis of MMS will show many weak points of the product but it will also encourage a
broader discussion of mobile services in general. A basic tenet of design is that it is very easy
to criticize and much harder to create. In order to show how this process can be used
Richard Harper, Leysia Palen and Alex Taylor (Eds), The Inside Text: Social, Cultural and Design Perspectives on SMS, 00–00.
©2004 Kluwer Academic Publishers. Printed in the Netherlands.
creatively, I’ll build on this analysis and finish the chapter by brainstorming a series of new
I wouldn't be surprised that many readers may not share my concerns over MMS. My goal isn't
to imply MMS is fatally flawed, only to say that there isn't a very mature way for us to discuss
its pros and cons. Take for example the fairly innocent phrase, "MMS is an extension of SMS
and therefore a natural progression for the industry". This is an exact quote I lifted from a well-
known industry speaker. I would claim there is an entire world of assumptions and potential
misunderstandings in that phrase. How can we pull it apart and discuss it further?
That phrase is infused with 'Default Thinking', which is the underlying way people manage the
complex process of product conceptualization. Product design is a rather soft discipline. There
are no equations for insight or guarantees of success. In order to cope, most people use a
personal set of beliefs to base their work. The problem is that such intuitive rules are rarely
exposed. In most cases, few people really have a strong understanding of how they internalize
the design process. They have a weak collection of axioms of design, broad market visions, or
rules of execution that aren’t clearly articulated. This collection exists in the background, much
like the assumption that gravity exists. These rules are so ingrained in how some people think,
certain ideas or positions are just obvious and therefore not questioned.
There is a popular parable that talks of fish not “seeing” the water. The point is that while fish
live in water their entire life, they never appreciate it surrounds them. It so encompasses their
life that it becomes part of the background. It just fades away and they focus on more obvious
things like finding food or escaping from predators. Default Thinking is our product design
'water'. What makes it so dangerous is it's near invisibility to normal argument. It isn't
composed of strong opinions that can you can discuss easily, but subtle beliefs that are hard to
pin down. In an attempt to make it more visible, I’d like to present four aspects of Default
Thinking and discuss them in more detail: Legacy Vision, Consumer Confusion, Design Syntax
and Design Semantics.
Legacy Vision is the approach we all take when first considering how to use a new technology:
it gives us our initial use. This approach is very understandable: we look backwards at what we
are doing, our current legacy, and apply the new technology in a manner to make this existing
problem better. We see this approach throughout the history of technology. As an example,
when movies were first invented, an expected use was to capture stage plays for later viewing.
The same was true for the telephone, which was envisioned to listen to opera from a distant
city. History has shown repeatedly that most new technologies are initially applied to an
existing, backward looking, legacy task. These first uses aren’t wrong, only naïve, as they apply
the new technology to well known problems. In retrospect, many technologies are disruptive,
changing the status quo and often end up being used in far more profound ways. The initial
uses are quickly forgotten. Movies clearly evolved into something very different from stage
plays and the phone was used for far more than opera. It takes time before the full impact of a
new technology is understood and applied in a more significant way.
Legacy Vision applies to mobile phones today. Once mobile phones had wireless data
capability, the ‘obvious’ thing to do was to look backwards and see the web. "The Web is hot,
phones are hot, therefore web + phone has got to be hotter!" was the manic belief in the late
1990s. This lunging grasp isn’t necessarily bad, only uninspiring. History implies there will
most likely be a much stronger and more meaningful use of data enabled phones that goes
well beyond the ‘web on a phone’ concept.
We need to be on guard against Legacy Vision as it implies a value that may not be there. In
the case of MMS, the new technologies that presented themselves were color screens and
small optical sensors. Digital cameras were a hot growth market so phones + cameras were
perceived to be much hotter. The obvious thing to do was to take this cool new raw
technology and extend SMS to send photos. The proven value of a previous market implies
success in the future concept. This doesn't mean the concept is wrong of course. The point is
that the value of the concept appears to be instantly valid; it just has to be sucessful. While
indeed, there appears to be an intuitive value to 'sending a photo', additional questions such a
"Do people really need this?" and "What are they doing in their lives where this has significant
value?" need to be asked. When Legacy Vision is very strong, these questions are often
considered trivial or worse, hopelessly academic. Just because something is a better message
doesn't mean that it is the right kind of message.
A second problem with Legacy Vision is it's strong negative impact on any explorations of
alternative uses of the same technology. The initial use is often seen as so compelling, it short-
circuits discussion in the longer-term solutions that will ultimately capture the true use of a
disruptive technology. For example, we jumped from the basic technology of color screens
and digital cameras straight into MMS. There are other product concepts that are potentially
more interesting, such as sending a photo during a voice call. Somehow concepts like this just
vanish from the landscape.
The original use of movies to capture stage plays wasn't wrong; it just wasn't ultimately all that
exciting. Something much more interesting happened as the use of the technology matured.
Most likely the same will happen with photos and phones. Something far more interesting will
most likely come along. We should be getting used to this pattern and anticipate it.
Most companies that target the consumer don’t intentionally try to design overly complex
products. The difficulty is in truly understanding the actual consumer. Design problems often
come using a naïve stereotype of the consumer that is more sophisticated than actually exists.
Here is a good example:
This image is shows how easy it is to miss the mark. This is a credit card pay terminal at a gas
station. The clear indication there is a problem is the professionally printed sticker imploring
you to 'PRESS THIS BUTTON’. This problem had to be so bad and widespread the company
had these stickers distributed to all stations and put onto each terminal after the original
terminal was deployed.
The strong implication of the "PRESS THIS BUTTON" sticker is that most people, at least
initially, just want to pay for gas the old-fashioned way: by going inside to the cashier. The first
mistake this design made was not to make the established task obvious to all users. It didn't
understand it had to meet consumers’ initial needs before introducing them to a new concept
such as pay at the pump.
The Implied Consumer, in this case, is someone who wants the use the new pay at the pump
feature. This may seem reasonable but it makes a common design mistake: confusing the need
to create a design for a new feature, with the needs of the target customers. While paying at
the pump is clearly a nice addition, it doesn’t appear to be the first concern of the consumer, at
least, not initially. While the designers of this terminal had no doubt good intensions, they
adopted an Implied Consumer that didn’t match the actual one.
Another aspect of the Implied Consumer is that they share all knowledge that the designers
possess. Implied Consumers know enough to make decisions that are fairly complex. This is
never intentional of course, but it is all too easy to put the consumer in a position to make a
decision that seems ‘obvious’.
This decision problem of the Implied Consumer is related to a graphic design rule introduced
by Edward Tufte, “1 + 1 = 3”. When you draw a single line on a page, you have only that.
However, drawing a second parallel line actually adds two additional items to the overall
graphic: the second line, and the white space in between. The white space acts like a third
graphical element. There is a strong correlation of this rule in product design. Adding a second
button to a product actually adds two additional aspects to the design: the second button and
the requirement to understand the difference between these two buttons. The user must
understand enough to intelligently choose which button to use.
The pay terminal has a severe case of 1+1=3. Its second design mistake is to have a completely
unneeded keypad on the left. There was most likely some future use for this keypad but as it is
completely unused, it just adds visual clutter, giving the consumer ‘another line’ to understand
and process. The third mistake is a series of instructions in the lower left that assumes the
consumer wants to pay at the pump. There is nothing about paying inside. When consumers
are confused, the only thing for them to read is something that doesn’t help.
These last two mistakes together create quite a bit for the consumer to process in order to make
the fairly simple decision to 'just pay as I always have'. In a sense, the consumer is put in the
position that they must 'understand what they can safely ignore'. This is a central problem of
the Implied Consumer. Asking consumers to make decisions unrelated to their primary task
often puts them into a stressful situation: they just stare at the device, not knowing what they
need to do. At this point, the need for the "PRESS THIS BUTTON" sticker is no longer so far
Most companies are sincere in trying to create easy to use products. The Implied Consumer
tends to sneaks up on designers, becoming stronger the longer they work with a problem.
Complex issues become commonplace and even obvious to the designers. It becomes easy to
assume the consumer knows enough to make complex decisions. Any interface, no matter
how simple in appearance will increase its difficulty if it asks novice users to make decisions.
This is most severe when decisions are required that the consumer doesn’t consider relevant to
their main task. Of course, there is a large trade off to be made as decisions allow for more
features to be presented. The point isn't to remove all decisions from a design but rather to
acknowledge that decisions, especially extraneous ones, add complexity. The significance of
this trade-off is rarely appreciated and can have a profound impact on many consumer-
The issues raised by the Implied Consumer go much further than just the initial use of a
consumer product. If this were the only issue, we would be in the age-old argument of "Why
doesn't the user just read the manual?" Decisions have an impact not only on novice users but
also for advanced users as they can be tired or just in a hurry. Interfaces that reduce decisions
can be a win for users of all levels. The point here is not that decisions are inherently bad, but
that unnecessary decisions have a large impact on both novice and advanced users.
This is the danger of using the Implied Consumer when designing a product. The underlying
assumption is that they share the same pc savvy approach the product team had when making
the design. This is why user testing is usually such a revelation to so many product teams. The
inability for a consumer to make the correct decision using a particular design in a particular
place is often stunning. Of course, this does not reflect on the ability of the consumer, but
more in the ability of the product to communicate. The Implied Consumer needs to be well
understood as a stumbling block and then exorcized from our product discussions so we can
properly appreciate the difficulties in using a product in the field.
Design Syntax and Semantics
The term 'User Interface' causes many problems. It has become so broadly used that it is an
imprecise term. A common request of most companies is that their next product has an 'easy to
use User Interface'. The problem is that it’s not clear what this actually means: nice graphics,
no hierarchical menus, simplified task structure, better cognitive model, or fewer features? This
is one of the key problems with Default Thinking as it considers the User Interface a little piece
of code to be tacked onto the end of a product design. The best way out of the problem is to
create a richer set of words to break down the vague phrase User Interface into smaller, more
precise terms. I will suggest a good start is to break up the larger concept into two pieces: the
Syntax and the Semantics of the design. The Syntax is the screen details. Examples of this are
the screen layout, the choice of buttons and the task flow of the design. The Semantics are the
broader motivational issues. Examples would be the context of use, the consumer’s
interpretation the design, and the consumer goals in using the product. Syntax is the HOW,
Semantics is the WHY.
Design Syntax issues were most acutely seen in the Mobile Industry during the early days of
WAP. There where significant problems ranging from screen legibility, scrolling screens,
hidden buttons, complex navigation, and technical reliability. These issues were rarely
discussed by most industry insiders yet seem likely to be a major reason for the lack of
In the case of SMS and MMS, Default Thinking considers them to be nearly the same. The
belief is that they both are just sending a message. This implies their UIs are also quite similar.
The diagrams below show the detailed interaction required to send a message using both
The SMS task flow for the user is fairly trivial. As diagrammed in figure 1, sending a text
involves a fairly simple set of steps: chose to send an SMS, compose the message as a long
string of text, pick a recipient and then it is gone. To be fair, not all handsets have an SMS task
flow this simple. It is, however, used with only small variations by many handset vendors.
MMS, as diagrammed in figure 2, is a bit more complex. First choose to send an MMS over an
SMS, leaving a blank window. To this window add text, which is much like an SMS, but it is
now possible to also add a photo, a sound clip, and a recipient. There is no sequence as you
can add any of these in any order. However, without a recipient, the SEND option is not
available to you. Note that both task flows are taken from the same handset company, one
from a phone that is SMS only, and the other that has SMS and MMS as options.
Three observations can be made at this point. The first is that the Design Syntax of SMS is fairly
linear: there is a sequentical path that must be followed. It is hard for the consumer to make a
mistake in that there are very few wrong turns to be made. The only real freedom they have is
in the composition of the text of the message itself.
In contrast with MMS, there are many choices to be made. The consumer is forced to choose
what type of message, what should be added in what order, where to place the photo, and
where to place the sound clip. These decisions assume that the user knows the answers to
these questions. The consumer must understand the fairly sophisticated document structure
required of MMS before they can navigate this decision hierarchy.
The second observation is this creates a more mature perspective when discussing why people
have trouble initially learning to use SMS. Breaking the SMS process down into this sequence
of steps shows the actual "User Interface" can be grouped into two sections: a fairly simple set
of sequential tasks and a fairly difficult process of composing text on a 12 button keypad. A
common argument used in favor of MMS is that "People had difficulty getting used to SMS at
first, there will be a similar period of adjustment for MMS". This statement must now to called
into question as the most difficult part of SMS is a skill that is completely transferable to MMS,
namely, the input of text. This implies that if people are indeed having trouble with MMS
today, it is not for the same reason that people had with SMS. You can't just equate the
learning curve of both products.
The final observation is that the concept of a message is completely different for MMS and
SMS. SMS is an empty vessel and all you can do is add text. MMS has a more robust model,
which is very much like a word processing document on a desktop computer. An MMS
message is like an empty document that can contain text, photos, sounds, and addresses. All of
these can be added in any order. If the message doesn't have an addressee, the SEND option
isn’t available and the message can’t be sent. In the SMS model, after composing the message,
the consumer is prompted for an addressee. The data richness of MMS creates the possibility
for the consumer to make mistakes that were impossible with SMS. For example, it is possible
to incorrectly place the photo so it is above the text instead of below. This forces the consumer
to delete content and replace the photo it in the correct position.
Default Thinking implies that these two products are very similar when we can now see that
this ‘simple little layer’, this trivial bit of Design Syntax, informs several deep observations of
these two products. By escaping our Default Thinking it is quite obvious that these products
differ not just in menu labels and icons but in a much more profound way at the data level and
even the the numbers of mistakes possible in the process.
Analyzing the Design Syntax of a product is a communications tool. It is a technique to break
through broad statements and get some detail into the discussion of a product. I find this tool
works at two very different levels. The first is as a reductionist tool, breaking up broad
statements with specific examples that can be called out, compared, and studied. This is what
we’ve just completed in comparing SMS and MMS.
The second level is as a motivational tool when working on a product design. The more you
take apart the Design Syntax of products, the more you appreciate how a complex flow creates
unnecessary problems for the consumers. When working on new products, making the Design
Syntax clean and simple becomes a much stronger, and more motivated goal.
Design Semantics examines the motivation and the values the consumer has in using the
product. This includes what they are trying to accomplish, how they interpret the design, or
why are they using the product at all. Too often the primary focus of a User Interface is the
superficial details of the design such as the icons and menus. Addition insights come from
looking at the consumer’s interpretation of the design and value it may have to them.
At first blush, SMS is perceived as a complex product, primarily as the text input is so difficult.
When first introduced, many people felt that the average consumer would not put in the effort
to become proficient in inputting text using a keypad. However, the value of sending an SMS
was so high that many, many people persevered. The value actually overcame the difficultly of
using early SMS handsets.
Most products rarely start with such a large motivational value upfront. Much more likely, the
exact opposite is true: the value for a particular product concept doesn’t really exist. No matter
how good the design of the product, it will not succeed. The reason for analyzing the Design
Semantics of a problem is much like doing user testing: understand how the consumer will use
that product. If there is no apparent value to a product concept, the user interface is irrelevant.
This concern of a false consumer value isn't a new concept. Most mature consumer industries
such as automobiles, prepared food, even beer, all do significant consumer trials before rolling
out a full product. It isn't an exact science but it attempts to make sure a product will have
value before the big money is spent bringing it to market. These industries appear to
understand the need to validate a consumer’s need, and reaction to a new product concept
before spending considerable time and energy on it. The problem with Default Thinking is that
it creates value too quickly, usually through Legacy Vision. This false value often motivates a
product too quickly. This doesn’t guarantee a product will fail but it does make it significantly
The mobile industry is maturing but still has a ways to go in comparison to other industries.
There aren’t enough studies of users to find out what consumers are doing to understand their
communication needs. This is clearly a very difficult problem to answer well. However it has
to be attempted so at least some discussion can occur, highlighting, if anything, at least the
assumptions that are being made.
Let's consider the value proposition of MMS. The marketing materials you see in most phone
shops make that value appear to be obvious, showing how fun it is to to receive a "Happy
Birthday!" message with a photo and embedded music, or even showing how to send photos
while on vacation. This makes for very enticing advertising copy. By examining the Design
Semantics a little more, I am attempting to discover the underlying value to the consumer. The
best place to start here is to understand the current value of SMS today. Once we have a better
understanding of what is driving SMS, we will be in a much stronger position to evaluate the
potential value of MMS.
Summary of Existing Work
There has been a considerable amount of work on the values that SMS currently provides, as
well as some of the misconceptions about those values. For example, it is still a common
belief that SMS has succeeded because it is a cheap alternative to voice communications. The
basis of this idea appears to be the accounts provided by teenagers explaining their behavior
by alluding to the minimal cost entailed. Close examination of their actual behavior shows
clearly that cost is a very minimal driver of usage. This is not to say that these same users are
without awareness of costs, but that other values provided by text are greater than and
supersede these. For example, Kasesniemi (2003) shows convincingly that the ways in which
SMS gets undertaken results in an increase of costs when compared to the costs that result from
voice only communication: though each individual text costs considerably less than an
individual voice call, Kasesniemi shows that texts are rarely if ever solitary, and are more often
made up of a series. It is the series of texts that creates cost; and these are often very
substantial. In addition, she and many others have shown that these costs are in practice
viewed as insignificant to the value that SMS provides.
The most obvious value of SMS is it’s low level of intrusion: a recipient is not required to
interrupt their activity to deal with the message nor is there any rule of conduct demanding
specific speed of reply (Kasesniem, 2003; but see also chapter 3 in this book looking at the
moral obligation to reply at some time). SMS also allows both parties to modulate their turn –
taking to suit their circumstances (chapters 5 and 6). A related value has to do with the time
shifting nature of SMS, whereby it affords a mix of real time and asynchronous messaging
(Ling, et al, chapter 4; for the benefits that this provides for minority groups see Bakken,
SMS also affords a degree of privacy that voice (and even MMS) does not, through allowing the
user to create and read messages without the activity being excessively conspicuous to those
around. One consequence of this is the still largely undocumented practice of using texts to
subvert particular contexts such as meetings and TV watching. A more thoroughly researched
area is on the use of text to develop relationships that would otherwise be subject to social
control, the most clear example of this being Philippino women using text to have emotional
and even sexual relationships with men (see chapter 10). Part of these usage patterns have to
do with the ability of texts to have gift like properties, binding both sender and receiver into
systems of obligation and exchange (chapter 14). Text also has some particular benefits when it
comes to the articulation of meaning, with many researchers noting that with text individuals
find that they can express themselves more clearly and with greater forethought than with real
time talk. This results in the gifting just mentioned and also other shifts in social practices.
Riviere & Licoppe for example (chapter 5) note that SMS is used by couples in France to
avoid the violence that sometimes attends difficult emotional conversations.
When SMS first arrived, most people felt it was a trivial product. However, it met a deep need
to communicate simply, less intensely, and in a time shifted manner, all of which enabled
people to communicate in ways they wouldn’t have normally done before.
In light of these studies, MMS’s value proposition doesn’t appear to be as strong. These studies
show how SMS is used primarily for social interaction and information coordination. Lots of
quick, small messages are going back and forth. People are sending lots of messages,
exchanging information over time, and setting up rendezvous. In reading these studies, many
ideas come to mind to extend these services, some of which will be discussed below.
However, the current interaction style that has been documented with these studies doesn’t
give the impression that sending photos is the obvious next step.
Of course, these studies do show potential uses for photos. The Filipino women using SMS to
augment the dating ritual has could clearly gain value from being able to send a photo.
However just because something can have initial value doesn’t mean that it has longer term
potential, i.e. it will continue to have value after initial use.
In addition, the gift exchange studies may indeed show a potential value. It is possible to
create quite a complex MMS, one that includes not only a picture but sound and text as well.
This has value as a gift. There clearly could be a small study in the gift giving groups to see
how they would respond to photos as gifts to understand this further. The real issue is not that
this could be a gift but more likely, how often it would be used. If used rarely, it has little long