|
[Dallas,
Aug 8th 2000]
The
knives are out for WAP technology. Following the huge hype of the
last 12 months, it's becoming apparent that it just isn't delivering.
Naturally, the naysayers have now become the "told-you-so-ers".
However, the disappointment of WAP technology needs some more careful
analysis. The industry deserves a good wrap on the knuckles. Much
of what has happened was avoidable. Will the lessons be learned?
First
of all it is important not to get hysterical about the supposed
"failure" of WAP. Naturally, the marketing men are rushing
out to paper over the emerging cracks. People are now openly talking
of waiting for the next generation of technology. By this they are
generally referring to GPRS, rather than 3G, of which more later.
It
has been shown repeatedly that introducing new technologies takes
a lot longer than first thought. Local area networking took almost
15 years to really reach the mass market. This 15 year figure has
been true of many new technologies in recent history. Indeed it
might be reasonable to say that the PC, first seen around 1977 (not
the IBM/Intel/MS version) didn't become a mainstream product until
the early 1990s when they became fast enough and cheap enough with
sufficient additional support that a mass market developed. Even
then there were at most 200 million users from a world population
of over 5 billion potential users. Slow introduction is nothing
new. Nothing to panic over.
Badly
managed expectations
First
let's take a look at expectations for WAP. There were two key groups
who had expectations. The buying public who strangely aren't all
that relevant in the current "backlash" and the businessmen
who are! The businessmen and investors looked at the growing internet
e-commerce activity and they looked at the number of mobile phones
in use and the growth of that business and put two and two together
and somehow came out with an answer in the billions. Why? Well they
had research reports which said that there would be 1 billion mobile
phones in use very soon, and other research that said that people
replace their phones, anything from every 9 months (Japan / Hong
Kong / Singapore) to every 2.5 years (elsewhere). So, it was speculated,
if you can have 1 billion phones in three years time and they are
all internet enabled then you can have 1 billion people doing m-Commerce
with their phones. Even without transaction charges, you could probably
charge 20 dollars a month as a subscription. Suddenly 2 and 2 adds
up to 10 billion dollars a month in revenue.
And
what do the genius executives have to do in order to get a slice
of this 120 billion dollar market? Well, gee! All they need to do
is get a 20 something web hack to work 100 hours per week and stick
a few WML pages up on a web server. Right?
Wrong!
Infrastructure
10
years ago it was an achievement to place 250 people on a client/server
system running on a LAN. If it was 15,000 people in a global corporate
running over a WAN then it was an engineering and logistical problem
which carried huge risk and often saw failure. If you had suggested
to those IT Managers that within a decade they would be asked to
build a distributed computing platform capable of delivering online
transaction processing to upwards of 200 million users, they would
have recommended you see a doctor.
However,
that is exactly what WAP Architects the world over are currently
being asked to deliver. A network running on very new Java J2EE
EJB platforms which will scale to at least a couple of hundred million
concurrent users over 5 years.
Is
it any wonder that some carriers have had trouble at modest numbers
around 50,000 concurrent users? Probably not. Somebody needed to
reset the businessmen's expectations with a reality check but they
didn't want to hear that story.
Write
Once, Run Anywhere!
Sun's
Java mantra became the chant of a whole industry obsessed in the
late 1990s. Despite numerous respected people, like Don Norman,
telling the world that "Write Once, Run Anywhere" just
didn't work for user interface, nobody listened. Even as web designers
were learning that they needed to modify or optimize web pages for
different browsers and different screen sizes, the technical boffins
were working on XML and its close cousin and derivative WML. XML
is an immensely useful technology but it hasn't and probably won't
deliver "write once, run anywhere". Once again expectations
weren't managed properly.
The
designers of WAP intended that an XML document type, WML, would
provide the facility to design content for limited screen phones
and allow one set of content to run across any device. Well that
might have been true, if they had persuaded the manufacturers to
produce a standard size screen with a standard deck size. However,
such standardization reduces the scope for "innovation"
and the manufacturers wanted to compete against each other whilst
offering minimal cooperation on an adequately loose standard. Varying
screen sizes are a fact of life. That's innovation!
It
was never going to be true that WML or even XML with XSLT would
offer a "write once, run anywhere" solution for web developers.
The reality is that producing content for WAP phones is expensive
and time consuming. For the content producers, the rewards from
the early market are minimal. Staff could be better employed working
on the wired web site. Without content it is difficult for carriers
and manufacturers to offer compelling products.
It
really was "the Content" Stupid!
The
customer's expectations were set by the marketing men keen to shift
more phones. Their goal was to grow the market. They can do that
by selling more phones to mainly mainstream majority market customers.
Even the laggards are now in the market for a phone.
However,
to sell internet ready phones with service packages, you must target
early adopters - people who have a phone already. So you have to
persuade them to trade up to a new one. Taking the UK as an example,
the carriers went straight in and said, "Surf the internet
on your phone". Somehow or another the postage stamp sized
screen wasn't going to allow that. The
customer has been disappointed. You simply cannot surf the internet
with such a device. Not in the sense they have become accustomed
to. Now these same carriers are having to recant.
Meanwhile,
the carriers were offering limited services through a menu of hyperlinks
on their server. These "walled gardens" as they are known
have proved unpopular. Why? Well it stops you from surfing which
is what the customer was promised.
The
critics have jumped on this and pointed to the need for the normal
wired internet experience. Naturally, this would work. In fact I
can do it already. I just carry my laptop, hook it up to my phone
and surf on. Doh! It took a genius to come up with that idea.
The
notion goes that making the screen bigger would allow developers
to use existing development tools and reduce the overhead for content
production. This is certainly true. There might be a reduction of
time and conventional HTML authoring tools could be used. However,
we have to take a serious look at this claim. Already NTT DoCoMo's
iMode service uses an HTML derivative, however, with screen sizes
smaller than a typical Palm Pilot, I doubt that it is possible to
simply deliver a standard web site to the smaller screen. The truth
is that "write once, run anywhere" just doesn't work for
human computer interaction across different screen sizes. "Write
once, run on a single form factor device" is the rule. Break
this rule and the user experience drops off quickly to a point where
you cannot offer the product to the mass market.
However,
it might have been possible to design some truly useful internet
enabled content for WAP phones, if the industry had taken a different
approach. Here's how...
WAP
Phone as an Information Appliance
The
biggest single mistake was to take the view that an internet enabled
phone is a general purpose device. It was an easy mistake to make.
A computer with web browser was a general purpose device. You can
use it to surf any web site. A phone is also a general purpose device.
You can use it to call any number. However, when you put the two
together, the combination is limited. To be completely general purpose
it would need to have a keyboard, a full size screen and a phone
transceiver built in together. It would need to be both a computer
and a phone, i.e. a laptop with a phone built in. Current WAP Phones
are still phones, but they are NOT computers in the sense of a PC.
This was the first mistake - marketing the device as if it was a
computer.
A
WAP Phone is an "invisible computer" or it ought to be.
A WAP Phone thought of as an invisible computer, becomes an information
appliance. Furthermore, each different form factor of WAP Phone
is a different information appliance. The whole industry failed
to realize this. Information appliances should be designed for a
specific purpose or a limited set of purposes. In other words, with
current technology, the "walled garden" approach was correct
providing what was inside the garden made sense for the specific
information appliance, as a single product. The marketing ought
to have reflected this approach.
The
purposes to which a WAP phone can be put is somewhat controlled
by the modality of the device. For example, a 4 line display is
very different from an 8 or 12 line display. A 2K deck size is very
different from an 8K deck size. As the devices change and grow more
powerful the range of tasks to which they are suitable grows larger.
Identifying and developing such compelling uses for the small screens
was always going to be difficult.
Screen
Size is not the only issue
In
the early days of WAP, the limited screen size was quoted as the
most likely reason for future failure. True enough, small screens
limit the usefulness and make designing compelling solutions more
difficult. Limited solutions means a limited market. As screen sizes
get bigger more options become possible.
However,
with a bigger screen, many of the current problems don't go away.
In fact, they get worse. We will still have the problem of many
varied form factors and consequently many different renderings of
the WML output. The content in that output will get more and more
elaborate as the temptation to place more and more on the screen
becomes overwhelming for the marketing people. So when 3G devices
arrive, the design problems will get more difficult, not less. Delivering
compelling and useful solutions will still be hard but more of them
will be possible.
Specialization
is the key
Forget
the notion that all the problems go away with a bigger screen and
an "always on" technology such as iMode cures the other
problems. If the market continues to tout the notion that they will
sell general purpose devices which can be customized then we will
see failure compound failure. Already the marketing guys are winding
up to proffer GPRS as a "replacement" for WAP. GPRS is
little more than WAP which is on all the time - an iMode for the
rest of us, if you like. For sure, removing the dial up latency
will improve the user experience but it won't solve the real problems.
Many
people will believe that the easy solution is to create a general
purpose wireless internet device with a modest but adequate screen
which can then be customized by the user. To those people I say,
"How many people do you know with a Palm Pilot? How many have
loaded anything other than the basic out-of-the-box applications?
Of those who did need something else, how many opted for the Handspring
Visor with its plug-in cartridges?"
Limited
modality devices and general purpose applications just don't go
together. If you want to be compelling and get the audience to use
the devices then you have to think about specific, narrow applications,
then go and get the content to make it compelling. Make that content
work for the specific device, optimally. Forget about customization
and pick'n'choose type configurations. Forget about delivery across
multiple form factors. Try to think like an appliance manufacturer
and less like a software developer. Identify an audience, design
a device for them and go and target them specifically. Concentrate
on delivering real value in small doses. Can the industry learn
to focus like this and avoid the mistakes? Naturally this will mean
that you can't capture that 200 million plus user base with just
a single offering and just "a few pages of WML". Then
again, who were those business types kidding? Only themselves!
|
