Management maneuvers
NetworkWorld Intraview
September
9, 1996 Volume 13, Number 37 Page 42-43
Amdahl Corp., still often associated with big
iron,
has shown considerable nimbleness in moving into the intranet arena.
Amdahl's
intranet involvement dates to 1994, when it began using the term
"intranet"
in a series of customer focus groups. The word was so new and different
that
Amdahl even thought about trademarking it. But alas, it didn't, and now
any
company even remotely related to the networking business has latched
onto
it. But these companies can't boast, as Amdahl can, that they've been
running
internal Web sites since the first browser went into beta. Steve
Telleen,
director of Amdahl's IntraNet Solutions Group in Sunnyvale, Calif.,
gives
his pioneer's perspective on intranet management in an interview with
Executive
Editor Beth Schultz.
NWW -- A lot of people probably still
think
of Amdahl Corp. first and foremost as a mainframe provider. The company
has
clearly been trying to change that image. How will your move into the
intranet
arena help?
STEVE -- Our traditional strengths as a company
have
always been based on our ability to deliver world-class service in
support
of enterprise-wide, business-critical problems. Our products have
always
had to integrate with those of our competitors and others to solve our
customers'
problems. Amdahl also has a long history of helping our customers
bridge
between legacy and emerging technologies, whether it was MVS and Unix,
mainframes and client-server or now tying it all together with an
intranet infrastructure.
It was from this heritage that we approached the
Internet/Web
technology in 1993. And it was this heritage that brought us to
focus
so quickly on both the intra-enterprise uses of the technology and the
issues
of how companies would actually implement and manage this technology as
an
enterprise-wide infrastructure.
Recognizing that our customers viewed our service
as
a major value and differentiating feature, Amdahl executives began to
focus
on the services aspect of our business a few years ago. They set a goal
to
have half our revenues come from services, and this year it looks as
though
they will.
NWW -- Does Amdahl's IntraNet Solutions Group
distinguish
itself among all the Web services providers out there?
STEVE -- The IntraNet Solutions Group began with a
focus
on the organizational and management issues that need to be addressed
when
implementing an Intranet. This focus has led us down a different path
from
other services that start with a focus on the technical issues of
implementing
an Intranet.
The technical issues - the tools and standards -
are
an evolutionary change from what we had before. The effect of the
technology
on organizations is revolutionary. We have focused on how to manage
that
revolutionary transition.
We developed an organizational architecture and a
complimentary
information management architecture that supports a distributed
decision-making,
self-publishing, user-pull model for enterprise information. We then
developed
a methodology for introducing these architectures into an enterprise.
We have packaged the methodology into four
components:
Executive Awareness, Goals Clarification, Implementation Planning, and
Implementation.
The latter two require three distinct focuses: organizational,
technical
and content. The real differentiating feature of our offering is the
organizational focus, the methodology for creating and rolling out the
roles, organizations and skills required to realize the business
potential of this technology.
NWW -- Discuss how Amdahl itself got into Web
technology
and building an intranet.
STEVE -- In April 1993, a few of the technical
experts
in Amdahl's Open Enterprise Systems (OES) organization acquired a copy
of
the Mosaic beta release and began playing with it. They hooked-up with
the
open systems competitive analyst, who had a volume problem making
information
available to our field sales organization. This resulted in a
skunkworks
pilot project focused on a problem inside our firewall.
The IntraNet Solutions Group grew out of the
Strategic
and Market Planning part of OES. In mid-1994, we began looking at how
Amdahl
might integrate the skunkworks pilot into our business. The most
striking
feature of web technology was how easy it was for non-technical people
to
use. And the development trend in that direction was picking up a
velocity
that was hard to miss. From this observation, we concluded that the
technology
was not going to be the difficult challenge; the challenge for both us
and our customers was going to be how to manage the result.
We developed the basic models and management
architecture,
then set out to implement the infrastructure inside Amdahl. We
convinced
the CIO to sponsor the project and used the formation of a
management-based
Web Council as the core of our roll-out. Over the next several months
we
learned a lot about what did and didn't work, and what was critical. We
also
learned something about the stages an intranet goes through during its
development. The results were refined into our IntraNet Methodology.
NWW --You have created a publishing model for
the
intranet. Describe it and explain why you think companies should follow
it.
STEVE -- The publishing model was created as a way
to
use the Intranet to manage the Intranet content. It is a combination of
explicitly
defined functions, roles and concepts and a logical architecture and
set
of processes that take advantage of the strengths of Intranet
technology.
The model allows local organizations to self-publish and manage the
information
for which they are responsible and, at the same time provides an easily
maintained structure that allows all the official information to be
found and browsed, in a logical context. At its core is a set of broker
pages for executives and managers to help them monitor the information
for which they are responsible. If this type of flexibility, local
control and management support is important to the company, then this
model will help.
NWW -- So, you've come up with this idea of a
broker
page. What is this?
It wasn't so much the idea of a broker page as
recognizing
that web pages had different purposes and met different needs. Today I
would
describe these as the need for content and the need for context. As the
volume
of information content grows, users have an increasing need to have it
presented in a relevant context. This is what a broker page does. It
provides context for the user. A broker page should have a definite
audience and purpose in
mind. In a web environment, making context information independent of
content
pages provides the greatest simplicity of management and
flexibility
of use.
Search engines are automated brokers. They are an
important
part of any large Intranet. However, it is important to recognize that
they
are a general broker that provides little or no specific context up
front. Users may need contextual help to formulate the right question,
they may need
support in following a certain sequence of activities, or they may need
help
screening information or identifying the official or sanctioned
information
sources. A search engine is not the appropriate broker for these
situations.
NWW -- What type of information do you suggest
companies
put on their intranet?
STEVE -- An Intranet is an enabling
infrastructure,
and one of the things it enables is the ability of those closest to an
activity
to develop and control the information and activities that best support
their
needs. It also is important to recognize that an intranet is made up of
three
generic types of information: official, departmental/project and
personal.
All three are useful and provide value to the enterprise.
Our methodology is designed to teach organizations
how
to use the infrastructure and to create situations that stimulate
creative
thinking. But each organization in the company should be enabled to
decide
which information and which projects make the most sense for them. The
measure
of a successful Intranet is how well it enables these organizations to
create
their processes and publish their information themselves, without an
MIS
development project or heavy-duty technical person.
The MIS goal should be to add
tools
and functionality that continually expand the capabilities that users
can
do for themselves.
It should not be to identify more projects to do for them. This is the
paradigm
or perspective shift that MIS groups need to make.
NWW -- What do you recommend for companies
starting
an intranet from scratch?
STEVE -- Start with the business goals and drive
the
implementation from the perspective of the management infrastructure.
When
you know what you are trying to accomplish and who is responsible for
publishing,
the technical infrastructure requirements become quite obvious.
Wherever
possible, stick to open standards. If you don't you are asking for
complexity
and costs in the future.
The first management issues are
learning
to clarify intranet business goals and putting the appropriate
measurements
in place.
On security, my suggestion is, don't be too anxious to slap
sophisticated access controls on the Intranet. When the time comes,
follow the enabling principle. Set up a system that allows the
publishers to define who gets access
to specific information themselves, but handle the authentication and
certification
of the user globally in the infrastructure.
Beyond the base technical infrastructure, the
creative
challenge is to continue improving the tools to make the users more
self
supporting.
The other major challenge is providing
intranet-based
workgroup capabilities. This includes E-mail, conferencing, shared
calendaring,
and workflow management. The challenge here is that integrated,
intranet-based
tools are just starting to emerge. The proprietary vendors already are
lined
up with intranet front ends to their proprietary back ends. Whether the
market
will tolerate these lightly masked proprietary solutions or find enough
advantage
in the more open and innovative emerging solutions remains to be seen.
NWW -- Many intranets spring up at a grassroots
level,
then get handed over or are absorbed under the corporate IT department.
Some
companies aren't sure how to handle this transition. Do you have any
suggestions?
STEVE -- When this happens, it is almost certain
that
a management infrastructure does not exist, and this is not an IT
problem;
it is a business management problem. The sooner the CIO can be made
aware
of this, the better for everyone. The technical infrastructure - that
is
the responsibility of IT - is going to look like the
client/server
world IT already manages. In fact, the chances are reasonably good that
it
already is managing much of the intranet technical infrastructure.
Once the management infrastructure has been
off-loaded
to the business managers, the IT department should take on the goal of
enabling
users to do for themselves. This means acquiring the software and
making
it available both for departmental servers and on servers that are
managed
by the IT department. Not surprisingly, when the IT department starts
to
offer servers that are backed up, supported, and that allow the users
to
easily manage their own information themselves, the official
information
finds its own way from the departmental servers to these IT servers.
The
departmental servers continue to support departmental, project, and
personal
information.
NWW -- How do you think the early intranets we
see
today will change?
The biggest change we will see is in what
constitutes
an intranet. Two years ago it was clear cut: an Intranet was a Web
behind
a firewall. In the past two years the definition has changed in two
directions.
Advances in network security have made firewalls increasingly virtual.
That
is, information and access can be protected on physically shared
networks. And there are more inter-enterprise demands for protected
networks - that is, EDI applications are coming to the Intranet.
I now look at intranets as protected webs. I think
this
will continue to evolve to the point that each individual in an
enterprise
may have a slightly different Intranet depending on the projects,
partners
and industry groups with whom they interact.
Copyright 1996 NetworkWorld |