|
|
| For more than two years, IBM has
been encouraging its customers and the world at large to adopt "e-business."
In short, e-business is about connecting valuable information to the people
who need it. The best way to do this, we think, is to combine the reach
of the Internet with the vast resources of information on traditional systems.
e-business also includes connecting people using Intranets and finding
new ways to streamline key business processes.
While we've been talking to the world about e-business, we've been in
the process of becoming an e-business ourselves. e-business offers a new
set of opportunities, needs, rules, and challenges. We are forging new
roads on our journey. There were no maps to point out the potholes and
no examples of successes or mistakes from which to learn. IBM became the
pioneer of the strategy and technology so that others could use our experience,
mistakes, and successes as a guide.
We approached our task in three stages: unifying the computer systems,
transforming key business processes, and Web-enabling core business processes.
Let me share with you the perspective we have on the continual journey to becoming an e-business leader. |
||
The Beginning
IBM's journey toward becoming an e-business began about 1993. That's
the year, you may recall, that our profitability took a dip which caused
a major reaction on Wall Street. That drop in market share and stock price
caused IBM to reevaluate every facet of the organization.
Then, IBM was a maze of complexity. We had 400,000 employees with offices
in more than 160 countries. We had more than 25,000 combined hardware and
software products. We went to market as twenty different businesses with
completely different fulfillment, accounting, human resources (HR), and
payroll systems. Redundancy was everywhere.
The IBM mission was to fix this and return the company to profitability.
The challenge, however, was to build a culture that would allow the IBM
organization to continually transform itself. By becoming a smarter, faster
organization and learning from our mistakes, we could bring the company
back in the black and in a position to prevent the slide from happening
again. Graphically, our vision is to set four waves into motion: |
 |
| The strategy was to transform the
core business processes, leverage knowledge and information, build new
applications, and run a scalable, available, safe infrastructure. To start,
we had to focus on one goal. For us, that was very clear. IBM had to go
to market as one integrated, global organization rather than 20 different
companies.
We very rapidly streamlined accounting by creating common general ledger
and payroll systems. Culturally, we began basing compensation on total
IBM performance instead of business unit performance. Other cultural changes
were made in both accounting and HR to ensure all of IBM was counting and
talking in the same way. The resulting system cut financial expenses-to-revenue
from 2.3% in 1993 to 1.2% today
The second wave began in 1995 with
"value-added processes." These were the key processes across the organization
that needed to be managed globally. |
||
|
||
| Once again, we started by eliminating redundant systems, such as the
25 fulfillment systems and the different infrastructures in every production
plant. For example, by consolidating the fulfillment and procurement systems
to an SAP R3 application, IBM was able to speed up the order-entry system. Information
that customers used to wait hours or days for, such as a shipping date,
is now provided instantly. Invoice accuracy is 98%.
In the procurement system, we can process purchase orders in 24 hours
instead of 30 days. Internal satisfaction has risen to 85% and, in 1999,
we expect to procure $12 billion in goods and services over the Web, saving
$240 million through the implementation of 3-procurement applications.
To build an integrated supply chain, we needed to substitute information
for inventory throughout all the production plants. In 1995, everything
IBM purchased was based on a forecast. Because the supplier couldn't see
what parts were being used, there was no way to replenish them until the
next forecast arrived. As a result, IBM had long order-to-delivery cycle
times, very high inventory, and substantial write-offs.
Now, we've got more than 80% of the inventory moving through production
on a regular basis. By allowing our partners and suppliers to see component
usage they can replenish based actual usage. Cycle time is down and we
saved about $800 million (US) just in inventory write-offs.
We also needed to create a common and accessible way to understand our
customers' needs and to develop products. In this second phase of transformation,
we consolidated terminology, methodologies, and systems across the organization.
In a nutshell, we broke down the barriers to communication and made sure
people could talk to each other. |
||
Moving to the Web |
||
| In 1997, we began the third wave of transformation by Web-enabling
the core processes under the headings of e-procurement, e-care,
and e-commerce. IBM finished 1998 with $3.3 billion (US) of goods and services sold though our e-commerce systems, primarily business-to-business, through OEM partners, Business
Partners, and ShopIBM. We expect that will increase by a factor of four or five in 1999.
Using e-procurement, we'll purchase roughly $12 billion (US) in goods and
services in 1999.
However, these systems are more than just a Web front end. Take the
e-procurement system for example. When we procure with a supplier over
the Web, we ask that suppler to upload their component prices -- how much
the vendor pays the supplier for every component that goes into the finished
part. The e-procurement system then automatically compares the component
prices to those purchased by a production plant elsewhere in IBM. If it
finds that the vendor can get a better price on a component, it notifies
the vendor. The vendor can then obtain the better price from its supplier
and pass the savings on to IBM. Similarly, we've found instances where
a vendor is getting a better price than IBM and that has allowed us to
go back and renegotiate with the supplier.
The e-care portion of the solution performs a similar service
for our customers. From one place, www.ibm.com,
you can have instant access to all the updates, tips, and product knowledge
in the organization. Customers have an instant way of communicating their
product questions and ideas back to IBM, too.
The key is not to Web-enable everything; it's ensuring that everything
you Web-enable ties back to the core business processes. That's where the
real power is.
|
 |
|
|
We created a framework for the systems we would build and deploy. At
the very base of the framework is our infrastructure. We're converging
our IT systems with the Web, including linking the data in our back-end
systems to Web-enabled front-end applications. This allows us to do all
sorts of transactions and information sharing over the Internet.
The part of this that relates specifically to our Internet activity
we call our Global Web Architecture (GWA). This has involved a worldwide
consolidation of servers and Web hosting, resulting in seven major server
farms around the world. GWA is AIX based
on SP2s and RS/6000s. The IT and Web infrastructure has been outsourced
to IBM Global Services, allowing us to deploy consistent architecture,
standards, and systems on a worldwide scale.
Web applications were needed to deliver capabilities such as commerce,
registration and profiling, and content management. Before we jumped in and started building the applications, we looked
at the future. It's nearly impossible to forecast future demand for an
application or a server. If the application is really valuable, its usage
explodes. We decided to make sure that any application we put on a server
could be quickly moved to another server as usage demanded.
We started focusing on open applications frameworks and industry standards
for what we would build. To minimize our cost and risk, we also looked
for applications that we could buy. For everything to work together, the
software vendors with whom we worked had to subscribe to similar standards
As we began making changes and deploying applications, we discovered
that the end users were constantly having to learn new navigational metaphors;
how to get around the application and get to the data. To simplify this
training, we decided to have only two navigation systems to interface with
the core applications: Lotus Notes or a Web browser.
New applications are developed in Java using
IBM VisualAge for Java. Applications both bought and built follow the
CORBA
and Internet standards.
Lotus Notes/Domino serves as the
server and interface to many of our core applications. The IBM PROFS system
is now Notes Mail. Many of the applications, like the SAP procurement system,
are accessed through Notes.
Almost all the transaction systems across IBM run on DB2
with Transaction Series (MQSeries
and CICS) as a hub to connect many of our new and legacy transaction-based
systems; e-commerce, accounting, procurement, etc.
We've recently started using the somewhat-new WebSphere
Application Server for both internal and external Web applications.
To reduce complexity and day-to-day management costs, we brought in
Netview and Tivoli
Systems, Inc.'s software products to provide end-to-end application,
system, and network management. |
||
The Data CenterWith the framework and the tools in place, it was time to look at the underlying structure.The transformation project began in 1993 with 155 data centers, 31 separate
networks, and hundreds of client configurations. There were more than 100
"local CIOs" who had each created standards and architectures for their
own area. Today, we have 25 data centers and we're consolidating those
to just six. We have one integrated worldwide network and we'll
soon have only four standard client configurations. There is now just one
person with the title of CIO; one person to create standards and architecture
for all of IBM.
By reducing data center expenses, we've reduced the IT budget by 26%.
Adding Tivoli software to manage and automate routine tasks has helped
free IT to concentrate on more important issues. The results have been
outstanding. Network availability is 99.5% and response time is consistently
under two seconds. Internet response time averages under six seconds and dial-up
availability for our mobile work force is 99%.
Overall, measuring performance, testing, and deployment has improved
reliability of the entire IT system by a factor of 40 and improved performance
by a factor of five. In an e-business, making sure the systems are running,
accurate, and fast is half of the battle. |
Lessons Learned
Always look to the future. While you can't accurately estimate, you can always anticipate. Be ready to move quickly when demand for the application explodes. |
Leveraging the InformationThe other half of the battle is leveraging the information and knowledge in the organization. The ideal situation is to get the needed information into people's hands all the time; to allow them to work on the programs and projects to help continually transform these core processes. We also want these core processes to be at the desktop in real time.To make this happen, we need a collaboration and messaging environment
that enables individual and organizational learning, and accessing existing
and creating new knowledge. We identified seven core business processes
that offer the greatest business value—and the best ROI—today. These
became our "e-business initiatives." The strategy focused on individual
efficiency and competency, and organizational responsiveness and innovation.
As you can see, these effect every business process in our company and
the Web puts new demands on them.
We began plotting each program on a grid that helped us measure the
strategic value of the program instead of the purpose of the investment.
We ask questions about each, for example "is this to improve infrastructure
costs or to improve a market-facing set of systems?"
With the grid and the questions, we can see where the budget money is
going—and rethink how the money is going to be spent.
The projects with the greatest strategic impact are those on which we concentrate.
Since we started this project in 1993, we've made substantial improvements
in terms of ROI, in terms of core processes, and in terms of budget. We're
now starting to focus on new channels of distribution to serve our customer
in an even more consistent manner, to meet customer needs better, and to be
more responsive.
When it all comes together, it's about how we can do things smarter,
learn from our mistakes the first time, and continue to become faster and
more responsive in this very complex and competitive world. We've got the
battle scars from doing this ourselves. It puts IBM in the unique position
of being able to lead our customers around the potholes we found on the
journey to becoming an e-business. |