Information Technology and the Firm 543
(some companies have thousands of PC clients), maintaining the software for all
of those clients could be a nightmare. Even with specialized tools developed for
that purpose, it never quite worked perfectly.
As companies recognized the opportunity to send data over the Internet,
whether for their customers or their employees, they started to migrate all of
their applications to a browser interface. This change has required companies
to rethink where the locus of their processing will occur. Prior to the 1990s,
companies’ networks were host-centric, where all of their processing was con-
ducted using one large mainframe. In the early 1990s, companies began using
client-server architecture. Today, with the current browser technology and the
Internet, the locus has shifted back to a host-centric environment. The differ-
ence, though, is that the browser on the users’ computers is used to display and
capture data, and the data processing actually occurs back at the central host
on a series of specialized servers, not on one large mainframe computer. The
only program users need is a standard browser, which solves the incompatibil-
ity problem presented by distributed data processing. No specialized software
is stored on the users’ computers.
Internet technology was largely responsible for many of the productivity
enhancements of the 1990s. Intel’s microprocessors, Sun and Hewlett Packard’s
servers, CISCO’s communications hardware, and Microsoft’s Windows operat-
ing systems have all facilitated this evolution. While Windows is the predomi-
nant client operating system, most servers operate on Windows NT or 2000,
UNIX or LINUX operating systems.
TODAY’S APPLICATION SYSTEMS
In the 1970s and 1980s, application software systems were stand-alone. There
was little sharing of data, leading to the frequent redundancy of information.
For example, in older systems, there might have been vendor data files for both
inventory and accounts payable, resulting in the possibility of multiple versions
of the truth. Each of the files may have contained address information, yet
each of the addresses may have been different for the same vendor. Today,
however, software applications are integrated across functional applications
(accounts payable, accounts receivable, marketing, sales, manufacturing, etc.).
Database systems contain only one vendor data location, which all systems uti-
lize. These changes in software architecture better ref lect the integration of
functions that has occurred within most companies.
Accounting systems, while used primarily for accounting data, also pro-
vide a source of data for sales and marketing. While retail stores’ point of sale
cash registers are used as a repository for cash and to account for it, they are
also the source of data for inventory, sales, and customer marketing. For exam-
ple, some major retailers ask their customers for their zip codes when point of
sale transactions are entered, and that data is shared by all of the companies’
major applications.