Advanced
How To Guides
Systems Integration
Approaches To Integration
Most
real-world scenarios in a typical SME involve multiple
applications that run on distinct physical machines
across the enterprise network, are developed in different
languages and run on different operating systems (Windows,
various favours of Unix including Solaris, HP-UX, AIX,
Linux, as well as legacy operating systems such as VAX
VMS, MV-S, etc.). As such, the typical scenario for
integration involves the flow of data between applications
distributed across a heterogeneous network.
Businesses today are coping with changes in the way
they interact with their customers. Improving customer
knowledge enables the company to maximise the value
of each - companies are investing in integrated customer
relationship management (CRM) solutions for this purpose.
Companies are also experiencing changes in the manufacturing
process, moving from mass manufacturing to stock to
mass customisation and manufacturing to order. This
requires visibility into the entire value chain, integrating
front-end order systems with backend production systems.
Companies are becoming motivated to integrate their
value chains as they aspire to become real-time enterprises.
This involves making the end-to-end process across the
value chain, from requisition to payment, as fast and
efficient as possible. The concept of the "Real-time
Enterprise" is gaining acceptance as companies
seek to accelerate business processes and reduce business
cycle times. And there is plenty of evidence from companies
already moving down this path that there are substantial
rewards for doing so.
Enterprise business integration is a means to an end.
The first key to success is to recognise that in today's
business environment the integration process begins
with understanding the problems the business
is trying to solve. While an agile infrastructure,
enabled by integration, can have substantial impact
on the overall success of the business, technology alone
has no value - two companies in the same business sector
can use the same technology and have very different
results.
Point-to-point hard coding and screen mapping
(sometimes also referred to as "screen scraping")
are relatively inexpensive and easy processes which
enhance a computer system into doing something the designers
hadn't given us the power to do in the first place.
For example, integrating separate Stock and Accounts
systems could be achieved by scheduling a task to run
every night to download a copy of the day's stock transactions,
and generating a journal to reflect the summary values
in the Accounts system. More ambitious processes can
involve automatically copying supplier price list data
from their Web site and using it to update internal
Purchasing system data files. However, such solutions
have definite downsides, not least of all because they
rely on the source data always being in the same format,
or being displayed in the same place on the screen.
They can be difficult or impossible to change, and are
good only in limited scenarios.
Data integration involves simultaneously updating
similar data fields in more than application. For example,
a process could be put in place to automatically update
customer information in an Accounts Payable system when
the Order Processing system data is modified, e.g. putting
the customer on hold for delivery and payment. While
more useful than hard coding and screen mapping, and
less costly than more full-blown systems, it bypasses
the target applications' logic, making it ideal only
when data synchronisation is the primary business driver.
Functional integration brings the logic of the
target system or systems into play. In the Accounts
Payable / Order Processing scenario above, one system
might have a field value "H" indicating that
the customer is on hold, while the other might have
a field value "N" indicating that deliveries
are not allowed. Functional integration involves validating
and, if necessary, transforming data before updating
each system. It is more adaptive than the other forms,
but is harder and more expensive to implement. Still,
when integrating three or more systems, or when the
applications involved may change, functional integration
may be the way to go.
Process
integration is needed to see business processes
end-to-end. It offers better management capabilities,
and can offer business users more of a business view
of the process, rather than just providing an IT management
eye on things. While more complex and expensive, process
integration can also generate large returns on investment
by optimising business processes and reducing business
cycle times. Simply put, process integration is the
big leagues. But it's complex and expensive.
B2B integration enable companies to goes even
further, integrating their systems with those of their
suppliers, partners and customers, and interactively
share data, develop plans and create products online.
It provides a very high ROI by cutting cycle times among
partners and suppliers and optimising the supply and
value chains. It also cuts transaction costs by trimming
transaction cycle times and cutting down on errors.
Integration
is far more than just connecting data and information
sources. Successful integration means that the information
delivered to people must be in a context which is
truly useful. Information is their lifeblood - they
spend increasingly more time hunting for it when it
isn't made available to them in the proper context.
A good analogy is a newspaper - virtually all of the
information in the daily newspaper is available somewhere
on the Internet . But to hunt through the millions
of sites to gather the information to you is impractical.
Instead, we rely on the newspaper editors to cull
through the sea of news and place it in the proper
context for us.
Delivering content in context is the only way to reduce
the real costs of company personnel hunting for what
they need. If SMEs can deliver information in context,
and reduce the time personnel spend searching, they
will have that much more time to do useful work with
relevant information. Companies that are working with
full sets of consistent, timely information can use
Knowledge Management (KM) and Business Intelligence
(BI) processes to make faster, more informed decisions,
and develop more knowledge-based long term strategies.
All signs point to the business requirement to make
information and content available in the proper context,
and to design Web applications and portals that consider
not only how content will be created and published,
but where and when it will be used.
Surveys have shown that a majority of companies spend
between 60 and 80 percent of their integration budget
on consulting services, most of which go towards programming
the infrastructure to connect and transfer data through
the enterprise. The reason for these continued high
costs lies in the fact that in real-world implementation
conditions, current integration solutions suffer from
some critical problems resulting in rigid, complex
implementations that are difficult to maintain and
modify, leading to delayed projects, poor return on
investment, and major cost over-runs.
The most successful implementations are those that
meet the business requirements and contribute to the
overall success of the business. They measure their
success with metrics reflecting Key Performance Indicators
(KPI). These KPIs are business measurements of performance
over a wide range of categories, such as revenue,
service delivery, stock quantities, cost, contract
management, and service quality - not IT metrics.
Meeting business expectations requires correctly defining
the drivers, intent, scope and metrics that measure
success.
|
