IT
in the 21st Century
In the 21st Century, successful IT organizations will move from reacting
to change to predicting its impacts. The ability to reliably implement
changes to our IT systems is no longer sufficient. We need the ability
to make accurate predictions about the impact of change. We need accurate
predictions about the impact that changes in IT will have on a business
and its competitive position in the market. We need accurate predictions
about the impact that changes in the business and in the markets in which
it competes will have on underlying IT infrastructures. We need these
capabilities to more effectively expend our efforts and resources in support
of the business' competitive position.
The
ability to make accurate predictions depends on the availability of accurate
data about the underlying components. The technologies and tools now exist
to deliver the individual pieces of data needed to produce these predictions.
This was not the case in the past. Recent advances in software test automation
tools, in particular, have delivered revolutionary capabilities. Driven
by Y2K and the Internet, advances in these tools now give us the ability
to build and maintain an accurate blueprint of how our systems work, how
they work together, and how our customers use them. What's missing is
an operating model that takes full advantage of the capabilities these
tools offer. We need a predictable process that produces reliable, repeatable
results. Once that process is put in place and the data begins to flow,
the ability to use that data to produce accurate predictions about the
impact of change will naturally follow.
The
intent of this article, the first of three in a series, is to introduce
the objective, Enterprise Impact Simulation (EIS), to identify the technology
breakthrough that makes it possible, and to the identify the demand drivers.
The second and third articles will explore the primary drivers for EIS
in more depth and go on to develop the model that will deliver it.
Henry
Ford was the first to implement the operating model that revolutionized
the automobile industry. The technologies that enabled his model had recently
come into existence. Ford's contribution was the development and implementation
of an operating model that took full advantage these technologies. The
IT industry is on the verge of a similar revolution. The required technologies
are here. The question that remains is this. Who will be first?
Enterprise
Impact Simulation: What is it?
Enterprise Impact Simulation (EIS) is a "dashboard" that delivers immediate,
accurate predictions about the effects that proposed changes to the business'
IT infrastructure (i.e., business applications, utilities, system software
and hardware) will have on the business, its customers and its trading
partners; and vice-versa.
In
a nutshell, EIS will give you the information you need to drive your business
at Internet speeds. IT executives have been forced to drive blind for
too long. In slower times, that was tolerable. Not pleasant, but tolerable.
Not any longer. Not on Internet time. EIS will give you facts about your
systems and your organizations. Facts about how those systems work, how
they work together, and how the systems are used. Facts about what it
will cost and how long it will take to change them.
EIS
will give you facts about your organization's behavior. Are you following
best practices or cutting corners that will lead to problems? EIS will
deliver facts that give you visibility into the organizations around you;
both the internal business units you serve and the external businesses
you need to collaborate with. How are they using the systems? Are usage
patterns shifting? EIS will give you the visibility you need to avoid
obstacles and to take the lead over your competitors.
For
years the IT executive's job has been to give the business units the tools
they need to run the business. Today, more and more, the IT organization
is the business. Enterprise Impact Simulation will give IT executives
the capabilities they need to accurately predict the impacts of change.
How
is it possible?
The ability to make accurate predictions depends on the availability of
accurate data about the underlying components. The technologies and tools
to deliver the individual pieces of data needed to produce these predictions
have only recently come to exist. Recent advances in software test automation
tools, like those from market leader Mercury Interactive,
now give you the ability to build and maintain an accurate blueprint of
your IT systems. Once you have that blueprint, the step to simulation
and prediction is a very short one.
For
years you have asked for facts and predictions. Who uses this data? What
will it cost to modify the system? How long will it take? Who will be
impacted? Simple questions. In return, you received complicated answers.
Instead of facts you get feelings. Answers typically include phrases that
hedge the facts; phrases like "we think it will cost about" or "it will
probably take around" The reason for this hedging is simple. Nobody really
knows.
Nobody
in your IT department really knows how the systems work. Sure, you have
employees who know part of the answer. But nobody knows the whole answer.
That's because the whole answer is too big, and everybody keeps their
part of the answer in their heads. In cases where you really need accurate
answers to these questions, it invariably takes an inordinate amount of
time. First, there are a series of meetings with all the gurus in attendance.
They put their heads together, debate the facts, break to check their
facts, have more meetings, until finally they arrive at a conclusion that,
more often than not, is still delivered with caveats and hedging. In the
end, the only real difference is that you're faced with a larger group
of people who all agree that they really don't know for sure.
The
latest generation of software test automation tools gives you the ability
to change all that. These tools can comprehensively capture the behavior
and usage of the systems with 100% accuracy. Tools like Mercury Interactive's
WinRunner product give you the ability to define and document a
system's behavior at the User Interface level. Tools like their TestDirector
product give you the ability to manage that definition. Their LoadRunner
tool allows you to define and verify the system's performance at various
usage levels. Mercury's new Topaz product allows you to monitor
the system to detect changes in usage patterns.
Tools
from vendors such as McCabe allow you to trace the flow of data
through a system, and from system to system. Their code coverage tools
allow you to identify the exact lines of source code executed as a result
of a specific transaction on a system. Other tools can monitor who's logging
on to a system, how often they use the system, and what they do while
they're logged on.
Individually,
these tools produce accurate data about a particular aspect of a system's
behavior. Collected and correlated, this data can be used to deliver answers
to important questions; instantly and accurately. By combining the use
of WinRunner and a code coverage tool, for example, you can answer the
question: "How many lines of code will be impacted if we change this field?"
Using data flow analysis tools in conjunction with WinRunner allows you
to answer the question: "If we change this field, what other systems will
be impacted?" Combining data about the identity and activity of the system's
users with an on-line org chart allows you to answer the question: "What
organizations will be impacted if we make a change to this system?"
Imagine
being able to get answers like these without having to go to the programming
staff. Imagine how much more productive the programming staff could be
if they had this information at hand and didn't have to go research the
answers by digging in to the system's source code. Imagine how many mistakes
and how much rework effort would be avoided by simply having these facts
readily available.
Once
you have the data you need, the step to simulation and accurate prediction
is a short one. Simulation tools exist today and are readily available.
All that remains is to implement a new operating model in your IT organizations
that makes the development and maintenance of that data a priority; an
integral component of your definition of a successful project.
Do
we need it?
The demand for change is being driven by past and present failures, and
by the promise of the future. Cost and schedule over-runs and the inability
to see them coming in time to prevent them, delivery of systems or enhancements
with less than required functionality, and unintended impacts to organizations
that weren't notified of changes to a system they use are typical examples
of past failures. Loss of important knowledge as employees leave for higher
paying positions due to the shortage of IT labor is one of today's biggest
challenges. Enterprise Impact Simulation addresses the root causes of
these problems and can make them a thing of the past.
C-commerce,
collaborative commerce, takes e-commerce to the next level. The promise
of c-commerce is attractive and compelling. The challenges it brings include
increasingly interconnected systems and jointly dependent projects that
today's typical IT organization are, based on past performance, ill equipped
to handle successfully. Enterprise Impact Simulation can equip you to
adequately meet these challenges.
Enterprise
Impact Simulation equips you for change. Your need for its capabilities
depends on how well your organizations have handled change in the past
and on how much change they see in our future. Some have more pressing
needs than others. But given the pervasive nature of technology and the
increasingly rapid pace of change, virtually every IT organization needs
to be taking steps today to improve its ability to manage change more
successfully.
Summary
It's not enough, in the 21st Century, to reliably implement changes to
your IT systems. You need the ability to reliably predict the impact of
change. Accurate prediction requires accurate data. Recent advances in
software test automation tools give you, for the first time, the ability
to collect and maintain the data you need. The need is real, driven by
the competitive realities of doing business at Internet speed. The solution
is at hand, waiting only for the implementation of an operating model
that makes the development and maintenance of IT blueprints an unavoidable
part of the process.
This
article, the first of three in a series, introduced the concept of Enterprise
Impact Simulation (EIS), identified the technology breakthrough that makes
it possible, and to identified the demand drivers. The second and third
articles will explore the primary drivers for EIS in more depth and go
on to develop the model that will deliver it.
About
The Author
Bill Walton has 15 years of IT experience including development, QA/Testing,
and management positions with IBM, Compaq Computer, and Sabre, Inc. His
background includes support of Engineering and Manufacturing as well as
back office functions. His breadth of experience gives him unique insight
into the problems facing IT today. His views on the most pressing of these
problems and their solution can be found on his web site at http://www.jstats.com.