Defining Enterprise Information Value
Let's say that an enterprise is a complex collection of semiautonomous units that share
some common goals. An Enterprise
Information System (EIS), then, is a distributed, networked, set of information
processing nodes that services the units of the enterprise, and especially
information transactions across the units.
This concept of EIS is
highly scalable by definition. An
arbitrarily large and growing set of processing nodes might comprise an
Enterprise Information System of Systems.
A processing node might be a microprocessor or a super computer. Subsets of EIS nodes comprise subsystems or
components. The ability to rapidly
compose systems from components, and systems of systems from systems, is a
central value attribute of an EIS. Accordingly, the design of an EIS must be
“open” to the extent that it allows member units of the enterprise to
autonomously expand and federate horizontally.
The Value Assurance
Framework (VAF) aims to improve the predicable level of global success at developing
large EIS. VAF is based on close
observation of many success and failure cases, and includes artifacts that
capture reusable, practices that have proven to be effective for designing,
building, validating, verifying, certifying, maintaining, and operating EIS.
Among
other definitions, Merriam Webster’s first definition for value is “a fair
return or equivalent in goods, services, or money for something exchanged.” In this sense, “value” is different than
“cost.” “Value” is perceived worth,
perhaps in terms of money, but not necessarily.
Possession
of a product or service sooner rather than later is often valued. Similarly, the value of product or service
usually changes over time. It might increase (wine, antiques, tickets as
the event approaches) or, decrease (used cars, electronic equipment,
out-of-style clothing) over time.
Cost is the price
of the utility delivered, perhaps in monetary terms, but not necessarily. Costs
might also be measured in terms of time.
Cost can also be associated with opportunity, i.e., the customer could
have bought something else, and/or the producer could have delivered something
else.
Value depends on utility, cost,
and time.
In
business, a consumer’s objective is to receive maximum utility-per-cost. A provider’s objective is to maximize the
margin between the consumer’s perception of utility, measured in terms of the
price she is willing to pay, and the cost to produce or provide a product or
service. In either case, best practices
evolve around ability to increase the perceived margin between costs (including
production time) and utility delivered.
 |
|
IT is evolving so quickly and unpredictably that
it makes sense to manage risks like the most successful financial portfolio
mangers. VAF maps traditional acquisition artifacts and processes to the
investment portfolio metaphor.
|
 |
|
VAF is informed by a broad baseline of research,
analysis, tools, and literature. VAF captures recurring best practices in
reusable templates. |
Ultimately,
the purpose of an EIS is to deliver value-added information. Information is useful if (and arguably only
if) it leads to better decisions. A
decision might be as mundane as which movie to watch, or as critical as whether
or not to launch a missile at one target or another. Regardless, measures of utility in this regard should address the confidence in, and
quality of, the decision supported.
The
concept of utility is more abstract than the concept of cost or time. That is, it can be difficult to quantify
utility while costs and time are much easier to measure. For example consumers value convenience as a
utility. What is the unit of
convenience? It might not be impossible to quantify convenience per se. On the other hand, time-and-effort-avoided are
essentially equivalent to convenience.
Time saved by making mouse clicks instead of trips to the shopping mall,
or by automatically downloading data instead of entering it manually, is
measurable.
Consumers
value confidence and quality. For
example they want certainty (confidence) that they are choosing good cost
options from among the many available alternatives of varying merit (quality).
Confidence and quality are not always directly quantifiable, but there are
viable proxies. For example, various
consumer guides use number-of-stars as a indicators of confidence and/or
quality. Stars can be keyed to
measurable quality factors such as security, reliability, safety, timeliness,
or performance in some cases. Sometimes,
even though the number of stars is ostensibly objective, that number is keyed
to subjective opinions of (hopefully) trusted evaluators. In both cases, star assignments that are
backed up by trusted logos (certifications) such as Consumer Report,
Underwriters’ Labs, National Institute of Standards (NIST,) or Oprah’s Book
Club, make the star ratings more trustworthy.
VAF
defines utility of delivered information, as its worthiness to support a
critical decision by increasing confidence in achieving desired outcomes.
Critical decisions include especially changes to planned courses of actions, or
allocation of resources to initiate new courses of action. Under this
definition, leading measures of information utility would likely include
combinations of information timeliness, breadth of data surveyed, pedigree of
data surveyed, security, relevance, and reliability. Utility might also include the
concept of surprise, i.e. discovery of important new insight. In this sense, information that provides new
perspective or insight, i.e. a new worldview, is significant. Significant
information provides more utility than information that is merely relevant. Proxy measures for significance
are the same Measures of Effectiveness (MoE) used to quantify targeted
operational outcomes. If delivered information enhances MoE, then the delivered
information must have been significant, i.e. had high utility.
VAF
uses objective measures of desired EIS outcomes, e.g. probability of good or
bad outcomes, proficiency scores, latency, reliability, accuracy, availability,
etc., as proxy measures of EIS utility.
Choosing these metrics carefully, and properly defining their enterprise
interrelationships to manage risks are critical to EIS success.
