Innovating Patterns of Interactions- Valutrics

The world’s network of interactions has undergone several profound changes. First,
the total volume of interactions has grown. The markets of “interaction enablers” (e.g.
telecommunication services, Internet, and media) as well as “enablers of enablers”
(e.g. telecommunication hardware and software, CRM, EDI, and e-procurement
systems) have largely been outpacing GDP growth over the past few decades. For
example, the share of IT spending in US GDP went from 2.5 percent in 1990 to
nearly 6 percent in 2000s.  Thus, interactive time is substituting non-interactive time.
Secondly, interaction patterns are tending to escape their historic boundaries.
Personal interaction span, defined as the number of unique persons an individual
interacts with over a one-year period, increased from approximately 25 in 1900 to
about 90 in 1967 to more than 250 in 2000s. Of course, connections in 1900 were
certainly different from those we have today through e-mail. Individual agents do not
only interact more than in the past, but they also interact more outside their
traditional scope of interactions. In network topology terms, the average span (i.e.,
the number of nodes a node is connected with) of interactions is widening, perhaps at
the expense of interaction depth (e.g. amount of personal mutual knowledge
involved in interactions, etc.).

These three trends in interaction patterns – increased volume of interactions,
increased span, and shortened degree – are occurring at the employee-to-employee
level (more inter-individual activities); at the intra-company level, where there are
more interactions between departments (more cross-functional activities or teams,
more meeting attendees coming from different departments); and at the company-
to-company level, (an increasing ratio of outside-facing employees to total employees
due to an increase in dis-intermediation).
These changes in interaction patterns can be explained by a shift in both interaction
supply and demand. The supply side analysis is rather straightforward. However,
supply side factors alone are insufficient to explain interaction volume growth.
Analysis of interaction demand is carried out in two steps: first, by describing the
problems, and second, by showing how interactions help solve them.

Hyper-competition and increasingly diversified
customer requirements have resulted in an extraordinary proliferation of products.
For example, between 1980 and 1998, the number of skin care products introduced
every year in the US increased from approximately 200 to around 1,200; the number
of new vitamins and dietary supplements increased from approximately 70 to 1,300;
the number of new varieties of ice cream and yogurt increased from 60 to 600.
Furthermore, the average number of stock keeping units (SKUs) at chain grocery
stores increased from less than 100 in 1930 to more than 20,000 in 1995.
In 1950s,  passenger cars may have consisted of one model that may exist in few versions only.
Today, , the number of different combinations the company offers has increased over 1,000-fold.
Yet, at the same time, product proliferation has increased signal uncertainty.
Statistical estimation theory, based on the law of large numbers, shows that product
proliferation makes understanding customers’ responses and related needs through
codified knowledge increasingly difficult or more uncertain.

One can make a broad distinction between products defined by a small number of
parameters, or commodities, and differentiated products, defined by a larger number
of parameters. Typically, commodities are fully defined by their price only, or by some
combination of variables leading to the overall price (e.g. oil). Differentiated prod-
ucts – such as luxury goods, software, or services – usually need a detailed description
(or perception) of their functional parameters. The total GDP for a country can be
broadly allocated to commodities or differentiated products.

Not only are signals more uncertain and granular but
they are also changing in time more frequently, resulting in higher temporal fre-
quency or signal volatility. Evidence of increased volatility can be most easily
seen in the stock markets. For example, the 120-month average of monthly amplitudes of
the Dow Jones Industrial Average increased from 5.5 percent in 1947 to approximately 9
percent in 2001. The 36-month average of the NASDAQ composite index jumped from 6 percent
in 1985 to 21 percent in 2001. Increased signal volatility means that the currently known
value of the signal will soon become inaccurate. Indeed, earnings have also become more
difficult to forecast; the average absolute error on analysts’ one-year forecasts has
increased from 4 percent to 8 percent over the last 10 years.

For some companies, despite increasing environmental complexity, this avoid strat-
egy can work. Some have the luxury to grow their narrow product line and related
technological portfolio without diversifying, or multiplying their various distribution
channels or geographic locations. However, these corporations will sooner or later be
confronted with complexity if they want to keep growing (through diversified prod-
uct ranges, new distribution channels, etc.). Savvier competitors could indeed start
building competitive barriers against these players by learning how to face complexity.
Other more diversified companies react to complexity by increasing their focus.
They streamline activities, simplify product ranges, separate apparently conflicting cus-
tomer demands (e.g. high quality and low prices) into different segments ascribed to
different product lines, or mechanically enhance signal-to-noise ratios by aggregating
market segments. Indeed, some major consumer goods companies shortened their US
product roster by a third, cutting new product launches by as much as 20 percent on
the grounds that “brand extensions had got out of hand.” Furthermore, many major automotive
players reduced their range by almost 20 percent.

Rules of organization design, which were meant to get rid of interaction needs,
thanks to a specific partitioning of the firm, must be radically re-written. The classic
rule of organizational design was to group activities, and then eventually define inter-
action patterns, i.e. linkages by means of hierarchical authority, functional coordina-
tion, or market mechanisms such as transfer prices. In following a “face” strategy,
corporations will need to follow a sequence in reverse: define the capability or know-
ledge needs, work out the patterns (topology, dynamics) of generative interactions
needed that would develop those capabilities, and eventually lay down the parts
involved in such interactions. Such an approach to design is very different from the
typical one derived from the cost perpective on interactions. Acknowledging the ICT-
enabled fall of such costs, the typical approach is to replace hierarchies with as many
market relationships as possible, including within organizations broken down into
very small units. This approach views interactions only as allocative and keeps the cost
perspective, as opposed to the one we advocate meant to harness the generative value
of interactions.