Overview of the concept of smart business networks and its implications for firms.

1. Journal of Information Technology (2004) 19, 228–233
& 2004 JIT Palgrave Macmillan Ltd. All rights reserved 0268-3962/04 $30.00
palgrave-journals.com/jit
Special edition
The emergence of smart business
networks
Peter Vervest1, Kenneth Preiss2, Eric van Heck1, Louis-Francois Pau1
¸
1
Rotterdam School of Management, Erasmus University Rotterdam, The Netherlands;
2
Ben Gurion University of the Negev, School of Management and Faculty of Engineering, Beer Sheva, Israel
Correspondence:
P Vervest, Rotterdam School of Management, Erasmus University, PO Box 1738, 3000 DR Rotterdam, The Netherlands.
Tel: þ 31 653 401485;
E-mail: vervest@d-age.com
Abstract
This article introduces the novel concept of smart business networks. The authors see the
future as a developing web of people and organizations, bound together in a dynamic and
unpredictable way, creating smart outcomes from quickly (re-) configuring links between
actors. The question is: What should be done to make the outcomes of such a network
‘smart’, that is, just a little better than that of your competitor? More agile, with less pain,
with more return to all the members of the network, now and over time? The technical
answer is to create a ‘business operating system’ that should run business processes on
different organizational platforms. Business processes would become portable: The end-
to-end management of processes running across many different organizations in many
different forms would become possible. This article presents an energizing discussion of
smart business networks and the research challenges ahead.
Journal of Information Technology (2004) 19, 228–233. doi:10.1057/palgrave.jit.2000024
Published online 30 November 2004
Keywords: smart business network; capabilities; agility; modularity; business logic; business operating
system
A joint vision yield an undefined direction, too uncertain for business
awhney and Parikh (2001) state that the intelligence executives. The correct approach therefore is to deal
S that is distributed and mutually supportive in a
network is much more than the simple summation
of the intelligence at the nodes of the network. Such
simultaneously with both short-term implementation issues
and long-term vision.
We should note that practice precedes theory in the
intelligence can enable executives and entrepreneurs to development of most technologies. Development of
decipher many of the phenomena shaping the future of the Internet preceeded before all the theory needed
their businesses. The intelligence of the network is was available. Or as another example, many iron bridges
augmented by its functionality – its ability to distribute, were built during the 18th and 19th centuries without
store, assemble, or modify information. ‘Dumb’ networks using theory and before the method of engineering statics
are simple pipes that transport information without and equilibrium was known. Even after today’s method of
enhancing it. A complex digital network can be ‘smart’; static force polygons had been invented, bridge designers
it can improve the utility of information in multiple did not use it. Eventually, the theory of static forces
ways. That, clearly, is synonymous with creating economic was used, an important step that allowed bridge designers
value. to extrapolate to much larger spans and enabled optimiza-
The pivotal question of this special issue concerns this tion of the use of materials. A similar process happens in
relationship between the intelligence of networks and the the case of smart business networks: these are being
smartness of the businesses that use these networks. The constructed today, and the practical construction is
papers in this issue aim to connect a long-term vision of followed by the development of suitable theories. Even-
smart business networks with short-term implementation tually, the theory should promote improved business
issues. Defining the long term without short-term details efficiency of such networks and should enable the
would not give practical results since the implementation expansion of business networks to much larger sizes. This
path would be unknown. On the other hand, defining special issue aims to contribute to the theory of smart
implementation details without a long-term view would business networks.

2. Emergence of smart business networks P Vervest et al
229
What are smart business networks? from the customer's direction the network
looks like a 'supply tree' or 'supply network'
We apply the word ‘smart’ to an action that is novel and
different, hence thought of as innovative. Smart actions
create remarkable, ‘better than usual’ results. Smart has a A
Eaton
connotation with fashionable and distinguished, but also C
with short-lived. What is smart today will be considered D E
General Motors dealer car owner
commonplace tomorrow. The word ‘smart’ in smart business
networks is therefore not an absolute but a relative term. B
Dana from the supplier to the (end) customer's
Smartness is a property whereby the network can create direction the network looks like a 'supply chain'
‘better’ results than other, less smart business networks or
other forms of business arrangement. While intelligence in goods flow
the communications systems and networks may have a more Figure 1 Seeing directional graph networks up or down the business network.
absolute meaning, smartness of business networks is
relative, time-bound and situation-bound. To be smart in
business is to be smarter than the competitors, just as an to GM, and Dana Corporation supplies subassemblies both
athlete considered fast means (s)he is faster than the others. to GM and to Eaton, making the relevant graph a network.
All three words in the title ‘smart business networks’ are If the supplier looks upward into the graph for a particular
necessary. The pair of words ‘smart business’ can apply to product, he will only see his own customer, that customer’s
any business without a network. A ‘smart network’ can customer, and so on. This structure is a chain. Hence we see
apply to a network that is not used for business or that the term ‘supply chain’ refers to the supplier’s view of
organization. A ‘business network’ is generic and includes where his product or service is taken into the end customer.
both smart and not-so-smart business networks. We define This term is, however, often used when the customer’s view
a smart business network as: is intended, where the correct term is ‘supply tree’ or
‘supply network’.
a group of participating businesses – organizational The pre-network economy changed slowly. Product shelf
entities or ‘actors’ – that form the nodes; life or turnover was slow. Once suppliers were chosen for a
linked together via one or more communication net- product, that structure remained in place for a long time.
works forming the links, or lines, between the nodes; The customer’s attention focused on the supply tree, and
with compatible goals; the supplier’s attention focused on the supply chain, since
interacting in novel ways; those were the long-lasting entities in the business structure
perceived by each participant as increasing its own value; that required only sporadic management attention. The
sustainable over time as a network; business network of qualified customers and suppliers
resilient if one or more businesses, nodes in the network, existed, but required little management attention. However,
malfunctions. the networked business environment is fast and agile.
Supply trees are selected from the network frequently and
rapidly, and they usually have short lifetimes because the
Smart networks, supply trees, and supply chains fleeting business opportunities have short lifetimes.
A network is a generalized graph consisting of nodes The Business Week (12 July 2004) issue gives many
connected by links. A fully connected network is a graph examples showing how, under the influence of modern
where each node is connected to all the others. According communication and production technology, allied to
to social network analysis the nodal degree would be gÀ1 globalization, markets are fragmenting. This development
for each node (g being the total number of nodes) and the was predicted in 1991 (Goldman et al., 1995) as an
network density would be 1 (Iacobucci, 1999). A business outgrowth of the inevitable move to an Agile competitive
network is unlikely to be fully connected but will be environment. Where the market leads, corporate strategic
partially connected. A tree is a hierarchical graph. It has no and structure necessarily follow. Smart business networks
cycles and each node has only one parent. A chain is a develop not only because technology permits them to
subgraph of a tree where each link connects exactly to one develop, but more significantly because markets and
other link. These constructs of ‘supply network’, ‘supply modern business competitiveness require such networks
tree’ or ‘supply chain’ are often confused. Let us try to in order to survive and thrive. Management attention then
explain. Imagine that we draw a graph of three nodes, A, B focuses on managing the network, on the processes for
and C, where each node represents a company, and we draw joining or leaving a network, and on processes by which to
a link between two nodes if one company (A) is a qualified select supplier trees from the network.
supplier or a qualified customer of the other company (B). Business networks and supplier trees and chains existed
Figure 1 shows an example with links between nodes A before and exist now; it is the centre of gravity of the
and B, B and C, and A and C. Note that the qualification of a organization and hence the focus of management attention
customer or supplier is a process that requires time and that has changed. Fine (1998), based on research into
money; it is far from instantaneous. When a customer several industry sectors including automotive and electro-
chooses specific suppliers from the list of qualified nics, wrote that the fundamental competitive capability was
suppliers in order to assemble a good or service, the graph not product or process, but the ability to construct and
as seen from the customer’s view will often (but not always) manage a supply chain. We can now go one stage further
be a tree. An example of a network is shown in Figure 1. and say that the fundamental competitive capability is to
The Eaton Corporation supplies automotive subassemblies construct and manage a smart business network.

3. Emergence of smart business networks P Vervest et al
230
Note that many of the papers in this issue that mention for unstructured tasks, for example, design and manufac-
supply chains, should have used the term supply trees or turing) and knowledge leverage (the creation of a network
networks. Such structures are particular instantiations from for leveraging skills and expertise). Clearly, such connec-
smart business networks. Business networks that are smart, tions are much more complicated to achieve and require
however, display quick connect and quick disconnect higher levels of mutual trust.
capabilities; they can pick the best capabilities from many What requires more attention is the capability to quickly
network actors, plug these capabilities together, and make disconnect, a process greatly influenced by risk and reward
these play in unison; they also control, or own, the business division (Goldman et al., 1995). This will be a vital element
logic for multi-actor execution of business processes. of a smart business network, because unless it is agreed
ahead of time how risk and reward, financial and otherwise,
will be allocated, serious problems of mistrust can develop.
What smart business networks should be able to do Methods for this allocation are being developed. Preiss and
The following capabilities are seen in smart business Ray (1998) published a method based on the relatively new
networks: mathematical ‘Cake-Splitting’ algorithm, that ensures
Establishment of common understandings: of meanings, elimination of jealousy. However, this method is suitable
words, of ethics and informal commitments, and of the only in the case of equal partners and divisible assets.
principles followed in contractual obligations; We need to understand the topology of business
Membership selection: the capabilities to decide which networks to fully realize their importance. Braha and Bar-
business entities can act as nodes of the network; Yam (2004, this issue) have done so. They examined the
Linking: the positioning and connecting of nodes to the statistical properties of networks of people engaged in
other parts of the network. The linking processes can distributed product development. The patterns of informa-
include the directories (search and select) and routing tion flows in such networks, say Braha and Bar-Yam,
(path finding) through the network as well as typical display similar statistical patterns as in other real-world
communications tasks such as handshake, authentica- networks of different origins such as information, biologi-
tion, and trust establishment; cal, and technological networks. Interestingly, the distribu-
Goal setting: the coordination mechanisms that deter- tion of incoming communication links always has a cutoff,
mine goals in the business network and the tasks and while the distribution of outgoing communication links is
responsibilities assigned to each member node; ´
scale-free (Barabasi, 2003). This could be consistent with
Interaction: the interactive, learning, and self-organizing Herbert Simon’s bounded rationality-argument (Simon,
capabilities that make the network generate novel results, 1969). It seems easier to transmit information (related to a
preferably those that no single member could achieve on network’s out-degree) than to process information (related
its own; to a network’s in-degree). Smartness would therefore be
Risk and reward management: the division of material related to the organizing capability of the information flows
results (profit and loss in a monetary but also in a fairly within the business network as well as to the topological
loose and generic sense) and the perceived value by each structure of the network.
of the participating business entities of its share;
Continual improvement: the capabilities and processes of
joining and leaving the network over time, of network Pick, plug, and play
renewal and sustainability; Establishing the – temporary – connection is not to say that
Fault tolerance: to malfunction of a node, for example a the network actors, or nodes, will interoperate. Interoper-
business malfunction or bankruptcy. ability can be facilitated by modularity. Garud et al. (2003)
define modularity as decomposability of a system by
grouping elements into a smaller number of subsystems.
Quick connect, and disconnect Schilling (2000) defines modularity as a continuous variable
It is particularly important that actors of a smart business of a system to separate and recombine its components as
network can ‘connect’ to other actors in the network. well as the rules governing the architecture for mixing and
Goldman et al. (1995) described this useful concept in their matching of these components. Products would be the
discussion of Virtual Organizations. Sanchez (1995) defines result of modular, Lego-like blocks combined in a specific
it as ‘quick connect electronic interfaces to coordinate way (Baldwin and Clark, 2000). According to Hoogeweegen
product creation resource chainsy to a network of product (1997), this would give the benefits of versatility (the
creation resources’. This capability of quickly-connected diverse set of products that an organization can produce)
plug-compatibility enables superior response speed and and agility (the ability to respond quickly to fulfil an
greater component variety when presented with new unpredictable customer order) while at the same time
product opportunities. delivering within the boundaries of allowed value chain
The concept of ‘quick connect’ is useful to smart business total costs and total lead times. Hoogeweegen et al. (1999)
networks: It includes a search-and-select behaviour by the develop a method to design modular business networks and
actors. Once the appropriate actor, or node, is found, and to optimize the allocation of tasks in the network. Wolters
the connection has been established, the process of (2002) discerns three dimensions of modularity: product
performing a business transaction can begin. Venkatraman modularity, process modularity, and value chain (business
(1994) defines the scope for business network redesign as network) modularity. Successful modular design, according
transaction processing and inventory management, fol- to Wolters, should not be restricted to products but must be
lowed by process linkage (interdependent process linkages concurrent in all three dimensions. It requires definition of

4. Emergence of smart business networks P Vervest et al
231
the function of each module and of the communication checks the events it receives against the current rules and
protocol between modules. It sees each module as a ‘black ‘fires’ the rules when their conditions are met.
box’ that will provide the functionality as required by the The creation of logic by individual actors in the business
modular design. This discussion may remind the reader of network takes a new meaning once this is linked together
objects in the software sense. It should not be surprising and managed through automata, independent of the
that ideas from one field, in this case computer program- originating actor(s). Such a business network operates a
ming, spill over and influence ideas in another field, in this ‘business operating system’ to run business processes on
case business management. different organizational platforms. Think of computer
However, modular designs require much more coordina- operating systems currently being developed to run
tion than nonmodular designs. What is an optimal degree application software on a large quantity of computers in a
of modularity (see Hoogeweegen and Vervest, 2005), or network, for example by using unused time on many PCs or
granularity of a system, or business network? Simon (1962) different hardware platforms that are networked together.
gives an important clue: formal process descriptions make Implementation of a business network operating system
us understand the relatively simple, dynamic laws that can encourages portablization of business processes, and
change states found in systems. Managing high degrees of facilitates the end-to-end management of processes running
modularity requires much more understanding of the across many different organizations in many different
processes that govern the plug compatibility of modular forms. It coordinates the processes among the networked
network components. businesses and its logic is embedded in the systems used by
Once a business network has been able to pick and plug these businesses. Competitive advantage is attained by
the appropriate product and process modules together, it smart humans who are augmented by the business
needs to be able to run, or ‘play’ these modules in the operating system. The question then is what to automate,
operational environments of the network actors. This is how to achieve that automation, how to connect the
what Webservices aim to achieve, apparently with some automation with humans, and where to draw the border
considerable success, see Van Hillegersberg et al. (2004). between the humans and the automation.
Own business logic Emerging structures of smart business networks
While workflows define how a process should run, the The levels at which one can analyse and understand the
business logic enacts, monitors and controls the process structure of business networks, are, from lower to top level:
flow in the technical environments of each of the business 1. the hardware and systems software infrastructure(s);
network actors; where necessary it passes control over to 2. the application software;
external systems to perform a task. This logic is controlled 3. management of an individual business – described in a
by business rules that take decisions on events depending networked systems concept as asset and event manage-
on the state of the various machines and processes linked to ment;
it. There are two critical components to this: the monitoring 4. the dynamic control and governance of the business
of all resources in the business network, and management network.
through rule-based event-correlation, see Figure 2. It
Studies such as Braha and Bar-Yam (2004, this issue) and
´
Barabasi (2003) suggest that the structure of a smart
business network emerges not only from the realities of the
business commerce which it enables, or the technology
being used, but also from the properties of the network
considered as a whole. A number of new determinants for
business network structure are proposed in:
Bounded group rationality that limits the actors’ group
mind share in a same way as for individuals (Simon,
1969). Measurements suggest that not only individual
human beings are limited by an inability to digest intense
input of data: a group of people, or a network of nodes,
show comparable limitations.
Dynamic emergence and decay of key information
brokers, information creators, and information users.
Measurement on networks shows that most nodes can be
categorised as one of these three types.
Resilience to the effects of both random and malicious
malfunction including business malfunctionings, for
example, bankruptcy and contractual breaches.
Each node in a business network has a minimum set
of attributes that define the node and link functions.
Figure 2 The network business operating system contains the overall business A minimal set of node functions includes capabilities
logic. for information generation, information use, and informa-

5. Emergence of smart business networks P Vervest et al
232
tion brokerage, the computational capabilities for each characterize smart business networks. Currently, we do
of these, and the allocation of attributes to each of these not yet have the suitable concepts and language to do so.
(such as permissions and rights to the capabilities, the right Does a requirement develop to embed higher-level
to create and delete nodes and links, and filtering functionality into lower levels? If so, should this be done
parameters). Link attributes include weights and amplifica- by including the functionality monolithically within the
tion factors. lower level, or should this be done while preserving
modularity? The former enhances static efficiency while
the latter enhances dynamic adaptability.
What smart business networks can do that no other How to design modularity? What spectrum of granula-
organizational form can do rities applies, what modules, architecture, and interface
We require a characterization of smart business networks specifications?
based on empirical observations. Examples from the past Is today’s limitation on the functionality of smart
give some clue. LiFung is Hong Kong’s largest export business networks the technical problem of managing
trader and innovator in the development of global supply complexity, or is it some other question, possibly in the
chain management (Magretta, 1998). It manages an field of social psychology?
expanding network of currently approximately 7500 Under what conditions can membership in a smart
suppliers around the world. ‘What we do is close to business network be so detrimental to the business
creating a customised value chain for every customer success of the individual company, that it best avoid such
order’ says Victor Fung. The LiFung operation captures membership?
the customer critical front end (design, engineering, There will come a day when smart business networks are
production planning) and back end (quality control, pervasively available and used. By then, complexity
testing, logistics) tasks. ‘We are smart about dissecting management will have a practical solution. What then
the value chain’. Smart business networks create a new will govern the design and operation of the smart
competitive game. The agile ability to quickly create or business network? Today, complexity management ap-
disconnect a specific tree or chain in a network, and the fact pears to constrain the expansion and management of a
that a given entity (company or profit centre) may belong smart business network, but what will be the limitation in
to different business networks for different strategic goals, the future? Research could already focus on this next
leads to three simultaneous levels of competition. A smart bottleneck to progress. Will complexity grow limited
business network may compete with another smart only by technology, or will trust and human relationships
business network; a smart business network may compete limit the design and operation of a smart business
in other strategies of one of its members, and the members network?
of a smart business network may compete in other
strategies while cooperating within the smart business
network. References
There are more examples of smart business networking.
Often, such examples are from using intelligent technolo- Baldwin, C.Y. and Clark, K.B. (2000). Design Rules: The Power of Modularity,
gies, identifiers, or ‘conscious’ machines. The interest of Cambridge, MA: The MIT Press.
´
Barabasi, A.-L. (2003). Linked – How Everything is Connected to Everything Else
business is rapidly growing. Bonabeau and Meyer (2001)
and What is Means for Business, Science, and Everyday Life, New York:
present studies in which companies such as Unilever, Penguin Group.
Southwest Airlines, Capital One, analyse operations as Bonabeau, E. and Meyer, C. (2001). Swarm Intelligence: A whole new way to
‘social insects’, because social insects work without super- think about business, Harvard Business Review 79(5): 107–114.
vision, are self-organizing, and can generate efficient Braha, D. and Bar-Yam, Y. (2004). Information Flow Structure in Large-Scale
solutions to difficult problems, although the interactions Product Development Organizational Networks, Journal of Information
Technology, this special issue.
themselves might be simple. Complex collective behaviour
Brooks, R.A. (2002). Flesh and Machines, New York, NY: Pantheon Books.
can emerge from individual network actors following Business Week (2004). The Vanishing Mass Market, Cover Story, McGraw-Hill:
simple rules (Brooks, 2002)! That is a powerful notion. New York, 12 July 2004.
Fine, C.H. (1998). Clock Speed: Winning Industry Control in the Age of
Temporary Advantage, Massachusetts/New York: Perseus Books.
Garud, R., Kumaraswamy, A. and Langlois, R.N. (2003). Managing in the Age of
The research challenges ahead
Modularity: Architectures, Networks, and Organizations, Malden, USA:
Smartness may emerge spontaneously and not be inten- Blackwell Publishers.
tionally designed. Today’s pervasive communications Goldman, S.L., Nagel, R.N. and Preiss, K. (1995). Agile Competitors and Virtual
technologies will have a much more profound impact on Organizations: Strategies for Enriching the Customer, New York: Van
how businesses cooperate and compete than we may be Nostrand Reinhold.
aware of today. We propose that smart business networks Van Hillegersberg, J., Boeke, R. and Van Den Heuvel, W.-J. (2004). The
should be a recognized direction for management and Potential of Webservices to Enable Smart Business Networks, Journal of
Information Technology, this special issue.
technology research. As a matter of priority, the following Hoogeweegen, M. (1997). Modular Network Design – Assessing the impact of
research questions are offered to advance the art, and EDI, Ph.D. Series in General Management 26, Rotterdam School of
science, of smart business networking: Management, TRAIL Research School, Rotterdam.
Hoogeweegen, M., Teunissen, W.J., Vervest, P.H.M. and Wagenaar, R. (1999).
It is common practice to characterize a single company, Modular Network Design: Using information and communications
usually by defining its product or service sector, possibly technology to allocate production tasks in a virtual organization, Decision
with other characteristics. We need a systematic way to Sciences 30(4): 1073–1103.

6. Emergence of smart business networks P Vervest et al
233
Hoogeweegen, M. and Vervest, P.H.M. (2005). How Much Modularity?, in: P. in digital age companies (London – Amersfoort –
Vervest, E. van Heck, K. Preiss and L. Pau (eds.) Smart Business Networks, Sunnyvale). His specific field of research concerns the
Berlin: Springer.
Iacobucci, D. (1999). Graphs and Matrices, in S. Wasserman, K. Faust (eds.)
development and application of enabling technologies for
Social Network Analysis: Methods and Applications, Cambridge: Cambridge smart business networks.
University Press.
Magretta, J. (1998). Fast, Global, and Entrepreneurial: Supply chain Kenneth Preiss holds the Sir Leon Bagrit chair at Ben
management Hong Kong style, Harvard Business Review 76(10): 103–114. Gurion University in Beer Sheva, Israel, is an honorary
Preiss, K. and Ray, M.R. (1998). A Method for Sharing Earnings in a Virtual member of the ASME, and is a member of many other
Organization, Agility and Global Competition 2(1): 33–40. national and international scientific and technical bodies.
Sanchez, R. (1995). Strategic Flexibility in Product Competition, Strategic He has published over 200 research papers and reports, and
Management Journal 16: 135–159.
Sawhney, M. and Parikh, D. (2001). Where Value Lives in a Networked World,
has co-authored with Steven Goldman and Roger Nagel the
Harvard Business Review 79(1): 79–86. books Agile Competitors and Virtual Organizations: Stra-
Schilling, M.A. (2000). Toward a General Modular Systems Theory and its tegies for Enriching The Customer, and Cooperate to
Applications to Interfirm Product Modularity, Academy of Management compete: Building Agile Business Relationships.
Review 35(2): 312–334.
Simon, H.A. (1962). The Architecture of Complexity, Proceedings of the Eric van Heck is a Professor of Electronic Markets at
American Philosophical Society, No. 106, pp. 467–482. Erasmus University’s Rotterdam School of Management,
Simon, H.A. (1969). The Sciences of the Artificial, Cambridge, MA: The MIT
where he teaches in the international M.B.A. program and
Press.
Venkatraman, N. (1994). IT-enabled Business Transformation: From
in the Global eManagement (GeM) program. His research
automation to business scope redefinition, MIT Sloan Management Review concentrates on the designs of electronic markets and that
35(2): 73–87. of business modularization. In his research he helps
Wolters, M.J. (2002). The Business of Modularity and the Modularity companies to develop innovative, electronic auctions.
of Business, ERIM Ph.D. Series in Management, No. 11, Trail
Thesis Series T2002/1, The Netherlands, TRAIL Research School. Louis-Franc¸ois Pau is Professor of Mobile Business at the
Rotterdam School of Management, besides being CTO of
the Network systems division of Ericsson. He is teaching
About the authors graduate and executive courses at RSM as well as at other
Peter Vervest is Professor of Business Telecommunications centers worldwide. His research at RSM focuses on mobile
at Erasmus University’s Rotterdam School of Management services, mobile and high-tech industries, and on innova-
and partner of D-Age counsellors and investment managers tive business models in the aforementioned.