A Model of the Diffusion of Technology into SMEs

Abstract

The paper considers technology diffusion and develops a model at the level of the SME. Technology diffusion in the form of new or improved technology, the transmission of knowledge or technical expertise is investigated. This involves spillovers through formal and informal networks enabling learning by interacting and an absorptive capacity to assimilate new technology developed elsewhere. A model of technology diffusion is developed including external sources, channels of technology transfer, and mechanisms involved in the transfer of technology into the innovative SME. The model is related not only to "best practice" but also to cases where "low" activity can be improved. Implications for policy relevant to technology and entrepreneurship arising from the model are also investigated and conclusions are drawn.

Full text

A MODEL OF THE DIFFUSION OF TECHNOLOGY INTO SME's
Thomas Brychan
Abstract
The paper considers technology diffusion and develops a model at the level of the SME. Technology diffusion in
the form of new or improved technology, the transmission of knowledge or technical expertise is investigated. This
involves spillovers through formal and informal networks enabling learning by interacting and an absorptive
capacity to assimilate new technology developed elsewhere. A model of technology diffusion is developed
including external sources, channels of technology transfer, and mechanisms involved in the transfer of technology
into the innovative SME. The model is related not only to "best practice" but also to cases where "low" activity can
be improved. Implications for policy relevant to technology and entrepreneurship arising from the model are also
investigated and conclusions are drawn.
1. Introduction
It is evident that Governments today regard technology diffusion as an important route to increased
competitiveness, especially diffusion into Small and Medium-sized Enterprises (SMEs) (La Rovere, 1998) with
advantages of flexibility, dynamism and responsiveness. However, SMEs have disadvantages related to the lack of
technological and financial resources which can lead not only to problems in their ability to source technology but
also in their capability to absorb it into their organisation and diffuse it into their industrial sector (Jones-Evans,
1998).
The objectives of the paper are threefold: first, to investigate technology diffusion in the form of new or improved
technology through formal and informal networks enabling learning by interacting; second, to develop a model of
technology diffusion including external sources, channels of technology transfer, and mechanisms involved in the
transfer of technology into the innovative SME; and third, to relate the model to "best practice" and to note
situations where "low activity" can be improved. Finally, the implications for policy relevant to technology and
entrepreneurship arising from the model of technology diffusion are investigated and conclusions drawn.
Since there is a time dimension involved in the study of the diffusion of technology into SMEs, similar to other
investigations of innovation, theories based on these studies will tend to lag behind the "best" current practices. All
models of technology diffusion, including refined models such as the Bass Norton model, are a simplification of
reality (Islam, Meade, 1997) and, therefore, have a measured influence upon policy. One theoretical model that has
informed policies is the Centre Periphery Model (Schon, 1971) which rests on three basic assumptions:
i) the technology to be diffused exists prior to its diffusion,
ii) technology diffusion takes, place from the source outwards to SMEs, and
iii) the support of technology diffusion involves incentives, provision of resources and training.

By applying the Centre-Periphery Model to SME Technology Transfer Network Theory it is possible to construct
what can be described as the "Hub and spoke" or "Star" network. This is a simple construct that can be used as a
building block for more intricate networks. Diffusion will take place from the source of the technology through
channels by a "diffuser", using a transfer mechanism, to the SME. The effectiveness of the system will depend
upon the resources available to the external source to enable the transfer, the efficiency of the diffuser and the
mechanism involved, and the ability of the SME to acquire technology. The scope of the system will vary directly
with the level of technology and the flow of information.
2. Technology Diffusion
When a new technique has been adopted the speed at which other SMEs adopt may differ widely. This leads to
what can be called the rate of diffusion (imitation). The rate of diffusion will be faster, the greater the improvement
over existing technology and, the lower the cost of the technology in general (Roy, Cross, 1975). Using the
definition of Bradley, et al, (Bradley, McErlean, Kirke, 1995) technology diffusion can be defined as the spread of
a new technique from one SME to another ('interfirm diffusion') (Stoneman and Karshenas, 1993). The two
principal types of technology diffusion are "disembodied" diffusion (the transmission of knowledge and technical
expertise) and "embodied" diffusion (the introduction into production processes of machinery, equipment and
components incorporating new technology) (Papaconstantinou, Sakurai, Wyckoff, 1995). Research spillovers are
the means by which new knowledge or technology developed by one firm become potentially available to others
and the absorptive capacity of the receiving firms will determine the extent to which the technology is incorporated.
The time pattern of adoption and the speed at which it takes place are distinct happenings. The exploration time
period when implementing an innovation can provide imitators with a " window of opportunity" to proliferate
(Jayanthi, 1998). Empirical studies suggest that the adoption of a new technology follows a bell-shaped, or normal,
distribution curve (Norris, Vaizey, 1973). By plotting cumulatively this shows the number of SMEs who have
adopted a new technology in any given year, and the distribution will give an 'S'- shaped curve. (It was Gabriel
Tarde who in the Laws of Imitations, 1903, proposed that adoptions plotted against time assume a normal
distribution, or if plotted cumulatively assume the 'S'-shaped curve.) (Baker, 1976) An 'S'-shaped distribution, not
necessarily derived from a normal distribution, shows the spread of most new technology. There are two general
reasons for the occurrence of this distribution.
The diffusion process for SMEs is a learning process.
SMEs who are potential users have to become aware of the technology and then to attempt to evaluate it.
Consequently they may use the technology on a trial basis. The learning process takes place at this stage.
Information about the technology has to be disseminated, and as it is adopted by other SMEs, or by the SME on an
experimental basis the information becomes more reliable.
The importance of accumulated knowledge and expertise is an important factor determining whether firms are
likely to adopt new technology or to act as sources of innovation (Gurisatti, Soli, Tattara, 1997). 'Bugs' will be
overcome, which will in turn reduce the risk of adopting the technology. The concept of the individual SME
learning curve can be extended to a network group of SMEs where experience with a new technology increases as
each successive SME adopts the new technology. As a result, the distribution of SMEs adopting a technology might
be expected to yield a normal curve.

An interaction effect occurs for SMEs.
When only a small number of SMEs have adopted a technology, there are a small number of diffusers who can
generate information on the technology and from whom the technological idea can spread. Diffusion rates at this
point are low. When the number of SMEs using the technology increases the "information base" broadens and
because there is still a considerable number of SMEs who have not adopted the new technology the rate of
diffusion increases. When there is a large proportion of SMEs using the technology the number of potential SMEs
still remaining becomes small. The remaining SMEs will be resistant to change and there will be a slow down in
the cumulative number of SMEs using the new technology. This will yield an 'S'-shaped curve. The first formal
study of diffusion was the spread of hybrid corn (Grilliches, 1960). The adoption rate in different states in the USA
was studied and it was found that there were significant differences between states in the rate of hybrid corn
adoption. Logistic growth curves were fitted by Grilliches to his data and the parameters found from the curves for
the different states showed wide variations.
Another formal study of the rate of diffusion was carried out by Mansfield who studied the rate of diffusion of
twelve innovations in four industries - coal, iron and steel, brewing and rail (Mansfield, 1961, 1968). Although
small firms were not included in the analysis, for medium-sized and large firms in most cases, the spread of
innovations over time approximated the 'S'-shaped curve. According to Mansfield the spread of innovations is best
described by a logistic curve.
Despite the shape of the curve for technology diffusion appearing 'S'-shaped, there will be differences in the speed
at which technology is diffused and the length of the diffusion process. Both within and between industries there
will be considerable variations in the rate of the diffusion of technology between SMEs.
Important factors which appear to affect the rate of diffusion (speed at which a new technology is accepted) are the
characteristics of the SME and the characteristics of the technology itself. Early work on the categories of adopters
found that further to adoption following a normal distribution curve the distribution could be used to show the
categories of adopters (Rogers, 1962). Table 1 shows the categories of adopters with the majority of adopters lying
between the mean and the mean minus/plus the standard deviation on the normal distribution curve.
Table 1 - The Categories of Adopters
Categories
Innovators Early Adopters Early Majority Late Majority Laggards
Number of Adopters
2.5% 13.5% 34% 34% 16%
x - 2σ x - σ x x + σ
Years

The categories of adopters can be described as follows:
•
Innovative SMEs are those who want to explore new technologies. They will have relationships with other
SMEs in their network, and with suppliers and customers.
•
Early adopters will be SMEs who will adopt new technology if it is to their advantage. Since they will act
as 'opinion leaders' their influence will be greater than innovative SMEs.
•
The early majority will be intentional while the late majority will be sceptical and will adopt when the
technology has diffused.
•
Last, the laggards will be so late adopting a new technology that it will have been superseded.
The categories of adopters shows that SMEs which adopt an innovation independently are innovators (Tassopoulos,
Papachroni, 1998). Early research studies aimed at defining the characteristics of adopters found that early adopters
relied to a greater extent on impersonal sources of information from wider and more sources (Rogers, 1962). They
used sources in close contact with the origin of new ideas including technical journals. SMEs that are early adopters
will tend to be "technically progressive" and will be close to the best that can be achieved in the practice of
applying technology (Carter, Williams, 1957). On this assumption a progressive SME will take a wide range of
authoritative technical journals, will have a variety of contacts with sources of technology including similar SMEs,
and will assess ideas from these sources. It is expected that communication within the SME will be well organised
and co-ordinated and there will be a willingness to share knowledge with other SMEs in its network. A progressive
SME will set its standards by reference to best practice in other SMEs.
The speed of diffusion will also be faster the greater the awareness of SMEs to the advantages of adopting a new
technology. The process of communication will be important here as well as the ability of SMEs to assess the
merits of the technological advance. An SME is more likely to adopt a new technology as it diffuses due to being
under increasing competitive pressure to do so, through the technology becoming more attractive, and as a result of
information about the technology being broadcast from an increasing base (Green, Morphet, 1975).
3. Technology Transfer Networks
Technology transfer networks are of particular importance to SMEs with little in-house resources and experience to
explore the potential of new technologies. SMEs usually lack awareness to the value of technology transfer, are
diffident to enabling services, and therefore rely on co-operation with others. Two basic mechanisms available to
SMEs are technology exchange (technology passed from one SME to another) and technology exploitation
(technology transferred to an SME from an external source).
Technology transfer networks enable SMEs to reach a common understanding regarding new technologies quickly.
Important aspects of SME technology transfer networks are the type and size of the network. Whereas, small
networks appear more efficient, since communications are easy and network dynamics controllable, large networks
benefit from a greater pool of resources. There are four principal types of networks. The "star" network has already
been reported. A "nodal linkage" network involves SMEs on an equal footing and is not suitable for SMEs with
different levels of experience. "Ad hoc" or "informal" networks are those without a formal structure where SMEs
intimately know each other concentrating communication where required. These tend to be mature networks, but
are not well suited for heterogeneous groupings, or those with little commonality between SMEs. "Regional"
networks are the most complex type consisting of multi-tiered structures linking local networks. These are suitable
for heterogeneous SMEs. The descriptions of these four types of network are exemplars in their purist form.
Networks adapt to changing internal and external factors and evolve from one (centre periphery) to another (multi-
tiered). Although co operation with other technology transfer networks provides the possibility of accessing a wider
contact base it carries with it some competitive risk.

4. A Model of Technology Diffusion
A model of the diffusion of technology into SMEs can be described as innovation (supply) from the source of
technology (origins) and diffusion (demand) to the SME (destination). The model can be expressed concisely in
algebraic form:
Origins i = 1, 2,... m
Destinations j = 1, 2,... n
Supply at each origin a
i
Demand at each destination b
j
Constraint: supply = demand []a
i
= []b
j
In order to find a solution we must specify the variable x
i j
as the unit(s) of technology transferred from origin i to
destination j over time t.
All supply
j
[]x
ij
= a
i
j = 1, 2 ... n
All demand
i
[]x
ij
= b
j
i = 1, 2 ... m
The diffusion of technology D can be expressed:
D=[
m
Σ
i=1
+
n
Σ
j=1
]x
ij
i = 1, 2,... m and j = 1, 2 ... n
The rate of diffusion of a new technology to SMEs can be likened to waves of adoption involving distinct time
packages. This is illustrated in Table 2.
Table 2 - The Rate of Diffusion
Innovators Imitators
Waves of
adoption
1
st
Wave 2
nd
Wave 3
rd
Wave 4
th
Wave 5
th
Wave
Categories
Innovators
Early
Adopters
Early
Majority
Late
Majority
Laggards
Number of
Adopters
2.5% 13.5% 34% 34% 16%
Time periods
Period 1 Period 2 Period 3 Period 4 Period 5
Diffusion for each
period
[
m
Σ
i=1
x
ij
]
1
[
n
Σ
j=1
x
ij
]
2
[
n
Σ
j=1
x
ij
]
3
[
n
Σ
j=1
x
ij
]
4
[
n
Σ
j=1
x
ij
]
5
The rate of diffusion (R) can be calculated according to time (t) (number of years) as follows:
R= [
m
Σ
i=1
+
n
Σ
j=1
] x
ij
/ t i= 1,2, ... m and j= 1, 2 ... n
This equation is a temporal model of technology diffusion which measures the speeds of diffusion (or rates of
technology transfer) (Bradley, McErlean, Kirke, 1995).
A hypothetical example of the first and second waves of diffusion involving sources of technology and SMEs is
illustrated in Figure 2--omitted.

The example illustrates that technology transfer is an active process whereby technology is carried across the
border of two or more social entities (the external source and the SME), and technology transfer channels are the
link between the entities (in which various technology transfer mechanisms are activated) (Autio, Laamanen,
1995). A technology transfer mechanism is defined as any specific form of interaction between entities during
which technology is transferred (Autio, Laamanen, 1995). The ability to establish external linkages is of critical
importance to SMEs and a critical mass of SME users will spread the usage and acceptance of the technology (Jain,
1997). The success or uptake of technology depends on how successful the performed community of (implied or
ideal) users match the characteristics of actual users (Woolgar, Vaux, Gomes, Ezingeard, Grieve, 1998). Success
can be achieved by "configuring the user". Further to this Malecki has stated that "as new technology and products
are learned, acquired, evaluated, and improved upon, a firm or region comes to know about best-practice
technology ..." (Malecki, 1991, p.122). Laranja calls these "cumulative processes of learning" (Laranja, 1994, p.
173).
5. "Best practice"
Technology transfer networks are one of the best forums for SMEs to learn from each other, to exchange
experiences, and to diffuse technology, The typical areas where the benefits of "best practice" can be found are
technology transfer skills (determining an SME's needs by auditing and drawing-up agreements and contracts),
technological expertise and know-how (including standards and regulatory issues), service provision (assembling
the provision of services), and management and organisation (public relations) (Commission of the European
Communities, 1998).
Networks are usually segmented by geographical region, industry sector or by technology and they can work with a
mixed sector-technology focus. The danger with specialisation is that is carries the disadvantage that eventually the
potential market will be exhausted. It is possible to overcome this by anticipating and looking for opportunities in
complementary technology areas.
"Best practice" procedures for the diffusion of technology within networks usually include minimum standards for
the SMEs, external funding apportionment, expected performance, and confidentiality. Procedures will usually
become less formal over time due to ideal size attainment and growth realisation. Good practice for the successful
operation of a network is the realisation by SMEs that it is not only an alliance of enterprises but also a partnership
of entrepreneurs. (Entrepreneurs will act as technological gatekeepers and will have an important role to play in the
operation of networks.) (Thomas, 1999) This needs to be reflected in network communications and good
relationships between the SMEs will form the basis of good practice for the operation of the network.
Success in the diffusion of technology within networks is often the result of SMEs following "best practice" and
this usually involves performance management. This is not easy to attain since the process of technology transfer
can be long and without success, the results of the network are difficult to define and there may be discrepancies
between the SMEs. "Low" activity may arise due to conflicts in a network. When these are efficiently managed and
resolved they provide opportunities for the SMEs to broaden their experience and widen their understanding of
other SMEs' views. When they are not conflict may lead to "low" activity. Conflict management and identification
will form part of the "best practice" of successful technology diffusion. Typical examples of "low" activity are
misunderstanding between SMEs, different objectives and motives and under-performance of an SME.

6. Implications for Policy
The implications for policy of a model of the diffusion of technology into SMEs, and the technology processes
involved, necessitates the need to formulate technology transfer related action. This includes raising SMEs'
awareness of the potential of technology transfer to help solve problems and the existence of networks to provide
practical support. Once SMEs comprehend the possible benefits of technology transfer they will need more help to
realise the benefits. Two further types of action to achieve this are specific support provided to individual SMEs
(assistance during the establishment of network relationships) and technology transfer support to SMEs in general
(to foster technological knowledge and establish network links from external sources such as universities and
research providers for the dissemination of know-how into SMEs).
Coupled to the three forms of policy action described above the three main types of external sources involved in the
diffusion of technology to SMEs are public and non-profit organisations (regional and national development
organisations (RDOs/NDOs), regional technology advice centres (RTACS) and chambers of commerce), private
consultants (technology brokers, management consultants, patent attorneys), and Research and Technology
Organisations (RTOs) (contract research firms, science parks and technology centres). Technology transfer
networks may comprise all three types although homogeneous networks are usually easier to form and develop.
Amongst the three types public bodies are best placed to undertake policy programmes, private companies
concentrate on providing focused assistance and RTOs provide technology knowledge and know-how. For SMEs
involved in technology transfer networks key mechanisms include information transfer (newsletters and databases),
technology transfer (R&D audits), skills transfer (training) and specialist support (financial guidance). Value for
money of the mechanisms will be a key policy measure. There will need to be care that changes in policy will not
make an SME withdraw from technology transfer activities and that policy reacts to difficult situations by
providing SMEs with incentives.
7. Conclusions
Although the variables involved in the model appear to be the most important influences on technology diffusion
into SMEs there will also be a multiplicity of influences that accelerate or alleviate the rate of diffusion. This
spectrum of influences on diffusion rates broadens when considering technology transfer among the various
different SMEs in multi-tiered networks. An extension of the hypothetical example of diffusion (Figure 2--omitted)
is the diffusion of technology into SMEs through multi-tiered networks (Figure 3--omitted). In these SMEs'
sociological forces will have an important role to play. The rate of adoption of a new technology will be faster if it
is compatible with the previous experience and present normative values of SMEs. Other influences on the speed of
diffusion include the complexity of the new technology and random influences.
The model illustrates that the successful diffusion of a new technology involves considerably more than technical
competence. Many complementary factors will be prominent and an SME may be retarded in its acquisition of
technology by other SMEs who are slow to adopt. 'Laggards' can have a deleterious effect on the diffusion of
technology by other SMEs. The rapid diffusion of a technology will be facilitated by a willingness of SMEs to
make adjustments.
Acknowledgements
I would like to thank Professor Dylan Jones-Evans, Director of the Welsh Enterprise Institute, for guiding me to an
investigation of a model of the diffusion of technology into SMEs and for his comments on the paper, and Dr. Rod
Gunn, Reader in the Business School, University of Glamorgan, for his advice on the algebraic form of the model.

References
Autio E., Laamanen T., (1995), Measurement and evaluation of technology transfer: review of technology transfer
mechanisms and indicators", Int. J Technology Management, Vol. 10, Nos. 7/8, pp. 643-664.
Baker M.J., (1976), Chapter 7, "Diffusion Theory and Marketing", in Marketing Theory and Practice, London,
Macmillan, pp. 119-131.
Bradley A., McErlean S., Kirke A., (1995), "Technology Transfer in the Northern Ireland food processing sector",
British Food Journal, Vol. 97, No. 10, pp. 32-35.
Carter C., Williams B., (1957), Industry and Technical Progress, London, Oxford U.P.
Commission of the European Communities, (1998), Good Practice in Technology Transfer, DGXIII
Telecommunications, Information Market and Exploitation of Research, Luxembourg, EU.
Green K., Morphet C., (1975), Section 7, "The Diffusion of Innovations", in Research and Technology as
Economic Activities, York, Science in a Social Context (SISCON), pp. 45-47.
Grilliches Z., (1960), "Hybrid Corn and the economics of innovation", Science, 29 July, 275-280.
Gurisatti P., Soli V., Tattara G., (1997), "Patterns of Diffusion of New Technologies in Small MetalWorking Firms:
The Case of an Italian Region", Industrial and Corporate Change, Vol.6, No. 2, March, pp. 275-312.
Islam T., Meade N., (1997), "The Diffusion of Successive Generations of a Technology: A More General Model",
Technological Forecasting and Social Change, Vol. 56, No. 1, pp. 49-60.
Jain R., (1997), "A Diffusion Model for Public Information Systems in Developing Countries", Journal of Global
Information Management, Vol. 15, No. 1, Winter, pp. 4-15.
Jayanthi S., (1998), "Modelling the Innovation Implementation Process in the Context of High-Technology
Manufacturing: An Innovation Diffusion Perspective", Cambridge, ESRC Centre for Business Research.
Jones-Evans D., (1998), "SMEs and Technology Transfer Networks - Project Summary", Pontypridd, Welsh
Enterprise Institute, University of Glamorgan.
La Rovere R.L., (1998), "Diffusion of information technologies and changes in the telecommunications sector: The
Case of Brazilian small- and medium-sized enterprises", Information Technology and People, Vol. 11, No. 3, pp.
194-206.
Laranja M., (1994), "How NTBFs Acquire, Accumulate and Transfer Technology: Implications for Catching-Up
Policies of Less Developed Countries such as Portugal", in New Technology-Based Firms in the 1990s, (ed. by
Oakey R.), London, Paul Chapman, pp. 169-180.
Malecki E.J., (1991), Technology and economic development: the dynamics of local, regional, and national
change, New York, Longman.
Mansfield E., (1961), "Technical change and the rate of imitation", Econometrica, October, pp. 741766.

Mansfield E., (1968), Chapter 4, "Innovation and the Diffusion of New Techniques", in The Economics of
Technological Change, New York, Norton, pp. 99-133.
Norris K, Vaizey J., (1973), Chapter 7, "The Diffusion of Innovations", in The Economics of Research and
Technology, London, George Allen and Unwin, pp. 86-103.
Papaconstantinou G., Sakurai N., Wyckoff AW., (1995), "Technology Diffusion, Productivity and
Competitiveness: An Empirical Analysis for 10 Countries, Part 1: Technology Diffusion Patterns", Brussels,
European Innovation Monitoring System (EIMS).
Rogers E., (1962), Diffusion of Innovations, New York, Collier-Macmillan.
Roy R., Cross N., (1975), Section 3.1.3, "Diffusion", in Technology and Society, T262 2-3, Milton Keynes, The
Open University Press, pp. 36-38.
Schon D.A., (1971), Chapter 4, "Diffusion of Innovation", in Beyond the Stable State, London, Temple Smith, pp.
80-115.
Stoneman P., Karshenas M., (1993), "The diffusion of new technology: extensions to theory and evidence", in New
Technologies and the Firm: Innovation and Competition (ed. By Swann P.), London, Routledge, pp. 177-200.
Tassopoulos A., Papachroni M., (1998), Penetration models of new technologies in Greek small and medium-sized
enterprises, Int. J Technology Management, Vol. 15, Nos. 6/7, pp. 710-720.
Thomas B., (1999), "The Role of Technological Gatekeepers in the Management of Innovation in SMEs: The
Regional Context", COrEx, 12 March.
Woolgar S., Vaux J., Gomes P., Ezingeard J.-N., Grieve R., (1998), "Abilities and competencies required,
particularly by small firms, to identify and acquire new technology", Technovation, Vol. 18, Nos 8/9, pp. 575-584.
About the Author
Brychan Thomas, Research Fellow, Welsh Enterprise Institute, Business School, University of
Glamorgan
Contact details:
Dr. Brychan Thomas
Welsh Enterprise Institute
Business School
University of Glamorgan
Pontypridd CF37 1DL
Wales
Tel: +44 1443 483290 Fax: +44 1443 482380 E-mail: bcthomas@glam.ac.uk