Social Networking, 2013, 2, 9-18
http://dx.doi.org/10.4236/sn.2013.21002 Published Online January 2013 (http://www.scirp.org/journal/sn)
Opinion Leaders and Lead Users in Marketing and
Management Accounting and Impact on
Small Business Performance
Katharina Simbeck
Entrepreneurship and Innovation Management, Technical University, Berlin, Germany
Received November 10, 2012; revised November 16, 2012; accepted January 8, 2013
ABSTRACT
This paper empirically investigates into the business performance benefit that lead users or opinion leaders among small
business owners draw from their higher involvement in management accounting or marketing topics. This work also
contributes to a better identification of network members’ roles solely through their ties between each other. Indeed,
lead users and opinion leaders can be differentiated by a higher degree centrality in comparison to their peers. However,
being an opinion leader or a lead user does not yield a measurable business benefit to the small businesses studied in
this sample.
Keywords: Social Network Analysis; Opinion Leaders; Lead Users; Small Businesses; Network Centrality
1. Introduction
The concept of opinion leadership was introduced by the
sociologists Lazarsfeld et al. [1] for a sub-group of indi-
viduals who have the power to shape a group’s percep-
tions. The concept of “lead users” was developed by von
Hippel [2,3] to characterize a sub-set of product users
that are ahead to their peers in terms of developing new
product needs. Von Hippel [2] argues that those lead
users can be closely integrated into new product develop-
ment. The concept of opinion leaders was subsequently
taken up also by researchers in marketing, aiming to un-
derstand how opinion leaders can be used to speed up
new product diffusion. Companies turn to lead users to
co-innovate relevant new products which have the poten-
tial of bigger commercial success [4]. Identifying lead
users and opinion leaders may significantly improve
marketing efficiency by targeting the right customers at
the right point in time of product life cycle at optimized
cost. Hence research in marketing has a track record of
identifying opinion leaders and lead users and their cha-
racteristics (see for example Darden [5]).
The concept of opinion leadership and lead user prop-
erty are widely used in social network analysis as they
emerge in the interaction between subjects and can thus
be considered as truely “social” properties of subjects. So
far, opinion leaders and lead users are being identified
using standard questionnaires or observed behavior (e.g.
purchase of product at early stage of product lifecycle).
With the increasing diffusion of electronic networks,
waste amounts of data on network position become avai-
lable. It is therefore of interest to identify lead users and
opinion leaders solely based on network data as it is po-
tentially cheaper and faster in comparison to using ques-
tionnaires or observing behavior.
Belz and Baumbach [6] have shown that using internet
ethnography methods (“netnography”), namely the ana-
lysis of posts in internet forums, classifies almost half of
internet users correctly as lead users/non-lead users. Bil-
gram et al. [7] have also studied internet communities
and identified several factors for the identification of lead
users, such as “being ahead of market trend, high ex-
pected benefits” or user expertise. Even though Belz and
Baumbach [6] and Bilgram et al. [7] refer to data avail-
able electronically on the internet, their approaches still
require comprehensive analysis of opinions expressed in
online communities. Hill et al. [8] successfully identified
potential customers to be targeted through marketing by
using real telecommunications data on people connected
to adopters of the product.
In contrast Kratzer [9] have found out that lead users
and opinion leaders among school children with regards
to toys can be distinguished by specific ego-network
properties, specifically betweenness for lead users and
degree centrality for opinion leaders.
In a given network structure, ego has a given set of
ego-network properties, e.g. a given centrality or be-
tweenness. However with regards to opinion leadership
C
opyright © 2013 SciRes. SN

K. SIMBECK 10
or lead user role ego can have different different interests
for different subjects. The groups of opinion leaders and
lead users tend to overlap [10]. That overlap was theo-
retically confirmed by Bilgram et al. [7]. Spann [11] ar-
gue that opinion leadership is a characteristic of the lead
user variable and therefore use an opinion leader ques-
tionnaire as one of three criteria in identifying lead users.
The object of this research is to study the ego-network
properties and lead user/opinion leader roles for two
management subjects, namely marketing and manage-
ment accounting.
Thus this research aims to contribute to a better under-
standing of network members solely through their ties
between each other. Based on the empirical data used for
this research lead users and opinion leaders cannot be
really differentiated based on ego network properties.
However, the tendency that lead users have a higher
number of contacts especially on a national as opposed to
local level, i.e. degree centrality and opinion leaders are
located on more paths between others (i.e. higher be-
tweenness centrality) can be confirmed. Furthermore we
find that lead users or opinion leaders do not benefit from
their higher involvement in management accounting or
marketing topics through better business performance.
2. Theory
2.1. Network Position and Network Role
Two distinctive network roles will be discussed: lead
users and opinion leaders. Lead users apply innovations
faster in comparison to peers. Opinion leaders commu-
nicate more than others on specific topics.
2.2. Network Position
Social network analysis has defined various concepts to
describe the position of an actor in a network. Centrality
is one of the concepts most frequently used. There are
several definitions for the centrality of an actor. It can be
defined among others based on the number of direct
contacts (degree centrality) or based on the paths be-
tween other actors that go through the actor (betweenness
centrality) [12,13] .
Degree centrality can be interpreted as a measure of
potential communication activity whereas betweenness is
interpreted as the opportunity to control communication
[13]. Betweenness centrality measures on how many
paths between other actors an actor lies and such poten-
tially controls their communication [11,12]. The property
of the ego-network can further by characterized by close-
ness centrality: This measure measures centrality in
terms of distance or closeness relative to the other actors
in the network [12].
2.3. Opinion Leaders
The concept of opinion leadership was described in soci-
ology first by Lazarsfeld et al. in his classical study on
opinion formation during presidential election campaign
in 1944 in Erie Countie [1]. Unexpectedly for the authors,
the study revealed, that mass media do not impact the
opinion of people directly but through opinion leaders [1].
The concept of opinion leadership was developed in the
framework of political convictions, but later applied
vastly to consumer behavior from a marketing perspec-
tive [14]. The opinion leadership attribute is linked to a
certain subject or area of interest, such as a product
category. Different persons can be opinion leaders for
different matters. In fact, King and Summers [15] found
that less than one third of almost 1000 respondents could
not be considered as opinion leader in one of 6 types of
product. However, opinion leaders for one product cate-
gory are likely to be opinion leaders for other, especially
adjacent categories, as well [15].
Feder and Savastano [16] have studied the role of
opinion leaders in the diffusion of innovations on the
example of pest control tools used by Indonesian farmers.
They conclude that opinion leaders facilitate the diffu-
sion of knowledge when they are only moderately more
successful (measured as socio-economic distance) than
followers.
2.4. Lead Users
A further classical study reveals the concept of lead users:
von Hippel [2] names such the group of product users,
who require certain features earlier than the mass market
and take advantage of innovations in this direction
strongly. Von Hippel [2] proposes to use lead users and
their specific requirements for market research purposes.
Today some companies establish relationships with lead
users to create product innovations [17].
Companies attempt to to use lead users to create and
identify relevant innovation [3]. Ideas generated by lead
users tend to be more commercially relevant, strategic
and innovative [4]. Lead users can be identified by using
a questionnaire; by screening customer databases or by
contacting product users and asking for referrals to other
users other needs [17]. Lead users tend to be more ex-
perienced and possess a higher level of knowledge in a
certain domain [18]. Schreier [18] propose to use know-
ledge and experience in combination with the individual
properties internal locus of control and innovativeness to
identify lead users. Spann [11] show how virtual stock
markets can be used to identify lead users for consumer
products.
Jeppesen and Laursen [19] find that lead users in an
online community do not only acquire new knowledge
outside the community but also import it into the com-
munity, this indicates that lead users might be especially
valuable acquaintances for other veterinarians.

K. SIMBECK 11
2.5. Association between the Network Roles and
Ego-Network Position
Without using the later developed concepts of lead users
or opinion leaders, Coleman [20] discuss the impact of
social ties to peers on innovation diffusion, namely the
adoption of a new drug among physicians in a city. They
find that those physicians that are friends with many oth-
ers, adopt the innovative drug earlier [20]. Kratzer and
Lettl [9] have shown, that lead users can be distinguished
by high betweenness centrality whereas opinion leaders
by high degree centrality. The link between opinion
leadership and network centrality is also confirmed by
Merwe [21]. Iyengar et al. [22] however find that opinion
leadership and central network position correlate but do
not completely overlap. They find that central network
position is actually a better measure of influencing lead-
ership as opinion leadership identified through sociomet-
ric questionnaires [22]. Cho et al. [23] model the impact
of network position on innovation diffusion using various
centrality concepts; however they implicitly assume that
lead users are always opinion leaders for the innovation.
In this specific application of the research question two
different sets of lead users and opinion leaders are ana-
lyzed: those for marketing and those for management
accounting questions. One agent with a given network
position can or cannot be a lead user/opinion leader in
both fields.
Often, a strong relationship between domain specific
opinion leadership (i.e. in a certain field) and general
opinion leadership is found: opinion leaders in one field
tend to consider themselves as opinion leaders in other
fields [18]. This leads to the first hypotheses.
H1 Network position is associated with opinion leader
role.
H1a The higher the degree centrality of a business
owner, the more likely the business owner can be identi-
fied as an opinion leader.
H2 Network position is associated with lead user role.
H2a The higher the betweenness centrality of a busi-
ness owner, the more likely the business owner can be
identified as a lead user.
2.6. Network Role and Performance
Interestingly spread of innovation is often considered a
similar process to learning. Literature describes both
opinion leaders and lead users as interested and knowl-
edgeable in their respective specialties. According to von
Hippel [2] lead users have a higher perceived need for
innovations which results in a higher perceived benefit
from adapting innovations. However, being a lead user
can also yield other benefits, such as peer recognition
[19]. Usually lead users and opinion leaders are assessed
by researchers with regards to their “usefulness” [17,23,
24,31]: How can they help to promote a product or a
healthcare treatment? What is the best approach to reach
lead users fast to establish a user base for an innovative
product? To our knowledge the outcome of opinion
leadership or lead user property on the subject is not
studies so far. The concept and definition of lead users
integrates that the first adopters of an innovation are
those that benefit most from it economically.
Opinion leaders are not only expected to share their
knowledge [25]. This could lead to the conclusion, that
opinion leaders and lead users are better in marketing or
management accounting activities than their colleagues
and consequently perform better with their businesses.
This leads to hypothesis H3:
H3: Opinion leadership and lead user property are
positively related with performance.
On the other hand, lead users are expected to try out a
greater variety of activities.
3. Description of Sample and Measures
A survey based empirical analysis of veterinarians in
Berlin is used to understand to what extend the hypothe-
ses are correct or need to be rejected. The survey was
sent out to all veterinarians practicing in Berlin from the
complete list of the Veterinary Chamber’s website,
downloaded in July 2010. The focus was on small animal
practices.
After removing veterinarians who do not practice
anymore or are specialized on horses or pathology and
consolidating those veterinarians practicing in a joint
practice/partnership, a base sample size of 324 practices
remains. The base universe consists of 283 single prac-
tices and 41 partnerships. In single practices, 138 veteri-
narians are male and 145 female. In the base universe 8
partnerships are male only, 17 female and 16 of mixed
gender (see Table 1).
Out of 121 valid answers received, 20 were from part-
nerships whereas 101 were single practices. The average
age of respondents was 50 years (approximated as 2010 -
year of birth). Practices were established on average
since 15 years (approximated as 2010 - year of estab-
lishment), 32 practices are young practices (<8 years of
establishment). Practices had on average 2.2 veterinary
assistants (including trainees) and employed 2.0 veteri-
narians including owner(s). Not all questionnaires were
returned 100% filled. Those with missing answers were
excluded pair wise for analysis.
There might be a non-response bias in the data from
practices with lower turnover because those practices
might be less interested in management questions. Table
2 shows the distribution of turnover in the sample and in
the base universe. There is no data available on turnover
for veterinary practices for Berlin. The sample data can
be compared to the German average turnover of small

K. SIMBECK 12
Table 1. Distribution in sample and base universe.
male % female % mixed % Total
Base Universe
Total 146 45.10% 162 50.0% 16 4.9% 324
Single practice 138 48.80% 145 51.2% 283
Partnership 8 19.50% 17 41.5% 16 39.0% 41
Sample
Total 56 46.30% 58 47.9% 7 5.8% 121
Single practice 50 49.50% 51 50.5% 101
Partnership 6 30.00% 7 35.0% 7 35.0% 20
Response Rate
Total 38% 36% 44% 37%
Single practice 36% 35% 36%
Partnership 75% 41% 44% 49%
Table 2. Turnover distribution in sample and base universe.
Sample (Berlin only) Small Animal Practices, GermanyDESTATIS (2009, p. 117)
Turnover in 1000 Eur Cases Grouped % Turnover in 1000 Eur Cases Grouped %
<40 11
40 - 80 22
80 - 125 24 57 53.3% <125 1567 1567 53.0%
125 - 250 31 31 29.0% 125 - 250 909 909 30.8%
>250 19 19 17.8% 250 - 500 398 479 16.2%
>500 81
Total cases 107 107 2955 2955
animal veterinary practices as collected by the German
Federal Statistical Office [26]. The distribution of the
turnover is surprisingly close to the distribution of turn-
over within veterinarians in Germany overall. This indi-
cates a good representation of practices in the sample. On
the other hand it needs to be expected that the turnover in
Berlin is comparably low [27]. 14 respondents did not
answer the question concerning their sales. The difficult
financial situation of veterinarians is comparable to the
one of other freelance professions: In his detailed study
of freelance engineering offices, Hommerich [28] finds
that about two thirds of the engineers had an annual
profit of less than 25,000 Eur, the average turnover per
person employed including owner was 62,000 Eur [28].
The average response rate was 37%, which is a very
satisfying response rate for this kind of survey.
3.1. Performance Measurement
In this study turnover will be used for measuring per-
formance. Turnover is used for three reasons: first, it is
easy to survey in the questionnaire. Second turnover is
less subject to tax optimizing efforts. Third it is assumed
that veterinary practices in Berlin have a similar cost
structure. Furthermore, veterinarians are expected to
share turnover more willingly with researcher than profit
figures. The turnover of small animal veterinarians con-
sists of several main components: on the one hand turn-
over from both curative and preventive veterinary ser-
vices, on the other hand turnover from the sale of prod-
ucts (drugs, pet food, dietary supplements). The sale of
products as opposed to services is linked to an own set of
managerial questions such as capital lockup in inventory,
optimization of order quantity, losses because of aging.
3.2. Measuring Network Position
In order to assess the network positions of the small busi-
nesses, every respondent was provided a list of all vete-
rinarians practicing in Berlin. Respondents were asked to
tick those of their peers, with whom they are acquainted.
The data was coded using 3-digit numbers. Further, small

K. SIMBECK 13
business owners were asked how many peers they are
acquainted with on a nationwide and potentially interna-
tional level, which can also be interpreted as a degree
centrality measure.
For every small business owner the properties of the
ego-network were calculated using UCInet [29] and en-
tered into the dataset for further statistical analysis,
namely degree centrality, in-degree centrality, out-degree
centrality, and betweenness. The complete network is
shown in Figure 1.
3.3. Measures for Opinion Leader and Lead User
Quality
Opinion leaders can be identified using the sociometric
approach (typical question: “To whom do you refer to
obtain information about...”), the key informant method
(by asking knowledgeable individuals) or through self-
designation (using a questionnaire) [30]. The sociometric
and the self-designation approaches are both widely used.
Interventions with the aim of improving medical practice
usually through opinion leaders usually use the socio-
metric approach [31]. Researchers aiming to use opinion
leaders to improve understanding of marketing processes
often turn to the King-Summers questionnaire [15,32].
According to Lazarsfeld et al. [1] opinion leaders will
answer the the following two questions positively:
 Have you lately consulted somebody in matters of ...?
 Has somebody asked you for advice lately in matters
of ...?
Fenton [33] have developed a questionnaire consisting
of six questions for opinion leadership that that is the
base for most questionnaires used today. Furthermore
Schenk [14] cites expertise and personal involvement as
further pre-requisites for opinion leadership. This is op-
erationalized in the questionnaire through the question “I
am earlier informed on new developments than others.”
The lead user concept was developed by von Hippel [2]
to describe those users of new products that have specific
needs earlier than typical users of the product category.
Figure 1. Complete Network: node size represents be-
tweenness; box shape represents high lead user score.
In this research however the focus is on new develop-
ments in management, namely marketing and manage-
ment accounting. Small business owners with a non-
management background face the challenge to compete
not only using their specialist knowledge, but also man-
agement processes, i.e. to gain new customers or to take
wise investment decisions. Within the set of businesses
competing in a certain market, some will be very ad-
vanced with regards to marketing and management ac-
counting processes while others will not care. By identi-
fying lead users of innovative management accounting
and marketing processes we will be able to test whether
those have specific network properties and whether their
innovative approach contributes to their business perfor-
mance.
Separate scales are built for the roles in marketing and
in management accounting: a person with a deep interest
in marketing might not have the same interest for man-
agement accounting. The opinion leader scale for this
study is built using the items “I recently consulted
somebody”, “Others ask my advice”, and “I am earlier
aware than others about new developments”. The lead
user scale uses the items “I am usually one of the first to
try out”, “I have done specific trainings”, “I like talking
about …”, and “I continuously read and learn about…”
Both scales were tested for scale reliability with ex-
cellent results: Cronbach’s Alpha was >0.8 for both opin-
ion leader and lead user scales for both fields of interest
marketing and management accounting. In comparison to
other studies, opinion leadership and lead userness cor-
relate relatively strong in this dataset: r = 0.732 for mar-
keting and r = 0.810 for management accounting. The
correlation between network role for marketing and
management accounting respectively is also high: r =
0.660 for opinion leader attribute and r = 0.707 for lead
user attribute. This indicates a very high overlap of lead
users and opinion leaders for marketing and management
accounting.
3.4. Control Variables
The average performance indicators are significantly
higher for male veterinarians. Consequently, gender of
the practice owner is used as a control variable. It takes a
certain time until a practice has found a sufficient num-
ber of clients and after a certain number of years of prac-
tice, the curve of customer base growth and management
learning flattens. The variable “years of establishment”
was approximated as the research year (2010) minus the
year of founding. The Variable “years of establishment”
is transformed into a dichotomous variable “young prac-
tice” which was set if the age of the practice is <8 years
and which is used as a control variable.
The concept of market orientation was described and
operationalized by Jaworski [35]; Kohli and Jaworski [36]

K. SIMBECK 14
and Narver and Slater [37] in the beginning of the 1990s.
The model brought forward by Kohli and Jaworski [36]
(abbreviated MARKOR) focuses on the elements of the
marketing process (intelligence generation, intelligence
dissemination and responsiveness. The authors predict a
strong link to company performance [35]. This has been
confirmed in many studies [34]. Therefore market orient-
tations shall be used as a control variable for the test of
the relationship between opinion leadership/lead userness
and performance.
The original scale to measure market orientation con-
tained 32 items and was proposed by them in 1993 [35]:
10 items to analyze marketing intelligence generation, 8
items for market intelligence dissemination, and 14 with
regards to the responsiveness (7 for response design and
7 for response implementation) [35]. While the ques-
tionnaire was shortened for the scope of this research, the
categories remained: marketing intelligence generation is
represented by marketing research (data collection), dis-
semination is represented by marketing research (data
analysis), response design is represented by marketing
strategy and response implementation is represented by
marketing strategy implementation. The scale was tested
successfully for reliability (Cronbach’s alpha 0.706).
4. Results
In line with prior studies [6,10], we find a relatively high
overlap between lead users and opinion leaders. This
agrees also with prior findings that lead users are not
only combining new insights from the outside with ex-
isting knowledge to create innovation but are also located
at the center of information sharing within their commu-
nity [19].
For correlation and regression analysis joint practices
and practices specializing in surgery, odontology or oph-
thalmology are excluded as they rely on recommenda-
tions from other veterinarians. Correlation analysis (Ta-
ble 3) shows, that both lead userness and opinion leader-
ship for marketing correlate significantly with in-degree
centrality. This is in line with the expectation from hy-
pothesis H1a that opinion leadership is linked to central
network position. Business owners that are well informed
about developments in management accounting and
marketing and like to share that knowledge tend to be
known by numerous peers. On the other hand, there is no
significant correlation for out-degree centrality. Out-
degree centrality measures how many peers a business
owner cites as contacts where as in-degree centrality re-
fers to the number of times somebody gets cited.
Citing many others as contacts is not linked to opinion
leadership nor to lead userness. The number of veterinary
contacts in Germany can be interpreted as a further de-
gree centrality measure looking at a national level, while
the other network figures measure the veterinary network
in the city of Berlin only. The number of veterinary con-
tacts in Germany is as expected also significantly and
Table 3. Correlation between network position and network role.
N 1 2 3 4 5 6 7 8 9 10 11 12 1314
1 Age 90 1
2 Gender 99 0.302** 1
3 Young Company
(<8 years) 99 −0.620** −0.407** 1
4 Market Orientation 83 0.0050.297** −0.0981
5 Degree Centrality 99 0.107−0.139 −0.1970.1271
6 Betweenness
Centrality 99 0.1920.021 −0.1740.2010.482** 1
7 # of peer contacts
in Germany 88 0.119−0.023 −0.1400.1650.479** 0.330** 1
8 Turnover 86 −0.142 0.415** −0.1400.456** 0.151 0.256*−0.0191
9 Opinion Leader
Marketing 95 0.0550.278** −0.1660.406** 0.114 0.290** 0.343** 0.308** 1
10 Lead User
Marketing 92 −0.038 0.214* −0.060 0.574** 0.166 0.367** 0.215 0.385** 0.656** 1
11 Opinion Leader
Controlling 95 0.0370.304** −0.1980.306** 0.0240.094 0.315** 0.280** 0.594** 0.533** 1
12 Lead User
Controlling 95 0.084 0.194 −0.156 0.550** 0.0600.116 0.266*0.293** 0.522** 0.674** 0.757** 1
13 In Degree
Centrality 99 0.1560.120 −0.0830.1350.481** 0.543** 0.417** 0.313** 0.254* 0.251* 0.124 0.1151
14 Out Degree
Centrality 99 0.113−0.144 −0.224*0.1670.744** 0.654**0.348** 0.1520.1200.198 0.047 0.0610.1851
*p < 0.05; **p < 0.01; ***p < 0.001; two-tailed.

K. SIMBECK 15
positively correlated with opinion leader role. Thus H1a
can be confirmed based on correlation analysis.
The results for lead userness are less clear. The corre-
lation between lead userness and betweeness correlates
positively and significantly for marketing, but not so for
management accounting. The assumption from hypothe-
sis H2a that lead users in management processes are lo-
cated between sub-groups in the peer network and thus
act as linking elements cannot be fully confirmed at this
stage.
A hierarchical linear regression analysis on standard-
ized independent variables is performed to validate those
findings. The significance levels of all regression coeffi-
cients are determined using bootstrapping.
In a first step linear regression with opinion leader
quality being the dependent variable is calculated for
control variables (see Table 4). The regression based on
age and gender explains only 4.9%/6.6% of variance in
opinion leader quality for marketing/management ac-
counting respectively. Introducing the ego network vari-
ables of degree centrality, betweenness and number of
veterinary contacts in Germany increases the explained
variance to 20.6%/16.3% for marketing/management
accounting respectively. The results are less clear then
the ones by Kratzer and Lettl [9] who can explain about
40% of the variance. Still, the regression co-efficients for
number of veterinary contacts in Germany which is also
a measure of degree centrality are significantly positive
in both models. As a conclusion H1 and H1a is con-
firmed. However there are indications that the number of
local contacts to peers has a negative effect. Interestingly,
opinion leaders can be identified in this dataset not so
much by the number of peers they know locally but more
so by the number of peers they know on a nationwide
level.
A similar hierarchical linear regression analysis is
performed to research the relationship between network
Table 4. Regression analysis opinion leader role.
Marketing Management Accounting
1. Regression with control variables
Age of Owner0.030 −0.046 −0.019 −0.045
Gender 0.260* 0.235* 0.304* 0.296*
corr. R² 0.049 0.066
Sig. 0.053 0.027
2. Regression with network variables
# of peer contacts within city −0.235* −0.164
Betweenness 0.298** 0.029
# of peer contacts in Germany 0.388* 0.404*
corr. R² 0.206 0.163
Sig. 0.001 0.003
*p < 0.05; **p < 0.01; ***p < 0.001; two-tailed.
position and lead user quality. Again, regression is cal-
culated for control variables first (see Table 5). The re-
gression based on age and gender explains only
3.5%/1.4% of variance in lead user quality for market-
ing/management accounting respectively. Introducing the
ego network variables of degree centrality, betweenness
centrality and number of veterinary contacts in Germany
increases the explained variance to 19.9%/8.0% for mar-
keting/management accounting respectively.
In comparison to the opinion leader regression model
the effect of betweenness is stronger and for marketing
opinion leadership also significant. The importance of
betweenness centrality and thus hypotheses H2a for lead
userness can only partially be confirmed through linear
regression (only for the model on opinion leadership for
marketing, not for management accounting). Hypothesis
H2 can be confirmed as the explained variance increases
significantly in both models. Figures 2 and 3 represent
the ego networks of 2 selected nodes with high/low lead
user scores.
In order to test hypothesis H3 correlation analysis is
performed between the performance measure turnover
and opinion leader/lead user factors. It turns out that both
opinion leadership and lead userness correlate signifi-
cantly with performance. On the other hand the correla-
tions for the control variables, namely market orientation
was strong as well. Consequently two-step hierarchical
regression analysis (Table 6) is applied in order to model
the effects of opinion leadership and lead userness on top
to the control variables. It becomes apparent that contrary
to the hypothesis neither opinion leadership nor lead
userness contributes to explaining performance. The
positive correlation between performance and opinion
leadership/lead userness is fully moderated by market
orientation.
Thus, neither lead users nor opinion leaders can trans-
late their assumed knowledge advantage into busines
Table 5. Regression analysis lead user role.
Marketing Management Accounting
1. Regression with control variables
Age of Owner−0.085 −0.188 0.064 0.033
Gender 0.258* 0.223* 0.168 0.164
corr. R² 0.035 0.014
Sig. 0.096 0.212
2. Regression with network variables
# of peer contacts within city−0.142 −0.110
Betweenness 0.435** 0.037
# of peer contacts in Germany0.195 0.347
corr. R² 0.199 0.080
Sig. 0.001 0.046
*p < 0.05; **p < 0.01; ***p < 0.001; two-tailed.

K. SIMBECK 16
Figure 2. Ego network of node 115 (high opinion leader/lead
user score).
Figure 3. Ego network of node 285 (low opinion leader/lead
user score).
Table 6. Regression Analysis Performance.
Turnover
1. Regression with control variables
Age of Owner −0.205* −0.192*
Gender 0.337** 0.319**
Market Orientation 0.361*** 0.378**
corr. R² 0.307
Sig. 0.000
2. Regression with network variables
Opinion Leader Marketing −0.041
Lead User Marketing 0.212
Opinion Leader Controlling 0.198
Lead User Controlling −0.310
corr. R² 0.303
Sig. 0.000
performance, however they have a higher probability of
being market oriented which itself relates to successful
performance. The concept of lead userness implies that
innovations are adapted earlier because of their needs. It
must be concluded that lead users adapt innovations ear-
liers not only for business reasons but also to meet other
needs, such as peer recognition or the joy of trying out
something new (see also [19]). Opinion leaders are keen
on sharing knowledge, which might not even benefit
themselves in terms of business performance.
5. Discussion and Conclusion
The objective of this research was twofold: on the one
hand we wanted to reconfirm the approach to identify
lead users and opinion leaders based on ego network
properties proposed by Kratzer and Lettl [9]. In order to
do so we replicated their experiment in a different loca-
tion with a different test group. We identified lead users
and opinion leaders in marketing and management ac-
counting among Berlin based veterinarians using a ques-
tionnaire. We also studied the network properties of the
veterinarians by asking them to which other veterinarians
they are acquainted.
The second objective of this research was to study the
link between opinion leadership and performance and
between lead user property and performance. We tested
whether the lead users and opinion leaders identified in
marketing and management accounting did perform bet-
ter in business. The results of this study show that this is
not the case. The higher involvement in marketing and
management accounting does not yield a measurable
business benefit to the small businesses studied in this
sample.
5.1. Implications for Theory
This research contributes to social network analysis, lead
user/opinion leadership research and entrepreneurship
research. With regards to social network analysis we
confirm that degree centrality can be used to identify
opinion leaders and potentially also lead users. Further
we find a difference between local and national networks.
Being acquainted with more distant peers increases the
likelihood of being an opinion leader. Further we con-
tribute to a better understanding of both antecedents (net-
work position) and consequences (business performance)
of lead user and opinion leader properties. Central net-
work position, i.e. being acquainted with peers is an an-
tecedent of opinion leadership. Being acquainted with
peers outside the local market and network seems to be
especially important. In contrast to the hypothesis, nei-
ther opinion leadership nor lead user attribute in man-
agement accounting or marketing are associated with
better business performance. In this respect this reseach
also adds to the complex field of success factor research
in entrepreneurship: a higher interest in management
topics does not seem to pay out for entrepreneurs.
5.2. Implications for Practice
From a network analysis point of view, we can confirm
that ego-network structure, especially degree centrality

K. SIMBECK 17
can help to identify high-involvement persons with a
higher probability of being lead users or opinion leaders
and thus improve targeting of marketing measures. Mar-
keteers that are trying to identify lead users and opinion
leaders based on network data should also consider the
difference between local acquaintances and more distant
acquaintances. According to the findings in this paper,
network members which are not only well connected on
a local level but also on a wider, say nationwide level,
are more likely to be opinion leaders or lead users. Since
opinion leaders or lead users in marketing and manage-
ment accounting do not show superior business per-
formance, scholars and consultants in management
should refrain from over-estimating the impact of their
science. Practioneers should focus on improving market
orientation much more then marketing or management
accounting skills.
5.3. Limitations and Directions
The explained variances in the models are relatively
small. With regards to the identification of opinion lead-
ers and lead users future research should try to find fur-
ther criteria which can be used to improve identify-
cation of lead users or opinion leaders besides ego- net-
work data.
This could be information that is available in elec-
tronic social networks such as hobbies or preferences or
information available in different databases such as
turnover with older product versions. Further, the data
does not differentiate well between lead users and opin-
ion leaders, which might be due to the fact that we have
not used product innovations but management processes
as subject.
With regards to the impact of lead user property and
opinion leadership in management on performance it
needs to be acknowledged that this may depend on size,
industry or complexity of the business. Also, innovative
new ventures might well benefit from technology lead
users in their management team while other service
businesses might benefit from opinion leaders in the spe-
cific service field in their sales team. The role of lead
users and opinion leaders for business performance thus
deserves further studies.
REFERENCES
[1] P. P. F. Lazarsfeld, B. Berelson and H. Gaudet, “Wahlen
und Wähler,” Luchterhand, Neuwied, 1969.
[2] E. von Hippel, “The Sources of Innovation (The Execu-
tive Bookshelf from the Sloan Management Review),”
Oxford University Press, Oxford, 1988.
[3] E. von Hippel, “Lead Users: A source of Novel Product
Concepts,” Management Science, Vol. 32, No. 7, 1986.
pp. 791-805. doi:10.1287/mnsc.32.7.791
[4] I. Eisenberg, “Lead-User Research for Breakthrough In-
novation,” Research Technology Management, Vol. 54,
No. 1, 2011. pp. 50-58.
[5] W. R. Darden and F. D. Reynolds, “Predicting opinion
leadership for men’s apparel fashions,” Journal of Market
Research, Vol. 9, No. 3, 1972, pp. 324-328.
doi:10.2307/3149547
[6] F. M. Belz and W. Baumbach, “Netnography as a Method
of Lead User Identification,” Creativity and Innovation
Management, Vol. 19, No. 3, 2010, pp. 304-313.
doi:10.1111/j.1467-8691.2010.00571.x
[7] V. Bilgram, A. Brem, K.-I. Voigt, “User-Centric Innova-
tions in New Product Development—Systematic Identi-
fication of Lead Users Harnessing Interactive and Col-
laborative Online-Tools,” International Journal of Inno-
vation Management, Vol. 12, No. 3, 2008, pp. 419-458.
doi:10.1142/S1363919608002096
[8] S. Hill, F. Provost and C. Vollinsky, “Network-Based
Marketing: Identifying Likely Adopters via Consumer
Networks,” Statistical Science, Vol. 21, No. 2, 2006, pp.
256-276. doi:10.1214/088342306000000222
[9] J. Kratzer and C. Lettl, “Distinctive Roles of Lead Users
and Opinion Leaders in the Social Networks of School-
children,” Journal of Consumer Research, Vol. 36, No. 4,
2009, pp. 646-659. doi:10.1086/599324
[10] P. D. Morrison, J. H. Roberts and D. F. Midgley, “Opin-
ion Leadership Amongst Leading Edge Users,” Austra-
lasian Marketing Journal, Vol. 8, No. 1, 2000, pp. 5-14.
doi:10.1016/S1441-3582(00)70181-9
[11] M. Spann, H. Ernst, B. Skiera and J. H. Soll, “Identifica-
tion of Lead Users for Consumer Products via Virtual
Stock Markets,” Journal of Product Innovation Manage-
ment, Vol. 26, No. 3, 2009, pp. 322-335.
doi:10.1111/j.1540-5885.2009.00661.x
[12] S. Wasserman and K. Faust, “Social Network Analysis:
Methods and Applications (Structural Analysis in the So-
cial Sciences),” 1st Edition, Cambridge University Press,
Cambridge, 1994. doi:10.1017/CBO9780511815478
[13] L. C. Freeman, “Centrality in Social Networks Concep-
tual Clarification,” Social Networks, Vol. 1, No. 3, 1978.
pp. 215-239. doi:10.1016/0378-8733(78)90021-7
[14] M. Schenk, “Finanz-Meinungsführer,” Spiegel Verlag,
2005.
[15] C. King and J. Summers, “Overlap of Opinion Leadership
Across Consumer Product Categories,” Journal of Mar-
keting Research, Vol. 7, No. 1, 1970, pp. 43-50.
doi:10.2307/3149505
[16] G. Feder and S. Savastano, “The Role of Opinion Leaders
in the Diffusion of New Knowledge: The Case of Inte-
grated Pest Management,” Policy Research Working Pa-
per Series, The World Bank, No. 3916, 2006.
[17] C. Luthje and C. Herstatt, “The Lead User Method: An
Outline of Empirical Findings and Issues for Future Re-
search,” R&D Management, Vol. 34, No. 5, 2004, pp.
553-568.doi:10.1111/j.1467-9310.2004.00362.x
[18] M. Schreier and R. Pruegl, “Extending Lead-User Theory:
Antecedents and Consequences of Consumers’ Lead
Userness,” The Journal of Product Innovation Manage-

K. SIMBECK 18
ment, Vol. 25, No. 4, 2008, pp. 331-346.
doi:10.1111/j.1540-5885.2008.00305.x
[19] L. B. Jeppesen and K. Laursen, “The Role of Lead Users
in Knowledge Sharing,” Research Policy, Vol. 38, No. 10,
2009, pp. 1582-1589. doi:10.1016/j.respol.2009.09.002
[20] J. Coleman, E. Katz and H. Menzel, “The Diffusion of an
Innovation among Physicians,” Sociometry, Vol. 20, No.
4, 1957, pp. 253-270. doi:10.2307/2785979
[21] R. Merwe and G. V. Van Der Heerden, “Finding and
Utilizing Opinion Leaders: Social Networks and the Power
of Relationships,” South African Journal of Business Ma-
nagement, Vol. 40, No. 3, 2009, pp. 65-77.
[22] R. Iyengar, C. van den Bulte, J. Eichert, B. West and T.
W. Valente, “How Social Networks and Opinion Leaders
Affect the Adoption of New Products,” GfK Marketing
Intelligence Review, Vol. 3, No. 1, 2011. pp. 16-26.
[23] Y. Cho, J. Hwang and D. Lee, “Identification of Effective
Opinion Leaders in the Diffusion of Technological Inno-
vation: A Social Network Approach,” Technological
Forecasting and Social Change, Vol. 79, No. 1, 2012. pp.
97-106. doi:10.1016/j.techfore.2011.06.003
[24] R. A. Coulter, L. F. Feick and L. L. Price, “Changing
Faces: Cosmetics Opinion Leadership among Women in
the New Hungary,” European Journal of Marketing, Vol.
36, No. 11-12, 2002, pp. 1287-1308.
doi:10.1108/03090560210445182
[25] W. Hoyer and J. Jacoby, “What if Opinion Leaders Didn’t
know more? A question of Nomological Validity,” Ad-
vances in Consumer Research, Vol. 8, 1955, pp. 299-303.
[26] DESTATIS, Kostenstruktur bei Arzt- und Zahnarztpraxen,
Praxen von psychologischen Psychotherapeuten Sowie
Tierarztpraxen, Fachserie 2. Statistisches Bundesamt.
2009.
[27] S. Bundesamt, “Einkommensteuer: Statistik 2004,” Deut-
sches Tierärzteblatt, Vol. 8, 2009, pp. 1036-1037.
[28] V. C. Hommerich, “Controlling Ist Zwingend Geboten,”
Deutsches IngenieurBlatt, No. 6, 2007, pp. 42-45.
[29] S. P. P. Borgatti, L. Everett and C. Freeman, “UCINET 6
for Windows: Software for Social Network Analysis,”
2002.
[30] R. E. Goldsmith and R. Desborde, “A Validity Study of a
Measure of Opinion Leadershipp,” Journal of Business
Research, Vol. 22, No. 1, 1991. pp. 11-19.
doi:10.1016/0148-2963(91)90010-U
[31] G. Flodgren, E. Parmelli, G. Doumit, M. Gattellari, M. A.
O’Brian, J. Grimshaw and M. Eccles, “Local Opinion
Leaders: Effects on Professional Practice and Health Care
Outcomes (Review),” The Cochrane Library, Vol. 8,
2011.
[32] L. R. Flynn, R. E. Goldsmith and J. K. Eastman, “The
King and Summers Opinion Leadership Scale: Revision
and Refinement,” Journal of Business Research, Vol. 31,
No. 1, 1994, pp. 55-64.
doi:10.1016/0148-2963(94)90046-9
[33] J. S. Fenton and T. R. Leggett, “A New Way to Find
Opinion Leaders,” Journal of Advertising Research, Vol.
11, No. 2, 1971, pp. 21-26.
[34] C. Rodriguez Cano, F. A. Carillat and F. Jaramillo, “A
Meta-Analysis of the Relationship between Market Ori-
entation and Business Performance: Evidence from Five
Continents,” International Journal of Research in Mar-
keting, Vol. 21, No. 2, 2004, pp. 179-200.
doi:10.1016/j.ijresmar.2003.07.001
[35] B. J. Jaworski and A. K. Kohli, “Market Orientation: An-
tecedents and Consequence,” Journal of Marketing, Vol.
57, No. 3, 1993, p. 53. doi:10.2307/1251854
[36] A. K. Kohli and B. J. Jaworski, “Market Orientation: The
Construct, Research Propositions, and Managerial Impli-
cations,” Journal of Marketing, Vol. 54, No. 2, 1990, pp.
1-18. doi:10.2307/1251866
[37] J. C. Narver and S. F. Slater, “The Effect of a Market
Orientation on Business Profitability,” Journal of Mar-
keting, Vol. 54, No. 4, 1990, pp. 20-35.
doi:10.2307/1251757