In Need of Change: The Scientific Practice of Representing

Business Models

Position Paper

Thomas John

and Dennis Kundisch

Cooperative Computing & Communication Laboratory, University of Paderborn, Paderborn, Germany

University of Paderborn, Paderborn, Germany

Keywords: Business Model, Business Modelling, Business Model Representation, Business Model Visualization.

Abstract: How best to represent business model knowledge is an open research issue. A number of approaches for

graphically representing business models have been proposed. These are said to facilitate, for example, the

understanding, analysis and innovation of a business model. Nonetheless, through the selective literature

review we perform, we find these approaches are largely neglected in case study research on business

models. We argue that this practice needs to change to make case study research on business models more

consistent and comparable, and thereby strengthen the cumulative character of business model research

altogether.

1 INTRODUCTION

A business model can be defined as “a conceptual

tool that contains a set of elements and their

relationships and allows expressing the business

logic of a specific firm” (Osterwalder et al., 2005, p.

5). In recent years, interest in the business model

concept has surged; the number of academic and

journalistic articles has “virtually exploded” (Zott,

Amit and Massa, 2011, p.

1023). In this context, Al-

Debei and Avison (2010, p. 15) call for research on

“the differential influences among approaches of

representing the (.) [business model] knowledge

(oral, textual, graphical)”.

A number of authors argue in favor of using a

graphical representation (e.g., Gordijn and

Akkermans, 2003; Osterwalder, Pigneur and Tucci,

2005) for reasons that include the facilitated

understanding and analysis of business models. A

number of approaches for representing business

models have been proposed (Kundisch et al., 2012).

In addition, in their recent review of the business

model literature, Zott et al. (2011) dedicate a

separate section to business model representations

(BMRs) – which is another indication of the

importance these approaches have. However, it is

not known whether these approaches are actually

used in scientific practice. To shed some light on

this issue, we perform a selective literature review of

case study research on business models.

Our contribution is that we provide first

indication for the hypothesis that BMRs are largely

neglected by researchers who employ the case study

method to study business models. This is regrettable,

because BMRs provide a valuable means to

consistently and transparently document a business

model, and thereby may help to remedy one of the

major shortcomings of current business model

research: “that researchers frequently adopt

idiosyncratic definitions that fit the purposes of their

studies but that are difficult to reconcile with each

other” (Zott et al., 2011, p. 1021). We argue that this

practice needs to change in order to make case study

research on business models more consistent and

comparable, and thereby strengthen the cumulative

character of business model research altogether.

2 BACKGROUND:

REPRESENTING BUSINESS

MODELS

2.1 Existing Approaches

BMRs have been developed in fields as diverse as

287

John T. and Kundisch D..

In Need of Change: The Scientiﬁc Practice of Representing Business Models - Position Paper.

DOI: 10.5220/0004115902870290

In Proceedings of the 14th International Conference on Enterprise Information Systems (ICEIS-2012), pages 287-290

ISBN: 978-989-8565-12-9

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

strategy, e-business, and accounting. Some focus on

a specific domain (e.g., e-government), most BMRs,

however, intend to be applicable to business models

in general. BMRs can be classified according to their

reach, perspective, and notation principle (Kundisch

et al., 2012): Reach illustrates which of the three

theoretical layers strategy, business model, and

processes a BMR covers. Perspective states whether

a BMR offers a single view (i.e., one diagram) to

represent a business model or whether it offers

multiple views (i.e., multiple diagrams), which

complement one another. The notation principle

indicates whether a BMR employs a network-based

approach to represent a business model or whether it

employs a map-based approach. The network-based

approach refers to BMRs that have a fixed number

of concepts (e.g., actors, goals; each having a

different graphical representation), which are

outlined in a network to represent the business

model. Map-based approaches, in turn, lay out the

concepts one by one, thereby providing a template

which spatially structures a specific business

model’s key characteristics. For sample BMRs see

table 1.

Table 1: Classification criteria notation principle and

perspective, and respective sample representations.

Notation principle

Map-based Network-based

Perspective

Single

view

Business Model

Canvas

(Osterwalder et al.,

2010)

e3-value

(Gordijn and

Akkermans, 2003)

Multiple

views

Not available

Strategic Business

Model Ontology

(Samavi et al., 2009)

2.2 Reasons for Use

Authors advocating the use of BMRs argue that

BMRs facilitate the following tasks:

(1) Understand a business model and

communicate about it (Eriksson and Penker, 2000;

Gordijn and Akkermans, 2003; Osterwalder et al.,

2005): A visual representation is seen to facilitate

overall comprehension and to be less ambiguous

than (informal) natural language, thereby increasing

understanding and reducing the potential for

misunderstandings.

(2) Analyze and evaluate a business model

(Gordijn and Akkermans, 2003): Informally stated

value propositions – due to their lack of structure –

are inherently difficult to analyze (e.g. regarding

their potential profitability), which calls for the more

structured basis for analysis that a BMR provides.

(3) Deduce requirements for the underlying

information systems (Eriksson and Penker, 2000;

Gordijn and Akkermans, 2003): It is easier to deduce

requirements from a codified representation than

from natural language descriptions, hence, the

resulting information systems are better aligned with

the business model. Eriksson and Penker (2000)

point out that in the case of a business model being

used as the starting point for several information

systems, the risk that different development teams

interpret reality differently (and develop

incompatible systems) is reduced by using a BMR.

(4) Innovate a business model (Chesbrough,

2010; Eriksson and Penker, 2000): Building explicit

representations of a business model is the basis for

experimenting with new variations of that business

model and, thus, supports its innovation.

(5) Support business model design through

software-based tools (Osterwalder et al., 2005;

Samavi, Yu and Topaloglou, 2009): These authors

envision practitioners to benefit from software-based

tools for business model design, for example, for

comparing or simulating business models. These

tools in turn need rigorous representations.

3 ANALYSIS: THE SCIENTIFIC

PRACTICE OF

REPRESENTING BUSINESS

MODELS

3.1 Methodology

We approach the question of to what extent BMRs

have actually diffused into research practice by

means of a selective literature review. BMRs

support the work with a concrete business model,

and research on concrete business models is usually

performed using the case study method. Our

approach is to identify case study articles on

business models and to analyze the approach for

representing the cases that is used in these articles.

For identifying relevant contributions, we searched

for articles containing in their title business model

and case study (and respective plural forms). The

databases we initially employed for our search were

Scopus and Web of Science. Search with these

databases yielded some 40 results only, therefore,

we sought to extend the literature base. We did so

using the freely available Google Scholar database.

The data provided there are typically of lower

ICEIS2012-14thInternationalConferenceonEnterpriseInformationSystems

288

quality than those of commercial databases. Hence,

they require substantially more cleaning effort.

However, Google Scholar has the advantage of a

very broad coverage. It constitutes a reasonable

complement to commercial databases in bibliometric

studies (Aguillo, 2012), and, consequently, also for

the selective literature review that we perform.

We reviewed articles published until December

2011. For Scopus and Web of Science, this

restriction allows for reproducibility of the search

strategy. For Google Scholar, however,

reproducibility cannot be assured, because Google

Scholar adds articles continuously and the output

cannot be restricted with regards to the articles’ time

of addition.

We restricted the review to English-language

articles and cleaned the initial set of articles from

duplicates and erroneous results (e.g., articles whose

titles had been crawled incorrectly). For quality

reasons, we removed articles that had been

published outside journals and conference

proceedings.

Due to the broad coverage that especially Google

Scholar provides, the problem arose that some

publications were not accessible for us. However,

we consider this to be a minor problem for the

following reasons: First, the respective publications

are published in journals that have a very low impact

factor or none at all. Second, and more important,

there is no indication to assume that omitting these

articles introduces a specific bias into our set of

publications (for or against use of a BMR) – and

only then omitting these articles would be

problematic.

3.2 Results

Altogether, our search strategy yielded a set of 57

(accessible) articles, the Venn diagram in figure 1

illustrates their origin. By far the largest number of

results was provided by Google Scholar (218

results). The cleaning process reduced this sample to

55 papers. Scopus and Web of Science yielded a

substantially smaller number of articles (Scopus: 33,

Web of Science: 40). For details, see table 2.

We assigned the articles to one of the following

three classes:

(I) Articles that employ a BMR: Articles are

assigned to class (I) if they represent a business

model using a dedicated graphical representation.

Such a representation features notation elements

whose semantics are defined within or outside that

article with the aim to promote repeated application

of the representation.

(II) Articles that employ an ad-hoc notation: If

within an article the graphical notation is only

introduced for illustration purposes (without

substantiating design choices, not seeking repeated

application), that article is assigned to class (II).

(III) Articles that do not visualize (parts of) the

business model they analyze.

Figure 1: Origin of the final set of articles.

The results of the review are summarized in table 3.

Given space considerations, we present the detailed

analysis of the reviewed articles in an appendix that

is available upon request from the authors.

Surprisingly, despite the reasons for using BMRs

provided above, more than three quarters of the

reviewed articles do not employ a graphical

representation at all. They mostly describe the

business models in natural language, to some extent

making use of tables. Nine out of the 57 articles

devise some ad-hoc notation. There are only three

articles that employ dedicated approaches for

representing business models. The causal loop

diagram presented by Casadesus-Masanell and

Ricart (2010) is employed in two conference articles

(which also share one author). e3-value (Gordijn and

Akkermans, 2003) is employed in one journal

article. Unfortunately, the respective authors do not

justify their choice of representational approach.

Table 2: Results of the search and cleaning process.

Database

Total

number of

results

Number of results after removing articles that are…

Final set of

(unique)

articles

Duplicates

Crawling

errors

No conference/journal articles

& Non-English

& Not accessible

Google Scholar 218 209 192 82 77 55

57 Scopus 33 32 31 -

Web of Science 40 34 - 32

InNeedofChange:TheScientificPracticeofRepresentingBusinessModels-PositionPaper

289

Table 3: Summary of classification results.

Final set of

(unique) articles

Representation used?

Yes Ad-hoc No

57 3 9 45

4 DISCUSSION & CONCLUSION

It is an open research issue how business model

knowledge is best represented (Al-Debei and

Avison, 2010). There are numerous proponents of

graphical representations. However, reviewing

scientific practice reveals that only a minor share of

authors graphically represents the business models

they analyze – and only a negligible fraction

employs a dedicated approach for this graphical

representation.

Given the arguments provided in favor of using a

BMR, this seems surprising. However, the

arguments such as facilitated innovation or

deduction of requirements, in our view, mainly

apply to practitioners. Researchers rarely experiment

with a business model or develop supporting

information systems. Rather, they analyze real-world

cases to derive universal, transferable knowledge on

how to design a business model. For this purpose,

the given arguments mostly do not seem to apply.

Still, there is another set of arguments which

better addresses researchers’ needs. These

arguments concern the mentioned shortcoming of

business model research with regards to the

application of idiosyncratic, difficult to reconcile

definitions (Zott et al., 2011). BMRs could play a

vital role in mitigating this shortcoming. Through

their predefined sets of notation elements they force

a researcher into a predefined frame of reference.

The potential advantages include a better

comparability of findings, an easier reception of

findings by the research community, and a more

comprehensive analysis of business models. A

prerequisite is, however, that adequate BMRs are

available. Therefore, research effort should be

devoted to analyzing the hurdles that prevent

researchers from using the existing BMRs and, if

necessary, refining representational approaches so

that they find their way into research practice.

The results of the literature review reinforce our

confidence in the chosen methodological approach

of performing searches across multiple databases. It

turned out that the number of unique articles is

highly dependent on the chosen database, varying

between 31 (Scopus) and 55 (Google Scholar). The

large overlap among the databases (24 out of 57

articles are contained in all three databases)

increases the confidence concerning the relevance

and comprehensiveness of the considered articles.

Future research could broaden the literature base

to receive a more complete picture of BMR use in

research practice. The articles that present dedicated

approaches such as the Business Model Canvas

(Osterwalder et al., 2010) or e3-value (Gordijn and

Akkermans, 2003) have several hundreds of

citations, and it would be valuable to find out how

the citing authors use these works. Following this

approach, however, a bias in favor of using a BMR

should be acknowledged: the fraction of articles

employing a BMR is likely to be higher than in our

findings (this had initially been the reason for us for

not using a search strategy based on citations – to

provide an unbiased view on BMR usage). In

addition, complementing our review of scientific

practice, a worthwhile endeavor would be to survey

practitioners about the awareness and usage of

approaches for representing business models.

REFERENCES

Aguillo, I. (2012). Is Google Scholar useful for

bibliometrics? A webometric analysis. Scientometrics,

in press.

Al-Debei, M., Avison, D. (2010). Developing a unified

framework of the business model concept. European

Journal of Information Systems, 19, 359-376.

Casadesus-Masanell, R., Ricart, J. (2010): From strategy

to business models and onto tactics. Long Range

Planning, 43, 195-215.

Chesbrough, H. (2010). Business model innovation:

Opportunities and barriers. Long Range Planning, 43,

354-363.

Eriksson, H. and Penker, M. (2000). Business modeling

with UML. New York: Wiley.

Gordijn, J., Akkermans, H. (2003). Value-based

requirements engineering: Exploring innovative e-

commerce ideas. Requirements Engineering, 8, 114-

134.

Kundisch, D, John, T., Honnacker, J., Meier, C. (2012).

Approaches for business model representation: An

overview. Proceedings of the Multikonferenz

Wirtschaftsinformatik.

Osterwalder, A., Pigneur, Y., Tucci, C. (2005). Clarifying

business models: Origins, present, and future of the

concept. Communications of the AIS, 15, 2-40.

Osterwalder, A., Pigneur, Y. (2010). Business model

generation: A handbook for visionaries, game

changers, and challengers. New Jersey: Wiley.

Samavi, R., Yu, E., Topaloglou, T. (2009). Strategic

reasoning about business models: A conceptual

modeling approach. Information Systems and E-

Business Management, 7, 171-198.

Zott, C., Amit, R., Massa, L. (2011) The business model:

Recent developments and future research. Journal of

Management, 37, 1019-1042.

ICEIS2012-14thInternationalConferenceonEnterpriseInformationSystems

290