For solving problems related to business/IT-alignment we propose a model called the Business Behavior Model. The main idea behind the model is to capture the motives that drive an agent to take decisions about what resources he should exchange in a business collaboration. The model draws from the rational agent theory, the resource-based view, the business model ontology, and causal graphs. The usefulness of the model is illustrated through a small case study. The result indicates that the business behavior model is interesting and useful as a complement to goal models and value models.
A major problem in the area of business/IT-alignment is to ensure that the information
technology available to an organization provides the support the organization needs.
One demand on the support is that it should be adapted as the organization adapts to
changing conditions [
In this paper we look further into the link between goal models and value models. We
argue that the information contained in both those models can be complemented in
order to give a more complete view of the link. Limiting our analysis to some
wellknown goal models (BMM [
The rest of the paper is structured as follows; in section 2 we overview theories that the BBM draws from. In section 3 we define, develop, and explain the BBM. We also discuss how it can be used. Section 4 contains an illustrative example of its use in the form of a case study. An analysis of the case study is done in section 5. Section 6 ends the paper with a concluding discussion and directions for future research. 2
The Rational Agent Theory. In this paper we assume that the agents being modeled
are rational. The Rational Agent [
The Business Model Ontology The Business Model Ontology (BMO) [
We learn from BMO that the resources handled by an agent have properties (reflecting four different aspects of the organization) and depending on from which aspect the organization is analyzed those properties become more or less relevant. The Causal graph. A Causal graph is a set of nodes and arcs. The Causal graph was chosen as the syntactical basis for the BBM as it is well-founded and contains the concepts we needed for BBM development structured in a coherent way. Table 1 overviews the basic concepts of the Causal Graph.
Informational: The out-node is considered before the in-node is analyzed.
Causal: The in-node has conditional probability to take a certain value considering a previous out-node.
Definitional: The in-node is composed of the all nodes linked to it.
Related models. For this research, some models from strategic and business layers
are used as comparison basis. For the strategic layer: i* and BMM and for the
business layer: e³value. i* is a goal and agent oriented framework developed to model
the goals of an agent or organization. The main idea of i* is to model an agents
intentions, i.e. its goals, beliefs, abilities, or commitments [
The definition of the BBM is based on three concepts that come directly from the Rational Agent theory and the Resource Based View – decision, resource and motivation. Those concepts are not independent and are therefore linked through causal relation with a value that indicates the intensity of the link (table 2). Definition: “The Business Behavior Model is a model which describes the impact of the participation of agents in a business by integrating their resources in a causal graph. The participation is realized through decisions and driven by motivations.”
described by a set of properties. One economic resource is present in the actor model if the actor rents or owns the resource. Decision nodes represent identification of (alternative chains of) goals and means in order to reach an objective The information from the out-node decision is available at the time the in-node decision is taken. Similar to a temporal meaning. Out-node has an impact on the value assigned to the in-node depending on the value link. The connected nodes are decision nodes. The purpose is to improve the definition of a decision by using sub-decisions (which are more detailed).
Creation link are used in order to trace the reason why a resources is analyzed. The reason is linked to a specific decision. XOR-relation
Strongly positive Positive Negative Strongly negative
connector Double bounded connector ++ + -
Those connectors act as constraint on the nodes attached to the links; at least one outnode have to be considered to grant the consideration of the in-node but not all of them.
A connector between links of same type. Those connectors act as constraint on the nodes attached to the links; all out-nodes have to be considered to grant the consideration of the in-node.
To complement the meta-model of figure 1 we use an additional methodological tool
which we call a categorization. The main point of categorizing a resource is to
emphasize from what aspect a resource‟s property is important for a particular
analysis. This tool is inspired by the Balanced Score Card (BSC) [18] approach. The
categories we use, however, come from the four pillars of BMO as those pillars are
more adapted to the RBV. Figure 4 shows this categorization of the MMOG resource
(rectangle box). Note that we do not prescribe that all resources should be subject to
categorization at all times; this is determined by the modeling purpose.
Motivation for syntax and semantics. A rational agent has beliefs, desires, and
intentions. It chooses from a set of available actions and performs one in order to
reach an optimal outcome. Therefore the model is structured according the following
pattern (figure 2): actors have motivations (desires-outcome) that are fulfilled by
actions and supported by decisions (Intention-actions) in the presence of
environmental constraints (belief). In order to integrate RBV, actions are led on
resources that are changed and exchanged through agent activities. Furthermore, to
provide a deepest view of resource, the model analyzed them through their properties
as proposed by Petit [
As we focus on the alignment of models on goal and business layers we use i* [
As shown in figure 1 the BBM model is based on three notions; decision, actor, and resource. Moreover, the model includes the notion of property and captures different kinds of relations. Motivation is, as said, a derived notion in the model. Decisions. As Means-End links in i* are an envisaged solution for the accomplishment of a goal, they are translatable in term of valuable decision. Indeed, a solution to fulfill a goal has to lead to a decision in the business process or otherwise the goal will not be achieved. Decision that implies actions toward another actor also emphasizes the Dependency Link between actors in i*. From a different perspective, a decision is taken as it generates valuable improvement for the motivation and Value Activities are themselves generating value. Therefore, a Decision can be transformed in a Value activity, but not all the value activities are related to a decision. Start stimulus are also interesting as they emphasize the initial need of the participation, therefore they are providing information on feasible initial decision. Considering the Business Motivation Model, decisions that appeal to factual means can be translated in terms of „Means‟.
Non-economic resources. The End node with a Vision semantic [
Economic resources. The Resources from i* are economic resources for BBM and their exchanges between actors in i* are modeled by causal links that cross economic resources: an exchange implies modification on properties of the resource (decrease of a resource to the profit of another). Resources are also present in the Value Model e³value within the Value Object.
Properties. Properties are related to the tasks, goals and soft goals of i* considering the fact that those elements are directed in the growth of aspects of a resource for the agent and therefore can provide indices on strategic properties. Properties are related to „Means‟, „Ends‟ or „Influencers‟ from BMM for the same reason. „Means‟ are usually related to low level properties at the opposite of „End‟ nodes. Influencers are external constraints that can be associated with properties from rented or purchased resources. Indeed, those resources possess properties that are not directly controllable by the actor. Table 3 emphasizes that it is possible to construct the Business Behavior Model on the basis of the other models or to construct (derive) those models on the basis of BBM. Constructing BBM on the basis of other models or the other way around results in models which are aligned on the same ideas – this reinforces the consistency among models and increases the alignment. For example, in the illustrative case in section 4, a BBM is constructed from an e³value model and subsequently a goal model (using the BMM notation) is constructed from the obtained BBM; the BBM bridges e³value and BMM. Another way of using BBM is to use it for simulation; the final objective is to optimize the motivation, hence the necessity to improve the related non-economic resource (attached to the motivation). When looking at the model, the improvement comes from Causal Links emerging from properties influenced by decisions (figure 1). Therefore, to optimize the motivation, the user has to optimize the improvement on the path through the non-economic resource by selecting the most efficient alternative decisions. By optimizing the improvement is meant comparing the value indicators on the causal links and selecting the one that provide the best end-effect. A simulation is also illustrated in the case study.
The following example is based on the case of a massive multimedia on-line game
(MMOG [
By analyzing figure 3 we can sort out the decisions (value activity) and resources (value exchange). The final model considers one resource for both the CD and the Game Access – the MMOG. Motivation (a non-economic resource) is not present in figure 3, but is derived from reasoning about why an actor participates in the business collaboration.
Economic resource
Property
Creation link
As no more information is available in the e³value, the modeler should start to furnish the model with properties and link them together. This information is present in the problem description, and the BBM in figure 4 has been complemented based on this information. The final model is obtained by adding a categorization for the MMOG resource.
Decision
Informational link variable by the considered agent) but also as variables whose values the decision makers can vary.
Figure 4 also illustrates simulation: the model proposes an alternative decision for the hosting resource which is to install a hosting service that would be owned by the game provider (shown in the lower right corner). When comparing Value Link on both out-relations from the two decisions: They cost the same (two double minus). For short term the renting is more advantageous, but in the long run owning is more advantageous. Quality varies (one is double minus and one is double plus). Renting provide the insurance of experience strengthen by contract. The installation requires experts that are maybe not present inside the companies; therefore, quality may be reduced.
In this case, the choice is quite easy; renting seems to be the better decision. Similar models are constructible for the customer‟s and the ISP‟s point of view.
Figure 5 shows a BMM built on the basis of the BBM of figure 4. We opted in this paper to show a BMM instead of an i* model as it is more compact. The used process did not consider the non-influencing decisions and focuses on the properties to sort them out in one of the following category of nodes: “End”, “Means” and “Influencer”. “End” nodes are top level properties in the graph of the BBM (figure 4). “Means” are low level properties (or leaves of the graph). “Influencer” nodes are external constraints. In our case the constraints come from the renting of the hosting service – the ISP is the one who fix the price and the capacity. The influencing decision (Transport CD) is also “means”. As it is visible in figure 5, the BMM conserves all the relation among nodes from figure 4. The negative causal relation between the “Game cost” and the “Quality of content” (in figure 4) is modeled through a dotted link in the figure below to avoid using Assessment elements from BMM. 5
This section highlights the value of two approaches of using BBM. The first consists in comparing the model in both the strategic level and in the business level with the BBM by pointing out some valuable added notions. The second focuses on improving the analysis for both the strategic and the business level.
Motivation oriented: Motivation is the engine that drives problem solving for a business. Therefore, modeling the engine of the participation is crucial to reach an optimal solution. Including motivation also brings the possibility of giving a nonprofit oriented view of the business by focusing on this non-economic resource. The motivation can be present in goal model, but in our case, the motivation is linked with actions on resources (figure 2).
Decision oriented: Decisions are the first step towards the achievement of a solution. Plus, alternative decisions provide the possibility of reflecting on which solutions are the best for the business through simulation. In terms of alignment, it bridges the establishment of goals with their application in the value proposition. Once again, some decisions could be drawn in the strategic layers but here we connect them with motivation and action on the resource (figure 2).
High resource granularity: Modeling resources as a set of properties gives insights about the weakest and the strongest points of resource configurations. The categorization of resource properties improves the structured view of the resource. Modeling inter-resource relations gives a wider and a sharper view of the studied system and detailed descriptions of property dependencies emphasizes different aspects of a system.
RBV (resource based view): The Business Behavior Model is not a tool to determine the strategic resources. However, it is a view of the internal and external mechanisms that involve those strategic resources. Indeed, the analysis of the properties and their impact on the global system gives a wider understanding of the engaged resources. As far as the BBM is connected to the RBV [19, 20] and gives interesting analysis of the resources management, it should be considered as a step towards obtaining a strategic advantage.
The use of the developed model improves the analysis of the business on several aspects thanks to the introduction of new aspects for the studied layers. Indeed, the business and the goal layers are focused on goal and value proposition. The BBM brings a new approach by the way of the motivation, the resources view and the possible decisions. The introduction of the decision concept allows analyzing whether or not the motivation is fulfilled by decisions and how. Decisions are also the basis of simulation for optimization through their alternativity. The developed model also emphasizes the weakest and strongest point of resources by pointing out their negative and positive impacts. The analyst gets a view on the mechanisms that are linked to the resource and therefore, he owns clues for further improvement of the organization (considering the RBV). The model also improves on the possibility to analyze interdependencies between resources as it shows those interdependencies at a sublevel (as relations between properties). 6
In this paper we have proposed a novel model, the Business Behavior Model, to be used when solving a part of the business/IT-alignment problem. The underlying idea of the model is to understand what are the motives that drive a collaborating agent to take decisions about resource exchanges.
The alignment problem is a complex issue that hits the organizations in their process of adaptation to the changing environment. In that context, this research aimed at achieving a support to improve the adaptation capacity. To do so, we had three goals; providing a clear and understandable definition of the developed model that we called the Business Behavior Model. The second was to define the context in which the BBM could be applied and used. The last one was to provide some clues on the usefulness of the model. This research has fulfilled the desired goals by the use of various theories, e.g. the causal graph, the resource-based view and BMO. The result of this research is a definition of the model and an indication of the usefulness of the model for solving the alignment problem. This is due to an analysis of the related model (BMM, i*, e³value) and the treatment of cases such as the MMOG. As shown, the BBM supports the bridging of two layers in an organization – the goal layer and the business layer. Through this research we also pointed out that the BBM could be used as an independent tool. It can emerge as a third kind of model next to the goal model and the value model with its own independent usage.
Future work: Improving the valuated causality relation among nodes by giving them
real values is the most relevant further work. Doing so opens the possibility of using
calculation on large and complex models that are based on the Markov theory [
Acknowledgements: We thank two anonymous reviewers for very valuable comments on an earlier version of this paper. 7