7.2.3    The enterprise model dissemination phase

The dissemination phase both resembles and differs from the development phase. Dissemination corresponds to the "outer loop" of figure 7.1.

Actual use of enterprise modeling

The actual use of enterprise modeling observed in the empirical studies has varied, but the most extensive use has been as a structuring device (O7).
Assertion 16 

An enterprise model intended for human sense-making and communication can be effectively used as a structuring device, introducing order in the domain that is modeled. 

By a structure is meant the arrangement of relationships that an actor perceive to exist between elements (e.g., as in a model structure discussed on page 145). By a structuring device is meant a phenomenon that divides a domain into elements and relationships.

Empirical evidence for assertion 16 can be found in all the four projects. In VPT, the observation that the elements of the enterprise model was used as a kind of pegs corresponds to the notion of structuring device. In PA30, the process and product aspects were used as a background for problems and opportunities (bombs and clouds). In Gazz, the use of models for holistic thinking can be considered structuring, as the model was used to focus on the relationships between elements. In TEK-S, observations from WG Gas on page 108 exemplifies the role of enterprise models as structuring devices (A23).

Activities in the dissemination phase

The activities in the dissemination phase concern distribution and feedback on models, and resemble the activities in the development phase.
Enterprise model dissemination versus enterprise model development
Enterprise model development and dissemination is not observed to be different in principle.
Assertion 17 

The enterprise model dissemination process involves the same activities and artifacts as the development process, except that the iteration cycle time is longer, the artifacts are more completed, and the need for more extensive explanations is higher. 

The activities in question are construction, presentation, interpretation and feedback (from assertion 6). By iteration cycle time is meant the time it takes between each presentation to the enterprise audience. More completed artifacts denote that selected intermediate artifacts are turned into exposed artifacts, and the explanation refers to the presentation that accompanies the enterprise model.

Empirical evidence for assertion 17 is found in TEK-S. Events associated with the dissemination process can be found in presentations from the WGs to the PG, from the PG to the customer, and from the PG/WG to users (at WS3). These events occurred less frequently than the presentations within the groups, the artifacts had been "polished" to a larger degree (although not finished, as was the case at WS3), and they were accompanied by more extensive explanations. Still, all the four activities were present: The project group constructed the model, presented it to the enterprise audience, the audience interpreted the model and the explanation, and provided feedback to be incorporated in the model.

From a theoretical point of view, dissemination (as for development) involves both perspective making and perspective taking in communities of knowing. The community developing models can be seen as contributing their perspective into a wider community of knowing that also includes the enterprise audience. Hence, their presentation is both an act of perspective taking (from the enterprise audience's point of view) and perspective making (as the models are modified due to the feedback from the audience).

Not foreseen needs for information
An issue related to the need for accompanying explanations to enterprise models is the need for human actors in the dissemination loop.
Assertion 18 

The work done in the development phase cannot be assumed to foresee all requests for information that occurs in the dissemination phase. Hence, there is a need for a "human in the loop". 

Empirical evidence that supports the need for human actors taking part in the dissemination process can be found in TEK-S. Feedback on presentations of enterprise models was often in the form of questions (like in WS3), asking for the rationale for an assessment or why a certain aspect was left out of the model. The strategy project handled the need for unforeseen information in various ways: They created a list of Frequently Asked Questions that were intended to provide answers to questions that frequently occurred in various forums where the strategy was presented. They were also available through various communication channels (email, telephone, face-to-face) as actors requested additional information on unclear issues.

An illustration of assertion 18 is that an enterprise model is not seen as a self-contained database representing corporate knowledge. Enterprise models are considered as a enablers of structured access to human actors in the enterprise and their knowledge.

Feedback on enterprise models
To receive feedback on enterprise models is an activity in both the development and the dissemination phase, but it was observed to be less frequent in the latter.
Assertion 19 

To receive feedback on enterprise models requires active stimulation of the enterprise audience in terms of explicitly stated expectations and availability of forums appropriate for communication. 

The assertion is based on observations from PA30 and TEK-S. In PA30, models were hung on walls together with markers, and plant workers were expected to modify the models as they passed by in the hallway. This did not happen. In TEK-S, the feedback on documents in the project database was scarce until after the final workshop, where actors understood that they had to state their (dis)agreement. At WS3, on the other hand, feedback on the presentations was ample, as the audience were both explicitly expected to respond and the forum (a meeting) was well suited.

There is a significant difference between what is potentially possible and what is actually done. Most actors in the Statoil company has more than enough to do already, and without an explicitly stated request for feedback or a high degree of self-interest in the enterprise model, feedback is not likely to be given.

Artifacts in the dissemination phase

The artifacts observed in the dissemination phase were improved versions of selected artifacts from the development phase. Hence, the artifacts played a role in both phases.
Exposed artifacts
Exposed artifacts are finished versions intended for dissemination (see page 145).
Assertion 20 

Intermediate artifacts that get disseminated without being developed into exposed artifacts may run the danger of being treated as exposed artifacts, even if the artifact is accompanied by an explicit statement that it is intermediate. 

By being treated as is meant criticized as, i.e., that the presenting actor may receive feedback as if the artifact was completed and not only a proposal.

The observations underlying the assertion were made in the strategy project: At WS3, two of the WG leaders commented on the perceived unfairness of the critique (page 118). In stage III, one of the SG actors refused to publish the preliminary assessments (page 123).

A problem following from the reluctance to present intermediate artifacts is that the enterprise audience is not involved in the process (and involvement was an explicitly stated objective in TEK-S, also as A5). From a theoretical point of view, lack of involvement implies that the perspective making process is not open to the entire community of knowing. Only a part of the community is contributing, being a community within the community. Not involving all actors may potentially lead to an inferior perspective being developed.

Boland and Tenkasi (1995:356) state that

"new perspectives need to be nurtured and given protection from strong demands for performance. Of necessity, they will not be able to compete with an established perspective in another community's way of knowing."
This is another way of saying that an artifact that is to be exposed has to be well considered in order to resist attacks from an enterprise audience that has not been taking part in the construction of the artifact.
Enterprise models and explanations
Explanations and enterprise models can be seen as complementary artifacts.
Assertion 21 

When an enterprise model is disseminated, the explanation accompanying the model may constitute most of the model's content. 

The distinction between model and explanation was discussed on page 137, and the presentations of enterprise models (A19) reported in the strategy project represent empirical evidence for assertion 21.

The assertion is also compliant with assertion 16, stating that models were used as structuring devices. In a presentation, the enterprise model has been observed used to structure the oral presentation, as in WS3. When used as a structuring device, the model provides a context or a background to interpret the presentation against.

Metaphors and enterprise modeling
In the strategy project, metaphors were observed to play a role in both the development and the dissemination phase.
Assertion 22 

A metaphor is a simple, yet effective means that may be used to make sense of an unknown concept in terms of a well-known concept.

By simple is meant that the concept of a metaphor is easy to understand. Being effective means that the understanding of the unknown concept is significantly improved.

Empirical evidence for assertion 22 was found in TEK-S, as discussed on page 137. The metaphors were simple (the actors did not seem to have any problems with stating and understanding metaphors) and effective (the right-hand sides were all well known, and actors seemed enlightened after introduction of the metaphor).

From a theoretical point of view, Lakoff and Johnson (1980:4) argue that metaphors are fundamental to the way humans perceive, think and communicate. They claim that the use of metaphors is a common way of making sense of observations and thus support assertion 22. A more theoretically oriented discussion of metaphors is found in section 9.2.10.

Storytelling and enterprise modeling
Storytelling has been observed as an activity in both development and dissemination of enterprise models.
Assertion 23 

A story (narrative) can be used in conjunction with an enterprise model to concretize and extend parts of an otherwise abstract model and thereby make the model easier to understand both in the development and dissemination phase. 

Empirical evidence supporting assertion 23 can be found in TEK-S. Storytelling (A22) was performed by the corporate staff representative in WS1, providing an account of the growth of Statoil, and as a means to extract the essence of the strategy at WS2. The stories (labeled examples in figure 6.9) that were included in the Web version of the strategy exemplified how Statoil could benefit from a strong technological position.

The existence and importance of narratives in workplace practice have been stressed and discussed by a number of researchers. Brown and Duguid (1991:45) point out two important aspects of creating and exchanging stories:

Weick (1995:60) is also a strong proponent for the role of stories in sense-making (italics added):
"If accuracy is nice but not necessary in sensemaking, then what is necessary? The answer is, something that preserves plausibility and coherence, something that is reasonable and memorable, something that embodies past experience and expectations, something that resonates with other people, something that can be constructed retrospectively, something that captures both feeling and thought, something that allows for embellishment to fit current oddities, something that is fun to construct. In short, what is necessary in sensemaking is a good story."
When enterprise modeling is dominated by sense-making, enterprise modeling may benefit from the construction and use of stories.
Medium for enterprise models
Enterprise models and subsidiary artifacts that are disseminated in the organization have to be manifest in a medium. The properties of this medium influence the dissemination process.
Assertion 24 

When choosing medium for enterprise model dissemination, the characteristics of available media must be compared with the requirements and objectives of the enterprise modeling effort. Without awareness of this potential pitfall, experiences and expectations may be transferred from a known medium with fundamentally different characteristics. A consequence is that the chosen medium is not fully exploited. 

The assertion is based on the use of media for enterprise models in TEK-S. Initially, the primary medium was World Wide Web on an Intranet. The primary medium of the final deliverable was paper. Web and paper have fundamentally different properties and possibilities when it comes to representing enterprise models, but this was not exploited. The Web version of the strategy document (including the enterprise models) was explicitly requested to be as similar as possible to the paper version. An explicit requirement was that printing of the document on paper had to be easy. Hence, Web as a medium was not applied on its own premises.

Looking at again, both the development and dissemination phases rely on communication. Selecting a medium for dissemination that are inefficient for communication purposes obviously reduce the quality of model dissemination.

Support for assertion 24 can also be found in (Orlikowski and Gash, 1994:191) in their discussion of introduction of information technology in organizations:

"…people tend to approach the new in terms of the old. The same may be expected of people confronting new technology."
If paper is the old, well-known medium and Web is the new, unknown medium, Web is approached with expectations more appropriate for paper.

Enterprise audience

The two final assertions concern the competencies and the attention of the enterprise audience.
Competence to access exposed artifacts
There have to be a fit between the competence of the enterprise audience and the media and forums used for dissemination of exposed artifacts.
Assertion 25 

The amount of feedback on the models may suffer when the enterprise audience have little competence in the use of forums and media for dissemination.

Empirical evidence for assertion 25 is again found in the strategy project. In WS3, the feedback was extensive (the forum was a meeting, and all actors mastered the meeting as a forum for communicating). When the same people were explicitly asked to access the Lotus Notes database and provide comments upon the preliminary strategy, they did so, but the feedback was sent by email to the database moderator and not inserted directly into the database. Still, people mastered Lotus Notes to the degree that they got access to the database. Finally, the Web version was introduced and a number of actors were asked to provide feedback. However, very little feedback was received. An alternative interpretation is that the number of actors with access to a Web browser was limited, but the amount of feedback was still disappointingly small.

Attention to make sense of the artifacts
The attention of an actor is a limited resource, as pointed out on page 146.
Assertion 26 

If the amount of attention required to make sense of disseminated artifacts becomes too large, the enterprise audience may choose to neglect the artifacts and consequently not gain the desired understanding.

The assertion may seem obvious, but is still vital to the successful dissemination of models and subsidiary artifacts. Empirical evidence can be found in the strategy project, where one of the WG leaders complained that the actors in his group did not take the time to read all the information that was handed out at WS1 or produced in the Lotus Notes database (A7). They did not have the time to read all of it in addition to producing their own notes. Hence, they were not able to discuss the contents of some notes in a meeting. The WG leader had to read through the note first and explain the contents.

From a theoretical point of view, neglecting artifacts implies that the communication process within the community of knowing becomes less effective, leading to a less complexified perspective. Communication is at the heart of perspective making.