Chapter 7 - Management anomaly specification

In this chapter, we propose to synthesize the anomalies highlighted by the cases’ results of the previous chapter before listing the requirements to design a new model extending the literature models of non-mutual conditioning between exploration and exploitation, in order to understand the observed phenomena.

The first feature raised through the descriptors and the comparison between models’ prediction is perhaps the continuity and orthogonality of exploration/exploitation that is dealt at different levels: generative processes, coordination mechanisms, environment cognition and organization design. So, we confirm the nuances expressed in (Gupta, Smith, and Shalley 2006). What is perhaps more surprising and natural at the same time is the fact that exploration can be based on exploitation characteristics. However, the cases show these can be beneficial or counter-productive for innovation management, thus requiring an careful management to counter the forces of traditional models of ambidexterity.

The second aspect is that we can assert that ambidexterity kills innovation. It is of course a bold statement, but the meta-language of exploration/exploitation tends to ignore the micro-foundations of collective action such as decision-making and generative processes (Birkinshaw and Gupta 2013). The actual implementation of ambidexterity, be it contextual, structural or sequential, induces numerous complications dealing with exploration/exploitation regime separation for practice. More specifically, it appears to be incompatible with practices where the unknown is effectively managed.

The cases show they follow organizational ambidexterity, but at the same time, have mitigated efficiency when it comes to exploration or exploitation standalone. Radical innovation is not really achieved for the originally intended exploration project, despite having several other valuable takeaways (Elmquist and Le Masson 2009). There is tendency to leave exploitation routines as a static reference point, as a fixation effect, and let exploration free float in other directions without carefully managing the search and generation directions.

Organization design being constricted by certification and market constraints, generative processes and organization ambidexterity clashes to support this enclosed change calling for a support of organizational metabolism and potential regenerative dynamic capabilities (Ambrosini, Bowman, and Collier 2009). Despite appropriate interactionism, we have seen that practices are not sufficient to sustain generative processes into a full settlement organizations. At best, in the case of the icing conditions detection, the settlement is partial as the alignment with BU management and Engineering department remains controversial. Consequently, reconsidering organization design supported/supporting generative processes appears crucial, as it probably implies considering the mirroring of the engineering efforts.

Synthesis of anomalies

The three cases presented in the previous chapter have insisted on different limitations of the three proposed literature models of non-mutual conditioning between exploration and exploitation:

• Adaptive model: the search-based assumption of exploration is overstretched by generative processes of different nature and locus. Exploration is guided by uncertain or unknown (i.e. to be designed) exploitation constraints as cleverly managed in the icing conditions detection case. Whereas, it can also be a lock-in despite valuable generativity on the user experience domain, as several fixation effects of existing business units are not revisited as presented in the ADT’s cases.

• Interactive model: creation of meaning and value circulation among (silent) designers around meetings and prototypes are not sufficient to sustain radical innovation and the associate unknown gradual shaping. The business class seat platform shows the interference and conflicts arising from the exploration/exploitation handover and agenda differences. It would require a management of deeper learning mechanisms. Exploration and exploitation are unfortunately both jeopardized.

• Encapsulated model: As already stressed in the literature on project-based management, the need to link project trajectories with underlying organizational ties, resources and routines is even more critical at the light of exploration projects searching for exploitation landing fields. Learning and change management tend to be dissociated in exploration and exploitation regimes.

Generativity biased by exploration/exploitation dichotomy

First of all, it is important to note again that our research agenda set from the beginning was to use advances in design reasoning and theory to understand how generative processes and associated management of unknown allows revisiting innovation management and the major construct of ambidexterity. The cases presented earlier show how the models black-box the nature of generative processes and their interferences with several managerial dimensions such as: coordination mechanisms, environment cognition and organization design.

Angela Dumas with Henry Mintzberg (Dumas and Mintzberg 1989) had stressed the importance of silent design occurring across the organization. The cases reveal how these distributed design activities can be framed and segregated by what exploration and exploitation are in an ambidextrous organization.

In the cases of the business class seat and ADT, the generative processes had used clear exploitation characteristics to base their exploration: seat packaging, integration and certification issues, and user pains. It is used as a reference point, but at the same time when balancing out exploration/exploitation, the generative processes require an extreme adaption to exploitation organizational constraints despite having made a strong effort on the cognitive and interactive dimensions.

The exploitation is conducted as if a simple extended or new list of requirements had been given to the engineering departments. Risks can be assessed, as well as maturity, but these registers are unfortunately rather static. They may not capture and manage epistemic interdependencies accordingly (Puranam, Raveendran, and Knudsen 2012).

Exploration efforts are then potentially biased by the non-mutual conditioning. And inversely, the exploitation is heavily burdened by exploration features that lack articulation with generated novelties.

Killing sustained innovation capabilities

The nature of unknown, its management and generativity have been underestimated in how they collide with adaptive, interactive and encapsulated models. They stretch the boundaries of these management models and challenge their viewpoint of collective action. The radical innovation intent is severed by their limitations.

For instance, among the expected capabilities, the role of leadership, top management and governance is stressed as being crucial. However, the actual practice and implementation is rarely discussed in detail. The different cases showed how management layers took an active role in projects trajectories. Valuable practices were implemented ranging from sensemaking, decision-making to sponsorship. These were meant to balance the contradictions of exploration/exploitation trade-offs and the results are partially successful and riddled by controversies.

The icing condition detection case creates a critical tension between its business environment legitimacy and BU strategy and Engineering appropriation. Following the retirement of the R&T manager, the researcher had relayed his idea of licensing the technology to an external firm capable (adequate cost-structure) to develop and industrialize it for small aircrafts. This move would be a means to actually set a foot on the aviation market, make a case for it before fully demonstrating its usefulness for heavier aircrafts enhancing flight safety. Again, such emergent strategy would be possible by fully considering separation nor balance of both, as they explore and exploit interconnected fields.

Ignoring organization design and its engineering design mirror

Organizational ambidexterity had highlighted ways of structuring the balancing of exploration/exploitation adding different typologies on how to sustain a competitive advantage. Oddly, separating it in space, time or leaving it in context, brings some organization design into play but only by sketching the organic dimension of it. For example, the centralization of ADT at the group level or creation of a standalone R&T team are fully prescribed. It still leaves numerous ties among members with exploitation activities and extended environment constraints. However the black-boxing of generative practices into these organic systems ignores the risks associated with the potential need for reconfiguration of stakeholders ties. Thus, the dynamics induced by design practices require, not only a partial-mirroring by organizations, but rather a mirroring dynamic.

For instance, the required adjustments, in terms of engineering capabilities to engineer a seat platform architecture, require a mirroring organization (Colfer and Baldwin 2016). The change is not addressed within generative process and with integrators contributing to interdependence shift (Stan and Puranam 2017). The managers in charge of the projects were acting as integrators. In the case of the seat platform, they a priori managed to articulate the integration for the Engineering Department and were later called back for further support. The same goes for ADT: numerous sessions were organized during the exploration but also after concept selection to follow up and adapt the preliminary designs trying to facilitate the integration. As specified in the limitations section (Stan and Puranam 2017, 1058):

A significant weakness of this study is that it does not provide direct evidence for the micro-mechanisms of how integrators help to cope with interdependence shifts—only that they do (and that these effects are visible even at intermediate performance stages).

The idea of a triple-loop of organizational learning (Leifer and Steinert 2011; McClory, Read, and Labib 2017), and the regenerative capabilities point in that direction. But as they engage in a recursive pattern, as discussed in the literature review, it appears crucial in the case studies to frame the organization of collective action with practices that are tightly linked to generative processes. In a more successful way, the icing conditions detection case shows how future exploitation constraints (regulation and norms) are shaped to support the technology search and selection.

Specifying requirements for a new model of ambidexterity

Coming back to our research questions, we have stressed how organizational ambidexterity can severe radical innovation practices embodied by different generative processes. One could say that it is the fault of organizational culture, but we have pointed in ZA’s presentation, there are no signs of bad intentions raising barriers again these radical innovation in the cases presentation and results. Several practices and management tools were used to circulate emergent value among stakeholders, including collective decision and co-design workshops.

So, our first assumption would be to put the emphasis on the micro-foundations. The nuts and bolts of generative processes and interactions with value management, action and decision-making would be absolutely critical to embody ambidextrous management. It fits the behavioural foundations of March’s agenda (Gavetti, Levinthal, and Ocasio 2007) and it lies at the crossroads of design theory and reasoning which can be understood as science extending H.Simon’s work (Simon 1996; Hatchuel 2001). Considering everyone as a designer in its own way and practice, generating actions, interacting with other (silent) designers and artefacts, could be a means to think of generative processes among all of these organs and processes, which call for a meta-process such as metabolism (Hatchuel, Weil, and Le Masson 2006; Segrestin et al. 2017).

Generative process and ambidexterity micro-foundations

From the beginning of our literature review up to the cases analyses, we have discussed the course of action, decision-making and generative processes. We have justified how organization ambidexterity can kill sustained innovation management.

So, we have to come back to foundations of the exploration versus exploitation model whilst also keeping in mind the limitations and perspectives identified in the interactive and encapsulated models. The model’s extension we propose to address in the following part, should then almost start from a similar reasoning introduced in the foundational paper (March 1991). However, the dichotomy of exploration and exploitation should be fully revisited by considering a mutual conditioning of both regimes as it was sketched by the practices highlighted in the projects’ trajectories. The nature of problems, their (re-)formulation and solving should be reworked around the uncertainty and unknown management. Decision and problem management should be enhanced by generative processes. For instance, in (Le Glatin, Le Masson, and Weil 2017) (see publication, we associate the possibility to deliberately play around the reversal of preferences to push decision-making into the unknown. We have proposed to call such pattern decisional ambidexterity. Such practice should aim at identifying fixation effects and their (epistemic) interdependencies. This technology of organizing should support creation of meaning, value management, engineering and a baseline for coordination and collective action.

Generative processes and organization design

The free-floating of exploration projects, as they look for organizational ties, should also think through the lens of design theory and reasoning. The idea of design-oriented organization (Hatchuel, Weil, and Le Masson 2006) already pointed in that direction. We need to specify with more clarity the organization design, as we have seen that assumptions of temporary organization or project-based organization tend to hide the complexities of changing engineering design briefs in exploration project management.

Design theories and reasoning are usually discussed for products and services, but one could consider the organization, its resources and routines as concept-enabling-knowledge for design. So, the key requirement of the model is to avoid black-boxing generative processes, as we have seen they expand onto several management dimensions. We should then put the emphasis on potential organization design fixation that are not naturally part of the engineering design practice, nor traditional project management. These are usually discussed in change (project) management and organizational change but would probably benefit from richer synergies if embedded in actual project management (Pollack 2017; Hornstein 2015). Consequently, our model of mutual conditioning between exploration and exploitation should also aim at managing an organizational metabolism necessary to sustain organization re-design as a necessary condition supporting the innovation potential of attraction of - so far black-boxed - generative processes in exploration project management.

Chapter synthesis

–

References