By Frantz Frebault – Square
Increasingly uncertain environments require organizational flexibility to enable innovation. But these organizations themselves need to stabilize their functioning to maintain their efficiency. If digital tools have always been a lever to control processes, they must also become scalable so as not to ossify business models.
Organizational ambidexterity: the paradox of control and innovation.
For Piaget, equilibrium is a dynamic that simultaneously pursues “aspects of functional continuity and structural discontinuity” (Maurice & Montangero, 1992). Because companies are subject to two contradictory challenges, that of controlling their operations to be efficient and that of adapting to increasingly changing environments, they find themselves torn between blocking their operational process and exploring innovative solutions. (March 1991).
These two issues push them towards opposing a priori organizational models. The first Taylor type is thought of as “mechanical” while the second, flexible and evolutionary, is “organic” in essence (Burns & Stalker, 1961). Choosing between return and risk, between efficiency and innovation, in an economy that is both competitive and changing, is a difficult choice that leads to the search for a form of organization capable of both plans: an ambidextrous organization (Duncan, 1976).
Digital transformation: the sustainability of control
The development of information technologies reflects this dual objective, as the budgets allocated to digital technology were and continue to be important both in their R&D applications and as a process control tool.
In the 1990s, given its innovative strength, many people saw in Toyotism the sign of a radical change in management systems by decentralizing decision-making power to the level of self-coordinated teams. However, the Japanese model was based on a balance that compensated workers’ autonomy for a cultural control typical of Japanese capitalism and impossible to duplicate in the West: Shikon Shyosai (samurai soul and merchant talent) introduced by Eiichi Shibusawa (1840-1931).
Western companies then began a transversalization of tasks, replacing this cultural lever with the computerization of processes, a form of “computer-assisted neo-Taylorism” that paroxysizes panoptic control (Coutrot, 1998), (Zuboff, 1988), ( Sewell, 1998 ), (Issac & Leclercq-Vandelannoitte, 2013)
The creation of CIOs, centralizing skills and resources in double bureaucratic structures, then completed the impossibility of actors in the field to try the slightest innovation by themselves, leading to an sometimes problematic “ossification” of organizations (Reix, 199).
Decompartmentalize teams and modularize architecture
The generally observed design is therefore a silo between the teams in charge of exploration and the teams in charge of exploration. This form of sequential ambidexterity, which results in a binary approach around the ” Ramp up ” and ” Run it is relevant only in a few limited cases (O’Reilly & Tushman, 2008). Today, we need to move towards a customer-driven global approach of continuous innovation.
Functional silos do not allow for the alignment of cultures or priorities and distance actors from customer concerns. Instead of matrix forms, divisional or holocratic structures that cover well-identified parts of the client’s needs should be preferred.
These units must include the agile and multidisciplinary teams they need and have full decision and action powers. So this means, by extension, having our own infrastructure and development environment to quickly design, deploy, and decommission.
Segmenting computer systems is an idea that goes against the grain, but here it takes on its full meaning. This is explained by Erwan Vezin, vice president responsible for enterprise architecture at AccorHotels in an interview with “LeMagIT” in 2017.. According to him, it is this type of project that allows us to better meet the functional needs of each profession and avoid the “Mikado effect”, which is the impossibility of altering a very fragile structure because it is intertwined in a very complex way.
At the same time, to maximize efficiency, it will be necessary to consider the level of stability expected of each component at each stage of the value chain. Isolate data fields and functionality that are expected to be static over time and across “business use cases” from those that are expected to be dynamic because they respond to uncertainty.
This approach will lead to the design of a main chassis or “operational backbone” that will support the platform of innovative solutions made available. (Ross, Beath and Mocker, 2019)