Playing the business ecosystem game
How Generative Technologies Demand Dynamic Control
Look at the standard playbook for launching a business ecosystem, and you will find a common assumption: the “keystone” firm must start with a crisp, definitive blueprint. Strategy textbooks tell you to map out user needs ex ante, design the technical interfaces, lock in your strategic bottleneck, and then orchestrate your partners along a roll-out plan to build momentum.
That approach works perfectly when user needs are well-known and downstream applications can be clearly imagined. But what happens when you are dealing with truly generative technologies, such as innovations deep within the digital or industrial infrastructure that can enable a virtually limitless, unpredictable range of future applications?
In a 2018 research paper published in the Academy of Management Journal and written with Professors Erkko Autio (Imperial) and Oliver Alexy (TU Munich), we proposed a process model to explain the creation of a de novo business ecosystem spawning from generative technologies.
While uncertainty is an epistemic challenge stemming from an agent’s limited access to existing knowledge, indeterminacy is fundamentally ontological. We are used to thinking that the future will unfold among a set of pre-existing options which we can map ex ante, like branches in a decision tree. This belongs to the realm of uncertainty: a purely epistemic challenge where, if we just had enough information, we could compute which branch will mostly occur. However, because of the generative nature of creation, we are facing an indeterminate reality whose very “options” cannot be known in advance. The French philosopher Henri Bergson, whose work I admire, famously argued that “the real creates the possible”. Because of human creative agency, the path can only be drawn retrospectively once it has already been “traveled”. This is not just an epistemic challenge, but an ontological constraint.
Hence, with truly generative technologies, you cannot blueprint a destination which you cannot even imagine. Trying to force a static vision too early leaves you vulnerable to a brutal, false-positive strategic trap: investing heavily to build a “fortress in the desert” which would be a beautifully designed set of control points that downstream players simply bypass as the market evolves.
To understand how pioneers successfully navigate this environment, managers must revise the linear ex ante model. Ecosystem creation is a highly organic, non-deterministic process driven by coupled feedback loops. Winning isn’t about executing a perfect plan; it requires a fundamental shift in how we understand strategic control.
The Architecture Paradox: Surrendering Ownership to Gain Traction
In a traditional, vertically integrated business model, control is simple: it is legally bound to asset ownership. But when you transition to an innovation ecosystem strategy around a generative technology, this classic command-and-control paradigm collapses.
The fundamental paradox of the ecosystem game is that to scale value creation, you must explicitly relinquish control over parts of your system architecture.
Because generative technologies enable an unbounded range of applications, no single firm has the resources or domain expertise to build every customer-facing variation alone. Trying to tightly restrict the architecture under a proprietary umbrella creates a fatal innovation bottleneck. To tap into the distributed creative energy of external complementors, the ecosystem orchestrator must systematically open up key layers of the technology stack.
From Static Ownership to Dynamic Control
Surrendering legal ownership of certain architectural layers does not mean accepting total chaos. Instead, strategic leverage shifts from static asset ownership to dynamic control.
Dynamic control is a continuous, closed-loop capability focused on steering collective action and maneuvering through highly uncertain feedback loops. Rather than resting on rigid barriers, it is defined by three operational levers:
The Power to Influence: Using non-coercive mechanisms—such as performative workshops, selective partner vetting, joint architectural review boards, and interoperability certifications—to subtly guide the rate and direction of independent external roadmaps toward a shared vision.
The Power to Monitor: Establishing a structural informational advantage by serving as a central, neutral broker. By constantly scanning external applications and exchanging bilateral roadmaps with complementors, you synthesize a macro view of the market landscape to detect macro technical trends before they manifest as market disruptions.
The Power to Update: Maintaining the organizational agility to iteratively pivot the overarching business case. When monitoring reveals that the ecosystem is drifting or that downstream control points are sliding away from you, you must swiftly redraw the boundaries of the blueprint or update your monetization strategy.
Ultimately, winning the ecosystem game involves ongoing maneuvering. It is an abductive search process where you manage the tension between leaving the architecture open enough to invite explosive external momentum, while retaining tight, strategic leverage over critical structural bottlenecks to protect long-term value appropriation.
Maneuvering in Poor Visibility
To ground these systems-dynamics principles, our research inductively tracked how technology pioneers navigate this very tension across four distinctive cases of de novo ecosystem creation from generative technologies. While our 2018 Academy of Management Journal study inductively tracks the granular details of these trajectories across multiple multi-year case studies, the high-level takeaway for today’s leaders is clear: when the future is indeterminate, you cannot play chess. You must learn to maneuver. Stop trying to build a fortress for a world that doesn’t exist yet. True strategic sovereignty belongs to those who don’t just predict the map, but who have the agility to steer the territory as it is being written. Are you clinging to an ex ante blueprint, or are you ready to master dynamic control?