Dr. Stephan Melzer talks about the risks of hegemonical platforms, semantic data models
as an instrument of power and the vital meaning of competition within ecosystems
Knowledge & Power
In ecosystems, knowledge is accessible and is used customer-oriented in manifold ways. That is the ideal image. How different is reality?
When we talk about data, information and knowledge in ecosystems, the context plays a central role. Because we can no longer assume a global context in which this triad is valid. Our environment is increasingly complex and virtual. It is organizing itself more in ecosystems that coin the context for the interpretation of knowledge. Therefore, the big question is who dominates the ecosystem – its rules and processes and its context. This force is more important than the original force over data, which means ownership, itself. Because, when I set the range of validity, I also own the interpretive sovereignty. I decide which data is true, valuable and usable and what fake news are, for example.
How does this interpretive sovereignty manifest itself in technological and economic ecosystems?
The first question to ask is who defines the semantic model. Amazon, for example, defines the semantic model of its platform and determines what an IoT date or a customer profile is and how to create it. And this in turn determines which information can be extracted from this and how this information can be interpreted. In an asymmetrical ecosystem, which is dominated by one player, this processes is simpler than in a symmetrical ecosystem such as GAIA-X. Here, one must tediously agree on a semantic model to enable the comparability of decentral data.
Are ecosystems more efficient and stronger when they have a hegemonic structure?
Of course, asymmetrical system are faster and also more efficient in the use of knowledge and when scaling business models. At least in the medium term. But they also produce uniformity of thinking and knowledge and this monoculture will eat itself up at one stage. Today, we are still pretty far away from this stage. Large, asymmetrical ecosystems still find enough resources to not suffocate intellectually but also politically. This also applies to a model that we can see in China, for example. When the government ultimately defines the data model, it is also very efficient. And even here, we are not yet at the point where the system is failing on its own.
Symmetrical ecosystems have greater start-up difficulties; they take longer to establish efficient processes. But they have a model, which can adapt much better to the environment, and that is permeable and changeable. Which is because they don't have to put up protective walls around them – the model's adjustments come naturally to them, so to speak.
An interview with
Dr. Stephan Melzer
More topics
"Ecosystems thrive on scaling as quickly as possible. That's why financial strength plays a crucial role in the early stages: The critical mass, the market share must be brought."
Scaling & dependencies
By start-up difficulties, do you mean the hockey stick effect?
Exactly. It's all about scaling as quickly as possible, because that's what ecosystems thrive on. That's why financial strength plays a crucial role in the early stages: The critical mass, the market share must be brought. How do most of the asymmetrical ecosystems work? Initially, you offer services free of charge, attract customers and partners to your own platform, you let them profit from the side effects and ensure that they become more and more structurally intertwined with this platform. Then, you gradually revert the process by monetizing services or by charging for better services whilst functionally drying up the cost-free solution.
This is how convenience turns into dependency, creating high barriers to exit. These dependencies are problematic in the long term; they are not sustainable. And that is also the reason why I’m convinced that we need strong symmetric ecosystems such as GAIA-X in Europe.
But can ecosystems be thought of only in these two extreme forms? Or are there ways to reduce dependencies without having to give up the advantages of asymmetric models?
In my view, the answer is quite simple. When you enter an asymmetrical system as a customer, you build up a very comprehensive data history over time. For example, health data that arises around a medical device and, related to this, around the optimization of therapy in collaboration with the attending physician, the organization of everyday life, contact with a community, the manufacturer, and so on. If you want – or need – to switch to a device from another manufacture, the question of the data ownership arises. And this data, typically, does not belong to the patient. You cannot take this data with you – and that is the key to the problem. As long as we do not connect the right to the data to the user, we will continue to have the problem.
This brings us back to the issue of semantic.
Yes – if we all had something like a personal data container, in which our health data is stored and a universal data semantics, this would run counter to the dependency. We would have the intensity of competition that is always demanded. However, this is ultimately a question of the legal framework. At least in regulated areas, such as in the healthcare sector, efficient federal structures and thus also federal ecosystems, which are not dominated by one player, are possible.
"Decisive is who dominates the ecosystem – its rules and processes and its context. This force is more important than the ownership of the data itself."
Federalism & Modularization
So far, one does not see much of such federal ecosystems.
But would there be a chance to create them? Yes. This would only require, for example, that the federal states relinquish the interpretive authority to the federal level and to harmonize their different systems and data models. The reasons why this does not happen are obvious. In the industry, for example, you can see that dealers do not share their knowledge of the customer with the OEM. Although it would make sense from the standpoint of the overall optimum. But it is also the question of interpretative sovereignty and the power within the value added chain. So it is always a question of who has the right to the data and whether the circulation of data, information and knowledge can be kept transparent and tracked.
However, there is definitely movement. In many corporations, silos are gradually being dismantled and fair approaches are being sought to integrate data with partners. It's also increasingly happening where there is a higher level of interest – for example, in the healthcare sector, in disaster response.
It is also important to remember that the willingness of users to make their data available in an asymmetric ecosystem without clear rules is decreasing worldwide. Similarly, legislators are becoming less willing to simply stand by as these business models scale, or to allow data to be used beyond its original purpose. In asymmetric systems today, enormous amounts of data are generated and the semantic model is clearly dominated. But the full potential for generating knowledge cannot be unleashed.
"If we do not connect the right to the data to the user, we will continue to have the problem of dependency."
By original purpose, you mean using data, which was collected as part of a preventive health program, to look at completely different contexts?
Yes. That's difficult socially, and that's difficult for business. Because causality and correlation are equated. One infers the future from the past and thus reproduces the past. And blind spots are not illuminated. In symmetrical ecosystems, the discussion of correlation and causality is broader, interpretive sovereignty is not monopolized and thus the process of interpretation and knowledge generation.
"The willingness of users to make their data available in an asymmetric ecosystem without clear rules is decreasing worldwide."
Doesn't a symmetrical and pluralistic ecosystem, as described by you, also mean that there is internal competition?
An ecosystem is much more vital and resilient the more open it is. Openness also means redundancy and competition. The question is how this competition will be fought and with what ambition. If the ambition is "the winner takes it all", if competition is organized destructively and unfairly, then competition ultimately destroys vitality
The goal must be that the ecosystem dominates its environment and not a single player within it. Another essential factor is modularization. This allows monolithic processes and structures to be cut in such a way that there are sufficient interfaces to which different partners can dock their services. And on the other hand, resources can be used jointly and flexibly. Be it data, IT platforms, or parts of a production or logistics chain that are not competitively differentiated. Finally, competition also helps to perceive external dynamics through a variety of perspectives and address them via different solution approaches. And that, in turn, is an essential prerequisite for scaling.
An interview with
Dr. Stephan Melzer
Contact us!
-
Core, periphery, vitality
Dr. Stephan Melzer about the design of vital ecosystems, what you can learn about their setup from Elon Musk and why fixed stars don’t shine brightly forever.
- 1
-
"Trying to control ecosystemsprevents their development."
Manuel Audi and Florian Weihe on minimum viable ecosystems, the central meaning of purposeand why technology is not a decisive factor for the success of ecosystems.
- 1
-
"Trust comes from transparency and a dialog on equal terms."
Manuel Audi and Florian Weihe explain why traditional networks cannot develop into open ecosystems, how decentralized structures promote innovation and why the discussion about data ownership is going in the wrong direction.
- 1