Enterprise architecture is pointing the way forward for the digital transformation.
In the following interview, the Chairman of the Board of the Cross-Business-Architecture Lab, Dr. Johannes Helbig, talks about the rapid changes enterprise architecture is undergoing and the role it is playing in the digital transformation.
How does CBA Lab view enterprise architecture?
Enterprise architecture is currently redefining its role and developing further into what is basically a top management discipline. The digital transformation has made its way into the center of the economy and is high up on the agenda at large companies, such as those that are members of our association. The focus of enterprise architecture has changed as well. The idea today is not so much about how to come up with the craziest disruptive idea as a startup but instead to develop the management skills that enable the large-scale implementation of the transformation and the establishment of such implementation as a permanent task. This aspect will be decisive in terms of business competitiveness – and the extent to which this will be the case can already be seen in the recent development of the Facebook share price as compared to Amazon’s.
When I think about EAM, I always imagine some isolated office with floor-to-ceiling wallpaper adorned with structural diagrams, process plans, and similar things. Is that an accurate picture? How are enterprise architects integrated into business organizations, and how do they go about their work?
The way architects work is changing very much at the moment. Things are moving away from governance and toward enablement. The digital transformation is extremely challenging for companies, especially large ones. Software is in fact eating the world”, as they say. Half of the Fortune 500 companies from the year 2000 are no longer on the Fortune 500 list. It’s all about speed now. Digitalization is a polarizing force: It strengthens those organizations that are already strong and exposes the weaknesses of those that are weak. While it’s true that structural errors have always been costly to those who make them, the price one pays for them today is much higher. It’s a winner-take- allgame, and
in order to become faster, companies need to have the right structure, which also has to be a very flexible structure that increases an organization’s ability to change. That’s the only way large companies can become as explorative as startups and be able to enter new markets quickly, adapt rapidly to early customer feedback, and achieve short iteration cycles. Such companies need to be able to explore markets that they have no experience with – and in some cases markets which still have to be created. The new role of enterprise architecture is to create the structures necessary for all of this, up to and including the establishment of a “composable enterprise” that can very rapidly reorchestrate its capabilities in order to more quickly create the new products and services that the market is demanding. Amazon offers a good example of this. Jeff Bezos’ most important doctrine perhaps is that it should be possible to recompose and externalize all internal services as marketable products. Amazon Web Services was originally created in response to internal excess capacity, for example.
In the past, enterprise architecture focused on controlling risks through ever more precise planning. However, the decisive indicator of the future viability of a company is no longer risk avoidance or risk prevention but instead adaptability. The management of risks will be replaced by the management of uncertainty.
Many companies believe they can address this challenge using innovation labs. Do you think that’s the right approach, and how can any innovations developed be integrated into a company later on? Is this also an architecture issue?
It’s fundamentally an architecture issue in terms of both technical and organizational aspects. The most important thing is to make sure all activities are coherent. While it’s true that many companies now do good work when it comes to innovation, only few are able to link their innovations to their inherent core business. It’s a little like soccer – and we recently saw in the World Cup what happens when offense and defense completely fall apart. The key challenge is to establish consistency between offense and defense, in other words between innovation and the core business – and that’s an architecture issue.
The question is how a company can turn itself into a complex overall cyber-physical system. Virtually all large companies have abandoned horizontal Taylorization over the last few decades and have divisionalized themselves as they move toward continuous vertical responsibility. This also occurs on lower levels – for example in fully product-centric teams that work autonomously, even at the risk of redundancy. Employees are networked, just like their digital counterparts, by which I mean the software agents. Hierarchies no longer work here; they’re too slow. This is a big challenge because such extensive decentralization makes implementing an appropriate governance system a more demanding process.
What does “making sure all activities are coherent” mean for enterprise architects specifically? What do they need to do? Do they need to look for synergies, for reusable elements in order to ensure the same things aren’t invented twice? Should they be embedding new services in a business plan or business architecture?
Eliminating redundancies and ensuring stability are much less important here than agility, adaptability, and speed. The architect’s role is no longer that of a central gatekeeper. Instead, this role is shifting to that of a coach who offers advice and acts as a facilitator in production teams. Architects are no longer the guardians of long-term architecture plans who document the actual situation and serve as central monitors at quality gates and verify the compliance of the solutions used. Instead, their work supports teams and is geared toward generating added value and creating effective solutions. Digitalization is blurring the lines between IT and specialist and technical departments, as well as the boundaries between the various stages of the software development process, whereby I’m thinking particularly about DevOps here. For the first time ever, IT doesn’t run transversely across specialist and technical departments but instead operates in parallel with these.
The product structure is the primary structure. However, the verticalization of individual responsibility requires a higher degree of autonomy. Technically speaking, things in this regard are heading in the direction of microservices, which basically represent an expansion of our old SOA principle of loose coupling to include ever smaller units – which is why I sometimes also use the expression “micro-divisionalization --; It’s now possible to create the semantic coherence here in an automated manner using data grids. From an organizational perspective, additional matrix-like structures, or “Chapters,” are forming above the vertical product teams, or “Squats” as they’re referred to in the Spotify organization. These Chapters ensure the consistency of individual agile activities, as well as their suitability within the specific context of the company. Enterprise architecture represents such a Chapter in a setup like that.
CBA Lab also has very traditional companies as members – for example Deutsche Bahn, Ergo, Bosch, and others. These companies are not at all like Spotify or Amazon. Does this therefore mean that parallel organizations are forming, or will form, in such traditional companies?. How is one to imagine such a development? And how are these companies going to get where they need to go?
We are currently seeing a tremendous transformation effort under way at all of our member companies across all business sectors. All of these companies are moving in the same direction, but many of them are still at the beginning of their journey. Digitalization has moved enterprise architecture to the center of strategic business transformation. There’s a great spirit of optimism among architects, who now feel as if they’re in demand and on the front line of new developments. In this situation, CEOs are basically becoming chief architects, as they are the ones who are designing and orchestrating the transformation. Some people, such as Volkmar Denner from Bosch, are completely aware of this and have actively taken on this new role. Volkmar Denner was the one who developed the vision of Bosch establishing connectivity, and he’s now moving the entire company in this direction, with regard to its products, processes, and organization as well – everything in fact.
People at companies rarely speak about architecture. Instead, managers tend to demand flexible platforms and customer-centric approaches, or else express their desire to have a data-driven company. Compared to these things, the term architecture always sounds a little like an activity with which one merely occupies themselves. Do you see this as a problem as well?
Yes, the word “architecture” is a bit of an unfortunate choice here because it’s generally associated with static principles. The word architecture doesn’t really express the idea of the actual task at hand, which is to manage dynamic processes. A word like “choreography” might be better here. Nevertheless, the type of work architects do has changed dramatically – the big ivory towers are truly a thing of the past.
What about architecture’s involvement in the innovation process?
Architecture is a facilitator of the innovation process at the interface to IT. At companies that are now changing their old business models, architecture is not replacing the traditional business strategy as it relates to what a company wants to be and what it should be doing for its customers. However, architecture can enable the establishment of interconnections that make it possible to turn existing assets into new products and services. Any company looking to transform itself into a data-driven enterprise needs to keep in mind that enterprise architects are the ones who know which data will be needed, which processes generate that data, and where the data can be obtained externally if necessary. For example, machine learning systems need an incredibly large amount of data in order to identify relevant patterns, and no company can produce and store such a large volume of data on its own. In addition, business innovation often occurs in a bottom-up process that starts with architecture. Take the blockchains that everyone’s always talking about: Distributed ledgers represent a structural approach; the algorithm part is a completely secondary component. The derived attributes from this structure are used to create new business ideas from the bottom up – for example for financial transactions without intermediaries. This is the opposite of a top-down development resulting from a business requirement, and the concept behind it is a textbook case of innovation.
How are architecture teams organized at companies – does the structure differ from that of a CIO or CDO office? Is there some kind of template that can be used or does every company need to come up with its own ideas here?
As I mentioned earlier, there’s a clear trend toward decentralized and product-centric organizations. The boundaries between business and IT are disappearing, and this is the case with regard to enterprise architecture as well. IT is no longer a secondary process; in the era of digitalization, IT is the name of the game, and we can expect to see a transformation of both structures and expertise. For decades, IT was outsourced – in people’s minds as well. In the future, every top manager will need to understand IT and manage it at least in the sense of being able to effectively utilize IT and its interactive relationships as strategic instruments. Because I have this view, it always surprises me to see how companies have two different organizations for the CDO and CIO. The data used in data-driven enterprises has to be generated by the sensors employed in new production processes, new machines, and new platform products, whereby all of this relates to primary processes. Architects need to be able to take systems as a whole into account, even as the boundaries for the scope of their activities continue to be pushed outward.
Which will make architecture more important, of course.
Absolutely. Its scope has already gotten much bigger. With today’s products, “the system” begins in the customer’s living room – for example with Alexa – and extends deep into the supply chains of supplier companies. Digital platforms are merging products and services into service packages that can involve multiple companies and multiple business sectors. Enterprise architecture no longer only relates to a single enterprise; it also involves the orchestration of entire value chains across companies and business sectors. That’s why we changed our name to Cross-Business-Architecture Lab. At the same time, everything is moving faster and faster. To use an example from physics, a higher motor rotation speed combined with a greater external mass due in our case to decentralization leads to a greater centrifugal force. Enterprise architecture can harness and control this force, which is why enterprise architecture is what makes autonomous decentralized product teams possible in the first place.
Dr. Johannes Helbig
Chairman of the Board
Cross-Business-Architecture Lab e. V.
Enterprise architecture no longer only relates to a single enterprise; it also involves the orchestration of entire value chains across companies and business sectors.