The four pillars of digital architecture — are you ready for excellence?

The Digital Neanderthal
14 min readMay 25, 2021

In 2018, Kim Hammonds, then the CIO of Deutsche Bank, described her employer as the most “dysfunctional” company she had ever seen. Her job was to transform and simplify Deutsche Bank’s digital landscape, reduce the number of systems, and make them more reliable. The CEO, John Cryan, had described Deutsche Bank’s IT as “lousy.” Even so, the board did not appreciate Hammond’s brash public commentary and unceremoniously fired her.

Technology is complex, and so is a company’s information technology. Digital leaders like Amazon and Google revolutionize workplaces and consumer behavior. They are also seriously threatening incumbent companies. One might expect that those challenged frantically develop an array of digital responses.

The bleak reality, however, is that most companies are unable to cope. Companies may strive to become data-driven and process-optimizing, yet their IT landscape resembles a map of medieval castles, full of roadblocks, walls, and barriers.

Rather than thinking of individual applications, C-level executives must think of their entire IT ecosystem as a system that needs to integrate many diverse applications. Unfortunately, stakeholders in business and IT are highly entrenched in old ways of working, used to a set of applications. As a result, companies fail to develop a digital landscape into a competitive system.

Four criteria determine the fighting power of your IT landscape: efficiency of resources, scalability to size, optimal diversity of components, and adaptability to circumstances. These four essential criteria are universal benchmarks for any complex system facing competitive pressures. They provide an intellectual framework by which a company can effectively evaluate and enhance its digital future.

Failure to rise up to the digital challenge should not be an option. Yet, when systems fail, they do so rapidly — and often with disastrous consequences.

Losing against all odds

History is full of technological failures: the Challenger incident in 1986, the Deepwater Horizon oil spill in 2010, and the Lidl pullout from its SAP project after investing half a billion Euros. The failures of management are obvious in hindsight and sometimes border on the incredible. The same is true for what is likely the most consequential technological breakdown in history.

In 1940, the French Army collapsed against the onslaught of the German Blitzkrieg campaign. Using war as an example might seem a bit dramatic, but it provides a few practical advantages: an ample supply of wrong decision making, excellent historic documentation, and proximity to technology. One might expect that the high stakes and the prospect of death would entice leaders to strive for excellence. Hopefully, the following example illustrates that this is not always the case.

At the beginning of 1940, the French High Command felt comfortable and strong against Hitler’s forces. By June 1940, German forces had rapidly advanced and French defenses were collapsing. A situation began to unfold that had been unimaginable just a few weeks before. The German commander, Heinz Guderian, commanded his troops from a mobile platform, utilizing state-of-the-art radio communication with his forces.

General Maxime Weygand directed troops from a remote castle where his only method of communication was a telephone next to the castle’s outhouse. Engulfed by latrinal stench, Weygand’s subordinate officers battled for telephone time. Making matters even worse, the telephone lady insisted on her lunch break from 12 to 2 p.m., entirely cutting off any communication during the heat of the most critical battles. From 6 p.m. onwards she went home, returning at 9 a.m. in the morning. General Weygand remained unfazed and apparently even refused to send letters via homing pigeon.

Testifying later in front of a national commission, junior officers would claim that the High Command had been “a submarine without a telescope.”

By the standards of a strategy consultant, France had prepared well: it built the Maginot Line, a multi-story bunker extending over hundreds of kilometers, enforced with steel and concrete. France possessed two to three times more planes and tanks than Germany. The country was part of an overwhelming alliance with access to far greater resources. Only a few dissenters, like General de Gaulle, claimed that the Army lacked mobility and was unprepared for a mobile war of the future.

It turns out that the lack of communication had disastrous consequences also on a tactical level.

The majority of the German Panzer troops had been considered ineffective as they lacked thick armor and were only equipped with machine guns. Their cannons were unable to penetrate the armor of the French Char B tanks.

Still, the French lost, as the Germans had tightly coordinated their tanks, airplanes, and ground troops by radio. If ground troops were in trouble, close air support eliminated resistance. Many other French tanks were destroyed by a powerful anti-aircraft gun, the Flak 8.8.

While equipped with more powerful cannons and better protected by thicker steel, the French tanks had no radio. To communicate, the commanders were forced to use signaling flags, only to be killed by waiting sharpshooters when exposed. French tanks required special refueling tanks which were mostly unavailable. German tanks received their fuel via canisters, dropped on the wayside by trucks or airplanes.

Germany won an overwhelming victory with significantly smaller resources than the British and French combined. It would unleash a humanitarian catastrophe and put France at the mercy of German conquest for almost five years.

Integrated systems

The Battle of France in 1940 holds an important lesson applicable to modern IT: technology needs to be understood as an integrated system. A functional IT architecture is an ecosystem of various applications, staff, vendors, and partners. Getting the mix right defines the essential criteria for winning systems. Getting it wrong ensures failure.

Simple technology (like canisters) may be better than a highly sophisticated solution. Linked weapons, cleverly combined, may transform into an innovative system exhibiting an overwhelming force against traditional tactics and weapons. The French generals refused to attack Germany. After the Battle of Poland, Germany had run out of ammunition and was left completely vulnerable. Without a strategy to incorporate emerging technology, any organization will fail once exposed to competitive pressure.

What makes a system competitive?

Over the last decades, natural science has increasingly embraced principles of self-organization and complexity theory. Totally different systems (social organizations, ecological habitats, planetary systems, and even mathematical functions) can exhibit similar behavior patterns because they observe evolutionary patterns.

Very rarely does evolutionary pressure drive a species into extremely coordinated behavior. On this planet, there are about twenty highly social species that have developed extreme altruism and cooperate in large social entities. Biologists call these biological systems superorganisms.

Most of these great collaborators are within the insect world, of which ants are the best researched. Ants, like other eusocial insects, have achieved overwhelming dominance within their habitat, evolving behaviors that facilitate agriculture, architecture, and organized warfare.

There are four characteristics that determine a winning system. Ant behavior provides an excellent illustration on what they are and how they are achieved:

Efficiency: Any system must be able to operate rapidly with a minimum of resources. Ants are incredibly efficient as they have learned to cooperate seamlessly and en masse. With little brainpower, they lay traces and follow them. They care for the wounded, have mortuaries, agriculture, farming, and organized warfare.

Scalability: Size matters in conflict and competition. Most ant colonies only have one queen. However, some ants have turned to be “hyperscalers” by accepting several queens within their nest. The supercolonies of Argentinian ants stretch for six thousands kilometers along the Mediterranean coast, from Spain to Italy. These colonies house trillions and trillions of ants and dominate their respective environments.

Diversity: Having different players allows for a division of labor and individuals that excel at certain tasks. There is an optimal level of diversity, as too much specialization requires too many resources. Army ants start to collaborate at a group size of six ants. However, most ants have three major casts: queens, drones, and workers. Often special soldier ants develop, with enhanced mandibles, bigger heads, and more power. Occasionally, there are also ants called repletes that act as living food storage, guardians with oversized heads that block the entrance of the nest, and gamergates, workers with the potential to turn into queens.

Adaptability: Systems need to cater to conflict, daily life, and the future. Evolution mercilessly puts survival to the test. Ants exist on all continents, save Antarctica. They survived the end of the dinosaurs and have existed on this planet for more than 160 million years. Lacking crisis resilience, even systems that have been around for eternities can fail catastrophically.

These general evolutionary principles govern the creation of any complexity. They apply to microscopic complexity like the folding of protein molecules in biological systems and, as some physicists suspect, extend even to the creation of the universe and its natural laws.

In today’s digital world, you ignore these principles at your peril.

Criteria for your system landscape

How does one measure the resilience of an IT system? And how can we define efficiency, scalability, diversity, and adaptability in terms of a digital landscape?

Efficiency

An IT system requires rapid connectivity and low maintenance costs. Therefore, an IT architecture that claims to be modern must qualify for seamless interfaces, data availability, and reasonably low headcount requirements. Primarily, it must be capable of coordinating various interconnecting applications.

General Weygand was unable to do so from his telephone. His counterpart, the German general Heinz Guderian directed his troops from the front lines. He rode on a state-of-the-art communications vehicle, accompanied by a group of his best radio operators. Weygand’s chateau was probably a very comfortable place, even if it only had an outhouse. Yet, speed and lost costs are of essence — then and now.

Scalability

Systems need to match up to economies of scale. When a business grows, costs should potentially scale down proportionally. As turnover grows, average license costs and the proportion of people to maintain systems should decline.

IT must be capable of supporting acquisitions. Very often, companies spend years trying to merge the IT of two companies into one. Particularly, enterprise resource systems must possess standards that enable them to be rolled out quickly to other entities.

Diversity

Systems vary according to the purposes they fulfill. For example, there may be unique systems for computer-aided design, transport planning, a customer rebate program, and production planning. Other systems provide standard functions such as billing, inventory control, order management, finance, and controlling.

The fine art of an astute general is to integrate new weapons to generate overwhelming fighting power. Similarly, executives must find the right mix of old and new technologies to drive their business.

Adaptability

A company needs to adapt to a highly competitive environment, and your IT systems must evolve accordingly. In addition, the systems function within a constantly changing environment; keeping track of rapidly changing websites, meeting the different demands of customers and products, and planning for resource shortages are all things your systems need to cope with.

Most IT systems are hopeless on resilience. Security threats are on the rise. Around half of all major companies have been hit by severe security incidents, often bringing their operations to a grinding halt. Companies expose themselves to a deeply criminal world. Hackers do not even shy away from stopping the works of a hospital or its intensive care units. Adequate preparation does not only reduce the risk of intrusion, it also makes your systems resilient once breached.

IT Architecture

What should an ideal IT landscape look like? First, of course, companies have hundreds, even thousands, of systems. However, systems providing business applications generally fall into three essential layers. These layers are universal and shared by almost every modern corporation.

ERP (Enterprise Resource Systems): These are systems like SAP, Microsoft Dynamics, Infor LN, and Oracle ERP. The ERP system is your core transactional system and the repository of data. It typically provides for functions such as order taking, delivery management, invoicing, limited production and logistics, and finance and controlling.

Advances in web technology, faster processing, and more powerful databases have streamlined these systems, but from the perspective of functionality, ERP systems have not evolved much since the 1980s. There is still little advanced functionality to benefit users. As a result, ERP systems are generally the least innovative applications within a system landscape.

ERP systems should scale in licensing and hardware, but mainly in the proportion of staff employed. The larger the system, the lower the costs should be. A substantial diversity in ERP systems is not advisable because there is little actual difference between vendors. Too much diversity results in complexity costs and lower efficiency. The ERP system should be as open to other systems as possible. The simpler the technology stack, the higher the system resilience.

The ERP system is the digital backbone of your company. It is the skeleton and should provide structure, but not be bendable to the whims of individual stakeholders or flimsy ideas.

Specific Applications: Apart from an ERP system, companies acquire many specific purpose-built applications. Examples for these applications are Customer Relations Management (Salesforce, Microsoft CRM, SAP CRM), reporting dashboards (Tableau, Qlik Sense, etc.), and various programs for computer-aided design. They may also cater to financial consolidation, production, warehousing, logistics planning, credit information tools, and duty management. Special applications help fulfill particular functions and provide great add-ons to the diversity of your systems. They ensure unique tools for specific tasks but are generally not good at delivering much else.

All ERP vendors have made attempts to extend their marketability by adding specific applications to their portfolios. However, more often than not, these solutions are expensive and poorly connected, even into the vendor’s technology stack. The new strategy of SAP’s CEO is to seamlessly integrate applications like Fieldglass, Ariba, SuccessFactors, and their SAP-CRM suite into the architecture. However, even if Christian Klein is successful, SAP’s specific applications will remain expensive.

Smaller vendors provide cheaper and more innovative solutions. Incorporating them into the IT landscape is the great art of digital excellence. Most of these vendors (but it still is imperative to check) will provide standard connection capabilities into ERP systems.

Software Development: Software development can meet the exact demands of business processes with extreme specialization. Software development caters best to internal process optimization and customer focusing applications. Properly built, it provides a platform to interconnect different systems and processes. Thus, software development plays a critical role in business process management, a rapidly growing area in application management.

Companies like Amazon mostly rely on software development to drive customer focus. Software development supplies diversity and efficiency at speed. In addition, software development is exceedingly scalable because a growing user base does not translate into extra license costs.

However, software development does pose barriers to entry: start-up costs are high, there are few sponsors within the business, and an ERP-centric IT department tends to reject bespoke development.

How to plan your ERP right

Traditionally, most corporates are still highly ERP-centric. ERP systems still perform the lion’s share of digital processes. They tend to be the focus of business and IT alike. However, treating an ERP solution as the sole driver of digital excellence is a grave mistake. Similarly, ignoring the integrity and stability of an ERP system would be an even bigger mistake. Using a multitude of ERP systems will increase the costs of diversity and will turn an IT landscape into an unmanageable supertanker.

If you like to ride a speedboat, keep the ERP clean and fit for purposes: accounting, controlling sales, inventory management, some logistics, and a limited set of production capabilities. Preferably, use one solution instead of many different ERPs, which will inflate staff and interfaces; interconnections will cost you dearly in time and money, and costs will balloon unnecessarily.

The cloud

The cloud is nothing more than externally hosted infrastructure services. Infrastructure can be expensive, especially when systems expand. Hyper-scalers can efficiently supply essential infrastructure services. Most software vendors will try to sell cloud as an added value or second product, often with a hefty premium above what a hyper-scaler would charge.

But why would you rely on a vendor of software for hosting services? Why wouldn’t you host the service in AWS, Azure, or elsewhere? Mixing the administration of your application with the hosting of infrastructure can make you very inflexible. For example, SAP’s cloud service is known to be rigid and error prone.

Don’t think of the cloud as the solution to all your problems. If done well, the cloud can be sufficient and scalable, but make sure you don’t pay through the nose and that you get the service you need.

Look at the facts!

The discussion on IT architecture needs to be open-minded and fact-based. Evaluate any system in the context of your ecosystem and its efficiency, scalability, diversity, and adaptability. Consider the ecosystem of your vendor. Do they have a good partner structure? What operating systems are they running? Will this partner be around 20 years from now? Keep an open mind and evaluate your options very carefully. Glossing over details can cost dearly as they may turn out to be very important later on.

Within your organization, business applications come with fiefdoms and islands of knowledge. Sales and marketing departments of major vendors are excellent at exploiting this weakness. But unfortunately, they rarely win on technological excellence.

In my experience, a solid blueprint written in a free-text format provides a much better basis for evaluation than PowerPoint slides. Note that Jeff Bezos of Amazon has banned PowerPoint from the company. Once IT projects turn into endless and unfocused townhall discussions, they are doomed. Consultancies, on the other hand, love them: they provide a rich source of income and great training grounds for their junior workforce.

Digital Excellence

First and foremost, excelling in the digital world requires intellectual astuteness. Times of technological transformation require solid and decisive leadership. Cognitive openness and fact-awareness will prevent a company from turning into a submarine without a periscope. Customers and efficient processes are the new strategic battleground. To achieve them, the IT ecosystem must focus evolve to optimal efficiency, scalability, diversity, and adaptability.

A well-built IT architecture is the first step to getting it right. The most expensive solutions are rarely the best, and creativity thrives on frugality. You may be better off using the digital equivalent of a petrol canister than an ultramodern, multimillion-dollar refueling tank.

Take a deep look at your system landscape. Break down each component by its four success factors. Are the systems interconnected and efficient enough? Will they scale sufficiently without incurring excessive license and hardware costs? Are your ERP systems too diverse or would the costs and timeframes of harmonization block your digital agenda for years to come? How adaptable are the systems to digitally native customers? Are they resilient against crises and security threats?

A definition of business needs and technological capabilities will provide a corporate-wide digital consensus. Expertly done, the exercise may only take a few months and will provide fact-based consensus. Respect for facts provides a common ground forward, essential for launching a digital strategy.

Since 2008, Deutsche Bank’s share price has dropped from a high of 93 EUR to 5,43 EUR in April 2020. Better late than never, Deutsche Bank is actively pushing a new digital agenda. Its new board member responsible for IT is Bernd Leukert. Leukert is an ex-executive of SAP, formerly responsible for the application platform S/4 Hana. Initially, he did not announce a radical change, nor did he attempt to bulldoze stakeholders brazenly. Instead, Leukert started with a deep dive into Deutsche’s architectural constraints and business needs. From there, he is leading the simplification and standardization of its IT. Given Deutsche’s desperate situation, he encounters less resistance and is likely to gain more support than Kim Hammonds.

Under his leadership, Deutsche and Google have allied. The objective is to launch a new suite of technology-based financial products.

Since then, Deutsche’s share price has steadily recovered to over 12 EUR (at the time of writing, 22.05.2021). The verdict on Leukert’s long-term strategy is still out, as it will only pay off in the coming years. Clearly, Deutsche’s new digital strategy is a calculated gamble; it may succeed or fail. However, to date, the digitally-driven strategy shows promise and may launch a comeback of Deutsche as a major player in the world’s banking community.

Treating IT as a complex ecosystem is the first step into embracing the digital challenge. Disjunct systems fail under competitive pressure. Evolving systems have the uncanny ability to start a life of their own and leverage unforeseen forces. Thriving in the new world requires dedication, intellect, and the right approach.

--

--