CHAPTER

11

BUILDING AND MANAGING TECHNOLOGY INFRASTRUCTURE

Because digital tools and technologies are all the rage, it’s sometimes easy to overlook the underlying infrastructure that supports them. In the case of digital transformation, this infrastructure comprises two core elements: IT infrastructure that is stable and flexible and high-quality data assets that are secure.

Technologies such as databases, servers, network infrastructure, and enterprise software applications are not especially glamorous, but the success of any digital project hinges on them. Artificial intelligence, for example, is not a self-contained technology. In fact, unless it’s plugged into a secure set of organizational systems, as well as accurate and compatible data sources, it’s useless. A well-functioning technology infrastructure supports the whole digital transformation value chain—from data capture to storage, analysis, and dissemination of results—all in (or close to) real time.

WHY IT MATTERS

Abraham Lincoln once said, “Give me six hours to chop down a tree and I will spend the first four sharpening the axe.”

The same logic applies to technology infrastructure. Without a proper infrastructure, management’s resources and attention will be consumed with executing and maintaining core processes rather than figuring out how digital tools can generate new sources of value.

Technology infrastructure consists of a set of integrated and shared systems, processes, and data that together ensure operational efficiency and transparency.¹ When executives are asked about the main obstacles to proper digital execution, the limitations of existing IT systems are often cited as the number one roadblock.²

According to BCG, two out of three successful transformations were linked to business-led, modern fit-for-purpose technology and data platforms that supported the development and scaling of digital use cases.³ Unfortunately, more than half of the companies in the study struggled with a lack of flexibility in their technology infrastructure.

GE invested billions in new technology, people, and processes to build an ambitious new division, GE Digital, to capitalize on the growth of digitization in the industrial world. However, the new division’s marquee product—the Predix industrial internet platform—never lived up to its billing. Its main problem? It simply didn’t work very well from a technology point of view.⁴ Getting your technology infrastructure right is critical for digital transformation execution.

BEST PRACTICES AND KEY INSIGHTS

Paradoxically, recent advances have made it both easier and tougher to build a solid digital platform.⁵ On the one hand, cloud, agile development methods, external code libraries, and easy-to-use development tools have enabled developers to build new digital solutions rapidly. On the other hand, these approaches can produce an uncoordinated mélange of tools and applications that collectively weaken, rather than strengthen, a digital transformation. Hence, the first step is to select a technology architecture that matches the organization’s digital ambitions. But that won’t be enough. Among the many challenges that a technology infrastructure must overcome, four have critical relevance. Two are linked to IT infrastructure (stability and flexibility), and two are related to data (quality and security).

Architecting Your Digital Infrastructure

The trouble with designing a technology architecture that will serve the organization’s ambitions is that it must be built for efficiency and agility. For that reason, companies that have successfully executed a digital transformation have layered their technology infrastructure.

The first component is an efficient operational and transactional backbone. This is a traditional domain for industrialization, optimization, and efficiency, and it covers areas such as back-office systems and systems of records.

The second layer is an agile customer-engagement platform that covers areas from key transactions (such as payments) to the delivery of customized experiences and connections with ecosystem partners. This is a domain for user experience and quality service delivery.

The third component is a data and analytics platform that will use internal data and capture connectivity data to perform advanced analytics and/or build and test algorithms. This is a domain for insights and innovation.

Because of their differing objectives, each layer requires specialized management, capabilities, technology partners, and metrics. If this sounds complex, that’s because it is. But choosing an architecture, up front, that matches the company’s business ambitions will guide future technology investments and ensure the coherence of the technology/IT operation.

IT Infrastructure Stability

From an IT perspective, stability is closely linked to reliability, consistency, and usability. Reliability means that the infrastructure is consistently up and running, and is also immune to various shocks and adverse events. Consistency refers to the infrastructure’s ability to provide a standard quality of service across operating environments, such as different operating systems or app ecosystems. Usability means that the infrastructure is easy to use. In sum, a stable IT infrastructure works most of the time, across a wide variety of use cases.

For example, most blockchain applications have relatively light infrastructure requirements: high processing speed and plenty of storage. Conversely, many AI solutions have high computing and network resource needs and require compatibility with data sources.

According to a study of CIOs, the enemy of IT infrastructure stability is complexity.⁶ Typically, the more complex the infrastructure, the less stable it is. A positive development has been the growth of cloud solutions that tend to be more stable than on-premise systems since they are centrally managed and fit for purpose. For this reason, many organizations are investing heavily in cloud infrastructure to support their digital transformation programs. However, linking cloud solutions to legacy systems remains a significant challenge.

IT Infrastructure Flexibility

On the surface, IT infrastructure flexibility may seem inconsistent with stability, but the two go hand in hand. Infrastructure that is stable but not flexible is dangerous, as it is unable to adjust to changing conditions. Unfortunately, because many legacy systems (and some cloud solutions) are not very flexible, organizations struggle to keep up when the environments change.

Good infrastructure allows for the establishment of feedback loops, whereby successes and failures can be quickly flagged, analyzed, and acted upon. For instance, when Ticketmaster wanted to tackle the growing problem of “scalpers”—people who buy tickets ahead of genuine customers in order to resell those tickets at a premium—it turned to machine learning algorithms.⁸ The company created a system that examined real-time ticket sales data, along with a holistic view of buyer activity, to block out resellers and reward legitimate customers with a smoother process. But as the company soon discovered, resellers adapted their strategies and tools in response to the new system. Ticketmaster’s infrastructure was flexible enough to modify its pricing tool to include feedback loops, allowing its algorithms to keep up with the resellers’ evolving techniques.

Sometimes it makes sense to differentiate between layers—e.g., between core back-office IT infrastructure and front-facing systems such as the websites and apps that work with customers and ecosystem partners. When it comes to integration with digital tools and technologies, this external-facing infrastructure is more than just a pretty front end.⁹ It performs key transactions and connects back to the core platform to complete the work. The front-facing infrastructure needs to be more flexible than the back-office systems and should be designed accordingly.

Data Quality

Bad data is more than an annoyance. In 2013, the US Postal Service identified 6.8 billion pieces of mail (4.3 percent of the total) that was undeliverable due to bad data, costing the service more than $1.5 billion.¹⁰ Fast-forward to 2020, when 20–30 percent of email addresses change each year,¹¹ and there are numerous data issues that can impact the effectiveness of a digital solution, including data incompatibility across different systems, duplicate data, lack of data currency, data inaccuracy, and data ambiguity. Indeed, one study published in the Harvard Business Review estimated that just 3 percent of companies’ data met basic quality standards.¹²

It’s often worthwhile to conduct a data quality audit.¹³ These audits first establish a baseline of data quality standards, rules, and expectations, and then run checks across the organization to determine whether the data meets the required benchmarks. This audit should be followed by a data cleansing process.

Data Security

Cyberattacks are so common today that the importance of data security is hard to overstate. However, several practices can reduce the likelihood of being compromised.

First, most attacks exploit known vulnerabilities and can be addressed with relatively easy fixes. When a vulnerability in a piece of software is identified, a clock starts ticking; now it becomes a race between the software vendor (to develop a patch) and hackers (hoping to exploit the vulnerability). This race does not stop when a patch is released, as many software customers fail to quickly patch their systems. For example, Maersk was hit badly by the NotPetya virus in 2017, even though Microsoft had already released a patch that would have protected Maersk’s systems from the virus, had Maersk installed it.

Second, routine maintenance and real-time monitoring can identify attacks and vulnerabilities and reduce the chances of future attacks. Standard practices should be followed, including the implementation of strong password policies, administrative privilege restrictions, and regular security audits. Random intrusion tests, including simulated phishing attacks, are also highly recommended. Because more than 90 percent of successful hacks and data breaches stem from phishing, the workforce should be educated on good and bad security policies, and best practices should be enforced for all users.

Finally, incentives should be aligned with security-friendly behavior. In the case of Maersk, the patch upgrade had been approved by senior management, but the IT administrators’ bonuses were linked to uptime and other measures of current performance. Thus, the administrators had no incentive to implement the upgrade, as this would have slowed system performance.

Hacker’s Toolbox

Fuse IT and business digital road maps. Technology infrastructure is often under the purview of the IT department, so close coordination with IT is important. Many first-wave digital transformations tried to go it alone (without IT) and failed.¹⁴ Today, IT leaders drive digital transformation in some companies, while digital and IT leaders work closely in others. As business leaders become more tech savvy, and IT leaders and their teams become more business savvy, getting digital and IT to cooperate on infrastructure is a more realizable goal than ever before.

Don’t avoid the tough fixes. Many organizations are using out-of-date, inferior, and unsecure systems, applications, and databases. Why? Because the prospect of changing them is so painful—e.g., if a system is critical to the company’s operations, taking it offline might have unpredictable and expensive consequences. Unfortunately, legacy technologies are not like fine wines. They don’t improve with age. In most cases, it’s better to update the infrastructure as soon as possible.

Think about your digital platform architecture early. Empower your enterprise architects. You will need to have an appropriate target architecture in mind as you invest and transform your technology infrastructure. This synergy is critical for executing your business strategy.

Establish clear accountability for data management. Data, whether customer insights or operations and connectivity information, will most likely be at the heart of your digital transformation. Today, 57 percent of Fortune 1000 companies have appointed a chief data officer.¹⁵ Even without a chief data officer, you will need clear data management accountability for governance, operations, innovation, analytics, and security of your data assets.

Educate the senior leaders. Don’t tackle the complexity of building a proper technology infrastructure on your own. Share with senior leaders your architecture plans, the trade-offs being made, and the progress of your technology buildup.

RELATED CHAPTERS

Choosing the Right Digital Governance Model (Chapter 8)

How to Make Digital and IT Work Together (Chapter 9)

Managing Digital Transformation Responsibly and Sustainably (Chapter 16)

Staying on Top of New Technologies (Chapter 29)