By Jeff Taylor and Paul Renno
Companies in all industries are grappling with the same critical decisions about how to build the digital capabilities that will support their future growth. But what next-generation technologies will they use in their quest for digital maturity?
Bain & Company wanted to determine the future role of containers, a new technology that enables software to run reliably when moved from one computing environment to another. We partnered with Red Hat, the world’s leading provider of open source software solutions, to survey 449 U.S. executives and IT leaders across industries.
Despite the mounting pressures, we found that few traditional enterprises have made substantial progress on their digital journey, and even the most digitally mature traditional enterprises are not on the bleeding edge. Yet, among the surveyed companies, the 15% that are farthest along on the digital maturity curve are 8 times more likely to have gained share than the 15% of companies that are least mature.
The most digitally mature firms invest in technology that delivers adaptability, resilience, speed and the ability to use analytics for better-informed decisions that improve the customer experience and operations. Among the most digitally advanced traditional enterprises, all report that they consider their architectures to be agile, adaptable and scalable, vs. only 16% of those that are least advanced.
More digitally mature companies also recognize that their existing infrastructure and applications do not need to be a drag on digital efforts, and see optimizing and modernizing existing technology as part of the overall journey.
When comparing more and less digitally mature companies, our survey showed that investment varies dramatically in cloud-based architecture, data access and advanced analytics (machine learning, big data), infrastructure optimization technologies, and modern application development and deployment platforms such as containers. In the newest of these technologies, modern application development, companies that are most digitally mature are three times more likely to invest than the least mature companies.
Containers directly or indirectly enable several key attributes of digital transformation. We found that respondents are beginning to beneﬁt from faster innovation as well as improved development and deployment cycles when they use containers. For example, adopters frequently report 15% to 30% reductions in development time. Adopters also report initial cost savings of 5% to 15% due to greater hardware and process efﬁciencies. Containers’ superior portability also improves the ﬂexibility and scalability of IT architectures, with some adopters mentioning containers as a step toward migration to more cloud-focused architectures.
Given these beneﬁts, container adoption is expected to grow across all phases of the application life cycle (development, testing and production), with the growth being most dramatic in production. Some perceived hurdles to adoption include security issues, the impression that most workloads cannot be containerized (i.e., applicability) and worries about the lack of enterprise-grade persistent storage options. Given these concerns, we wanted to understand if containers could replicate the rapid adoption of analog technologies such as server virtualization once the perceived obstacles are lessened. Our research suggests that there are reasons to be optimistic about the future of containers.
Many companies evaluating containers ask a fundamental question: Are they as broadly applicable as server virtualization to a spectrum of workloads? While overcoming the applicability barrier will be key to containers’ future path, survey respondents report a growing set of workloads being prioritized for containerization. We see progress on initially more difﬁcult-to-containerize stateful applications, and companies are showing high interest in moving beyond web applications to containerize more traditional applications over the next three years. In addition, while today early container adopters are prioritizing net new applications engineered in ways that are easy to containerize (e.g., microservices), we are increasingly seeing examples of companies containerizing older, monolithic applications, further expanding containers’ applicability.
As with other applications, containerizing these legacy apps has potential to improve adaptability and cost efﬁciency by increasing portability, decreasing complexity and streamlining the installation, upgrade and rollback process. Among the companies beneﬁting from the new technology, Dell has publicly stated that it used the industry-leading Docker container format to containerize a 20-year-old monolithic in-band systems management tool that remained fully functional while leaving no footprint on the host. Containerization simpliﬁed installation, upgrade and rollback, and could run on any Linux distribution supporting the Docker format.
Now, as containers grow in value, industry momentum will build around them. Battle lines are being drawn in container orchestration and management between lead innovator Docker, which continues to drive the container format standard, and solutions from incumbent software vendors as well as leading cloud vendors (e.g., Google’s Kubernetes), and many respondents are indicating that they are implementing multiple solutions or even developing their own.
We expect containers to continue to coexist with virtual machines (VMs) in the near future, as respondents indicated that over the next three years they will increasingly deploy containers on VMs as opposed to bare metal. Despite the limitations on cost efﬁciencies from containers when implemented on VMs, companies continue to value VMs’ security, familiarity and integrated solution suites.
Whether or not they replicate VMs’ rapid ascent, as container usage continues to evolve and containers become more applicable, we see them becoming an increasingly attractive option for companies across many industries, as they struggle to meet the intensifying digital imperative.
Jeff Taylor is a Bain & Company partner based in Boston. Paul Renno is a partner based in San Francisco. Both are members of Bain’s Technology practice.