The beautiful thing about cloud is one can switch providers, or reduce risk by having components platformed by different providers. Choice is a wonderful thing. Enterprises are taking advantage of today's range of choices -- and seeing marked improvements in software productivity as well.
That's the gist of a recent study out of the IBM Institute for Business Value, in collaboration with Oxford Economics, which surveyed 7,164 C-suite executives across 29 industries about their cloud formations. Only three percent reported using a single private or public cloud in 2021, down from 29% in 2019, the survey finds.
Avoiding vendor lock-in was the prime motivation for moving to multiple clouds. At least 79% of respondents said workloads being completely portable with no vendor lock-in is important or extremely important to the success of their digital initiatives. Plus, 69% said vendor lock-in is a significant obstacle to improving business performance in most or all parts of their cloud estate.
There have been four generations of cloud, the survey's authors -- Howard Boville, Hillery Hunter, and Richard Warrick -- point out. The first two stages were characterized by single-purpose subscriptions and credit-card purchases of SaaS applications. Most companies are now in the third stage of the cloud evolution, in which conventional applications, compute and networking are being shifted to cloud providers, the survey's authors state. In the next stage, which a few forward-looking companies are entering, begins to digitize the entire business, both inside and out.
Along with the end of the single-cloud era, the survey unveiled some additional surprises. First, contrary to conventional wisdom, enterprises in the survey do not appear to be hamstrung by talent shortages in cloud development and maintenance. "Among all the obstacles listed in the survey, talent gaps were among the weakest," the survey's authors -- Howard Boville, Hillery Hunter, and Richard Warrick -- indicate. More than half, 54%, report that leadership talent is "not a significant obstacle" anywhere in the cloud estate, and 47% reported that the availability of "people with the right skills and experience" is "not a significant obstacle" anywhere in the cloud estate. "High levels of SaaS adoption may be making talent shortages less acute," the authors state, implying that much of the backend and software work is being relegated to cloud providers.
The current permutation of cloud -- the cloudification of software development, compute resources and networks -- is resulting in a more productive operating model for the business, responding executives agree. Seventy-three percent state that the cloud model is delivering or sustaining "Improvements in product and service delivery productivity, Another 72% are seeing "improvements in software/digital product quality -- reducing defects, and rework."
When drilling down further into the impact of cloudification on software development and delivery, executives report "significant improvements to many dimensions of software development: customer outcomes, velocity, quality, safety, value from data, and the like," Boville and his co-authors observe. The leading benefit is cloud services helping to reduce dependencies between teams and components of software, cited by 78%. Another 75% report progress in "improving developer productivity by automating repetitive tasks," and 71% say cloudification means 'working with a high degree of psychological safety" -- very important for developers working with high-visibility projects.
Technical debt -- a festering concern at many technology-intensive enterprises -- also gets alleviated, as mentioned b 69% of executives.
The current third generation of cloud implementations "is distinct because it incorporates other technologies and practices," according to Boville and his co-authors. "Cloud, software development, data engineering, and a host of modern practices such as design thinking, agile, SecDevOps, and site reliability engineering are meshed together."
The report's authors offer some recommendations for taking the next step into cloud as a vehicle for more far-reaching digital transformation:
Start with end-to-end workflow redesign as an entry point to advanced cloud. Boville and his co-authors suggest asking questions such as: "Where are the most powerful, transformational opportunities for reinvention? How much business value could be unlocked via cloud technologies, including APIs, data fabric, and AI? How much business value could be unlocked by improving full product lifecycle delivery velocity?"
Leverage existing assets. "Mainframe computers are one example," the co-authors explain. "The combination of mainframe- driven applications and a hybrid cloud platform can be powerful. Monolithic systems, which can be converted to microservices modules of software, are another. And cloud can use data fabric to unlock and democratize hugely valuable stores of customer and operational data."
Fund advanced cloud initiatives as products, not projects. "It's better to develop digital products as living assets that begin with a prototype and advance with many cycles of learning and development. Products need to be funded based on the requirements of the product life cycle and on frequent customer feedback loops."
Think unconventionally when it comes to selecting cloud leaders. Advanced cloud "will require a heavyweight product manager/change agent to drive the program forward. The candidate should probably not be a conventional choice for running a large enterprise IT program, and probably should not be someone who already owns other important programs."