On-demand technologies: the numbers don't add up.

Increasing the ability to innovate, improving operational agility, and strengthening competitiveness: this is a dogma for millions of companies engaged in digital transformation, who see public cloud services, Software as a Service (SaaS) applications, and generative AI tools as the go-to technologies to achieve this goal. But while the benefits of adopting these solutions are clear, a recent study by Capgemini Research Institute ("The On-Demand tech paradox: balancing speed and spend") also reveals a dark side, including poor cost visibility, management complexity, underutilized resources, and governance gaps. All these factors, combined with a mindset still rooted in the past, risk compromising the scope of these benefits.

According to the report previewed by Il Sole24ore, companies are progressively shifting from capital-intensive investments to more flexible, consumption-based models, fueling a trend that has seen the share of on-demand technologies in IT budgets grow from 29% to 41% in just twelve months. More specifically, 82% of the approximately 1,000 executives of very large companies interviewed globally reported a significant increase in spending on these solutions, 76% of companies exceeded their public cloud budgets (by an average of 10%), and 68% overspent on Gen AI, paying the price for underutilized resources and a lack of centralized purchasing. Finally, only 29% of companies achieved expected cost savings from SaaS, and only 38% actually accelerated innovation with generative AI tools.

"These technologies," observes Ernesto De Ruggiero, Managing Director of Cloud Infrastructure Services at Capgemini in Italy, "have transformed the way companies operate, but they have inevitable economic implications, with spending set to double in the next three to four years." This clearly summarizes a scenario that requires CEOs and CIOs to make important decisions. In this scenario, those companies capable of aligning their cloud migration strategy with business objectives (thanks to more mature and advanced FinOps practices, which are currently still underdeveloped) and designing scalable and modular architectures, leveraging AI-based automation to optimize processes and reduce inefficiencies, will be the ones best positioned to generate value from their investments. The majority of managers surveyed (77% to be precise) consider the scalability and performance guaranteed by cloud computing to be crucial for growth, as they allow for accelerating transformation and reducing time to market, but there are many cases of companies that find it difficult to keep the management and operating costs of these solutions under control.

There's thus a sort of paradox looming over companies, linked to the risk of not seeing returns from on-demand technologies. To overcome it, according to De Ruggiero, there's essentially only one way forward: "a holistic approach in which strategy, solution design, corporate culture, governance, tools, and operational processes are aligned and communicate with each other coherently." There's no shortage of companies (even in Italy) that have maximized the benefits of adopting SaaS and cloud-native solutions by following this path. One example comes from the manufacturing sector, where Capgemini and Google Cloud implemented a Gen AI solution (comprising a chatbot and a forecasting engine) for a large multinational corporation. The solution provides real-time responses and accurate demand forecasts, addressing ineffective order management in warehouses and creating the conditions for inventory optimization and cost reduction. Another successful example is National Gas, the UK's leading gas network operator, which has moved many of its processes to a fully automated cloud environment on Microsoft Azure. In general, the Capgemini manager concluded, “we need to go beyond simple technology implementation and develop a data-driven strategy supported by IT investments based on solid business cases and clear, shared metrics, as well as integrated cost planning from the initial design stages.”