AI fuels nearly 30% increase in IT modernisation spend

There is a clear demand for modernization and tech investment: 59% are worried their organizations’ ability to manage data won’t meet GenAI’s demands without significant investment. With the right approach to this investment, enterprises will be better prepared to overcome productivity challenges and satisfy end users who demand continuously improving experiences.

Enterprises plan to spend on average $35.5 million on IT modernization in 2024. More than a third of that will be on AI, with the average enterprise investing over $21 million on the technology in 2023-24, and $6.7 million on generative AI (GenAI) specifically. The drivers for this are clear: rapidly prototyping and testing new ideas, making employees more efficient, and identifying and capitalizing on new business trends. Yet enterprises recognize there are challenges ahead — from ensuring AI can be used effectively and safely, to having sufficient compute power and data center infrastructure in place.

“Enterprises have entered the AI age, but so far are only scratching the surface,” said Matt McDonough, SVP of product and partners at Couchbase. “Almost every enterprise we surveyed has specific goals to use GenAI in 2024, and if used correctly this technology will be key to managing the challenges facing organizations. From keeping pace with end-user expectations for adaptable applications, to meeting ever-accelerating productivity demands, GenAI-powered applications can provide the agility and productivity enterprises need. Enterprises must be certain that their data architecture can cope with GenAI’s demands, as without high-speed access to accurate, tightly managed data it can easily guide individuals and organizations down the wrong path.”

Key findings include:

• Businesses are unprepared for data demands: 54% do not have all the elements of a data strategy suitable for GenAI in place. Only 18% of enterprises have a vector database that can store, manage and index vector data efficiently. Enabling capabilities such as control over data storage, access and use; the ability to access, share and use data in real time; the ability to use vector search to improve GenAI performance; and a consolidated database infrastructure to prevent applications from accessing multiple versions of data will be critical to building a strategy that meets GenAI’s data demands.

• Reliance on legacy technology is stalling modernization: Despite increased investment in modernization, factors such as a reliance on legacy technology that cannot meet new digital requirements is either causing projects to fail, suffer delays or be scaled back, or be prevented from ever happening. The result is an average $4 million wasted investment per year, and an 18-week delay on strategic projects.

• Targeted spending: Respondents are aware of how investment can help their GenAI capabilities. 73% are increasing investment in AI tools to help developers work more effectively and create new GenAI applications faster, while 65% say edge computing will be critical for enabling new AI applications — by reducing latency and placing data and computing power together.

• The dangers of rushing into AI: 64% of respondents believed most organizations have rushed to adopt GenAI without understanding what’s needed to use it effectively and safely. Worryingly, this may have been achieved by weakening other areas. 26% of enterprises diverted spending from other areas to meet AI objectives — most often from IT support and maintenance, and from security.

• Meeting the productivity challenge: 71% of IT departments are under growing pressure to do more with less. On average, enterprises need to increase productivity by 33% year-on-year simply to remain competitive. This could explain why 98% of respondents have specific goals to use GenAI in 2024.

• Investing in infrastructure: 60% of respondents are worried about whether their organization has sufficient compute power and data center infrastructure to support GenAI, while 61% say their corporate social responsibility and environmental responsibilities mean they cannot fully adopt GenAI unless based on more efficient infrastructure. Some respondents may be unaware of potential solutions — 66% believe they would need to invest in multiple databases to get all necessary capabilities to support GenAI, despite the existence of solutions that support all multipurpose access needs.

• Adaptability is key to meeting end-user demands: 61% of enterprises are under pressure to continually deliver improved experiences for end users, with the average consumer-facing application falling behind expectations in 19 months, and the average employee-facing application in 20. To counteract this, 45% of respondents say adaptability — the ability to change what the application offers the user as needed — will be the most essential attribute for applications.

“Investing in the right data management and infrastructure architecture will help unlock GenAI’s transformative potential,” continued McDonough. “For instance, organizations don’t need vast, complex ‘jack of all trades’ applications to improve productivity and meet expectations, and nor do they need multiple, costly databases to meet their needs. An adaptive application that can use GenAI to enhance a specific end-user experience will be equally effective while also having a much faster time to market. And a modern multipurpose database with all necessary functionalities will help keep architectures and costs as streamlined as possible.”