For 2025, technology architecture leaders are seeing some glimmers of hope in the wider economy, which is bringing some cautious optimism. Inflation is down from its record highs, and while cost pressures, economic volatility, and geopolitical risk remain, tech leaders are planning to focus investment on the tech infrastructure, software, and data assets required to realize the potential of AI and generative AI investments. To offset some of the financial and other impacts, such as increased carbon emissions associated with AI infrastructure and services usage, organizations are closely looking at how they can make the most of their investments to gain the benefits of AI while optimizing their spending.
To help technology architecture leaders assess their spending across the board and deliver maximum business value, our new Budget Planning Guide 2025: Technology Architecture And Delivery provides guidance on where to increase and defend investments, as well as where to decrease investment — and where to experiment with emerging technologies in 2025.
Increase Or Defend: Cloud And AI
Overall, cloud and AI will both be key areas of investment in 2025. Organizations are investing even more heavily in cloud, both for AI infrastructure and services support, along with the migration of more of their workloads to the cloud. Into next year, tech architecture and delivery leaders will note a continued emphasis on AI, with some more specific recommended AI use cases that will require budget. With this in mind, we recommend that tech leaders increase or defend investments in these areas for 2025:
- Optimization practices to deploy workloads in the optimum cloud and price point
- TuringBots to build software faster, better, and cheaper
- AI agents to drive employee efficiency and productivity
- VMware alternatives to blunt the impact of massive cost increases
- Specialized large language models (LLMs) and infrastructure for better accuracy and reduced costs
- API developer portals to facilitate access to AI services
Decrease Or Avoid: Routine Enterprise Software Apps
While tech budgets will increase in 2025, the only way that tech architecture leaders will be able to free up enough budget for new investment and experimentation is by letting go of certain investments that no longer make sense for their businesses. For many tech leaders, single-function enterprise apps or zombie applications resulting from long, unloved acquisitions from enterprise software mega-vendors will be what gets dropped in 2025. Specifically, we recommend that tech leaders decrease or avoid investments in these areas:
- Single-function enterprise business applications or zombie products
- Stand-alone hybrid cloud management solutions
- IT service desk unsupported by AI
Key Emerging Technologies Worth Experimentation
Tech leaders are keeping a close watch on emerging technologies that are at early stages of maturity but could positively disrupt the business in the future. For many tech organizations, there have been some budgetary lessons learned from the rush to generative AI in the past two years. While economic conditions are a little less constrained heading into 2025 than they were in 2023 and 2024, tech leaders and chief technology officers should still consider closely the cost and budget implications of the emerging technology that they choose to pilot next year. In the coming fiscal budget cycle, technology leaders should consider evaluating these technologies (and allocating budget to fund proofs of concept) to determine organizational fit:
- AI PCs capable of running LLMs locally
- AI cost management to control costs before they spiral
- Edge intelligence to capture actionable insight from mobile and IoT devices
Learn More
To get more details on where you should invest, divest, or continue experimentation, download our Budget Planning Guide 2025: Technology Architecture And Delivery. And be sure to register for our upcoming webinar on September 24, where we’ll dive deep into all of these technology areas and answer your questions live.
Source link
lol
