Shifting Priorities in Finops: Optimising Cloud Computing Costs and Managing AI Expenses

As the economic landscape evolves, finops teams are adapting their priorities to reduce waste and manage commitment-based discounts. In 2023, the top concerns revolve around finding ways to reduce cloud computing costs, such as purchasing resources ahead of need. However, the rise of artificial intelligence (AI) and machine learning (ML) is expected to significantly impact finops practices in 2024. Here are some of the shifting priorities in finops, the challenges faced, and the need for optimisation and cost governance. 

Reducing Waste and Optimising Cloud Computing Costs:

In response to economic pressure, companies are increasingly focused on reducing waste and optimising cloud computing costs. This includes strategies like purchasing resources in advance to meet future needs. Compute spending is the most heavily optimised area, but there is still room for improvement in storage, databases, and emerging technologies like AI. While finops systems can account for usage, cost-effective utilisation of cloud resources remains a significant challenge for IT organisations.

The Challenge of Optimisation Saturation:

While optimisation efforts have yielded cost savings, there is a concern that the benefits may diminish as wasted resources are reduced. In 2023, the finops community created a library of optimisation opportunities for major public cloud providers like AWS, Google Cloud, and Microsoft Azure. However, the challenge lies in optimising across the full cloud continuum, including cloud, traditional, edge, and mobile. Companies may need to consider moving processing off public clouds and back to on-premises if it proves more cost-effective.

The Need for Cost Governance and Training:

To address the optimisation challenge, cost governance becomes crucial. Establishing effective cost governance practices will help companies avoid costly mistakes as they undergo significant transformations in cloud computing consumption. However, there is a need for more training and approaches to help finops teams gain a broader view across all systems. Currently, there is a clear focus on public cloud cost savings, but a more significant problem to solve is optimising a range of platforms and ensuring cost-effectiveness across the board.

Improving Forecasting Capabilities:

Another area that requires improvement is finops forecasting capabilities. Finops teams are seeking better features to gain a handle on future costs and adjust spending accordingly. Real-time decision-making enabled by self-service finops reports is highly valued by engineers. By catching cost-saving opportunities before code is even written, companies can solve problems before they arise. This proactive approach is crucial in avoiding mistakes like overprovisioning resources from an infrastructure-as-code-enabled application.

Managing AI Expenses:

While AI and ML hold great promise, the costs associated with generative AI are not yet impacting the majority of finops practices. However, for large cloud spenders, AI is becoming a rapidly increasing source of variable costs that needs to be managed. As organisations with higher overall cloud spend recognise the impact of AI/ML, the need for effective finops practices and tools to ensure value from AI spending will become more apparent.

The shifting priorities in finops reflect the evolving economic landscape and the increasing importance of optimising cloud computing costs and managing AI expenses. By staying proactive and adaptable, finops teams can navigate the changing landscape and ensure cost-effective utilisation of resources.

Damian Corneal Warren Tucker Richard Gott Reggie Kelley Andrew Hogan Kevin Doig Stewart Wilson