26 Apr 2024

Unlocking value from the cloud – and how GenAI is changing the game (KPMG Whitepaper)

The following extract is from a new KPMG Whitepaper on Cloud and Gen AI.

In recent years, the cloud – once seen as a disruptor – has become standard practice.

Most businesses have migrated some or all of their technology landscapes to the cloud. Having moved from the ‘crawl’ to the ‘walk’ phase of their cloud journeys, most are now starting to realise the benefits: cost savings, faster innovation and greater agility.

But presence in the cloud is just the first step. The priority now is to stay ahead of the curve. That means to continuing to embrace new technologies on their cloud estates, and getting the maximum value from them. That’s why organisations are forecast to invest 45% of their IT spend in the cloud by 2024.

But there’s a problem: the cloud is maturing too rapidly for most firms to keep up. Cloud service providers (CSPs) – the hyperscalers in particular – are releasing new features faster than ever. And as if the landscape wasn’t moving quickly enough, enter Generative AI.

GenAI is a gamechanger for the cloud – and is a central focus for the hyperscalers as a result. Gartner predicts that by 2026, more than 80% of firms will have used GenAI models or APIs; and/or deployed GenAI-enabled applications in their production environments. That’s an eye-watering acceleration – in 2023, the proportion was just 5%.

Cloud adoption is no longer purely a technology decision; it’s an enterprise-wide strategy. The challenge is to balance speed and value, while enabling continuous transformation and harnessing the power of GenAI.

Driving cloud value: challenges and success factors

If businesses’ priority is to realise value from the cloud, what might be preventing them from doing so? And how do they overcome the barriers?

Challenges

Many organisations still lack a properly defined cloud strategy, and so may be tempted to follow their competitors or industry leaders. But the reality is that there’s no one-size-fits-all solution to optimising the cloud.

A lack of focused strategy can lead to lost or delayed value in three main ways – which together, can erase the cloud’s potential value altogether:

delayed adoption of cloud services
adoption of ‘shadow’ (redundant) cloud services
unrealised use cases, mostly focused on cost reduction, not value creation

This last point is especially important. As organisations invest more in their cloud estates, they must focus on two imperatives: minimising cost and generating value:

Most firms have a cloud cost-management capability, but in many cases, it’s maturing gradually. Many therefore struggle to get their product teams to prioritise cost while delivering functions and features for the business. And there are additional upward pressures in the current climate: inflation, the rising cost of AI workloads, and a growing regulatory burden. These are bringing more volatility to the cloud’s economic model.
A properly defined cloud strategy is vital to the ability to generate value – as it should drive a focus on productivity, automation, speed, agility, and control over how product features are rolled out.

Success factors

The ability to maximise cloud value depends on many aspects of an organisation: its technological landscape and maturity, and pace of innovation. How good it is at revamping existing systems and architecture (brownfield projects), and creating them from scratch (greenfield initiatives).

And crucially, its understanding of whether it’s ready for the cloud in the first place. There’s no framework in which to judge this. It comes down to a combination of the leadership team’s subjective, and objective inputs such as quantitative data.

Start with functional transformation

Don’t try to take a big-bang, all-at-once approach to cloud adoption. Begin with the business function (or functions) that will benefit the most from cloud transformation.

That will have two key advantages. It will narrow down the scope to a manageable level; and provide key learnings for enterprise-wide migration.

Where to start should be informed by your business strategy. For example, if your overarching objective is cost reduction or optimisation, then you might want to start with cloud-enabled HR transformation. In the HR function, cloud technology can reduce overheads, automate manual procedures, and streamline global processes to remove redundancies and overlaps.

From there, broader implementation can be an iterative process, with future waves of investment focused on each next function or group of functions to adopt the cloud.

You’ll also need to think about the right implementation method for each function – for example:

lift-and-shift for low-criticality applications
lift-modify-and-shift for end-of-lifecycle workloads
private cloud configuration for highly sensitive data
SaaS for customer-facing processes
cloud-native and/or microservices-based architecture for some advanced functionalities.

Set accountability for cost control

Without proper planning and implementation, it’s easy to misconfigure and underuse cloud resources. That can lead to poor performance, downtime and security issues – all of which will drive unexpected costs.

Cloud optimisation stems from economies of scale, abstraction and shared services. However, accountability frameworks are vital to ensure that cloud spend is properly tracked and managed. Without a designated individual or team in charge, your spend will quickly mount up

To establish that accountability, we recommend using a Cloud FinOps strategy – which involves:

capturing your cloud costs
allocating them to specific projects, functions and stakeholders via robust charge-back mechanisms
setting up budgets and alerts
regular reporting on cloud spend
reducing cloud waste, so that ineffective services don’t go unnoticed.

Capture synergies between cloud and GenAI

Integrating Gen AI into your cloud environment can transform the return on your investment, by boosting productivity and innovation. As the major CSPs are also the main providers of GenAI services, there are inherent synergies between the two technologies.

Organisations can exploit these synergies by:

Using cloud to support GenAI initiatives. The cloud is an ideal platform for the massive compute, storage and networking demands of GenAI applications, as its scalable architecture can handle complex GenAI models.
Using GenAI to accelerate cloud transformation. GenAI can radically transform functional processes automation and remediation – a capability we can expect huge demand for.