Major waves of innovation in delivering enterprise technology, from mainframe to the Internet, IT outsourcing and now the cloud, reset the cost and capability levels that enterprises can achieve. But this reset is inherently difficult to manage precisely because it changes the established ways of doing things.
So, it is no surprise that RightScale’s annual State of Cloud Survey found that managing cloud costs was top of companies’ 2019 priority list for the third year in a row.
Our experience and client feedback suggests that the benefits that cloud is designed to unlock for businesses is by no means guaranteed. So how best can companies exploit cloud for value, whilst remaining cost effective?
Take, for example, a transformation which moves hundreds of workloads from on-premise infrastructure to the cloud. That alone doesn’t change a customer journey: what you do afterwards, with new capabilities on cloud like analytics or automation, does.
Or take migrating your ERP to SaaS: without effective change management, an eye for the user experience and process transformation, we have seen clients experience tens of thousands of process deviations from standard process.
It’s clear that value has to be fought for to be achieved in transformation but businesses rightly see the opportunity as being worth it. The KPMG/Harvey Nash CIO survey’s positive findings also revealed that, as digital and business strategies converge, more UK CIOs (54% up from 49% last year) are seeing budget increases this year than at any other time in the last 15 years.
The first question, therefore, is how to achieve value from cloud investments.
As businesses approach the process of engineering major digital change or transformation, the integration of cloud services is no longer a disruptive practice but a standard expectation. The potential is clear: better end-user experiences, enhanced agility, greater security, lower cost models, breakthrough technologies, and faster time to market.
For example, in the past 12 months alone, we have helped clients use AI to automate reviews of invoices in their supply chains and identify billing errors. We’re actively transitioning our supplier audit business to use this cognitive intelligence and we couldn’t do this without the leaps in AI technology available on the public cloud.
Firstly, our ethos on cloud is to start with a solid understanding of what your business needs from cloud. For some, speed to market is important. Others can make breakthroughs with analytics. Others may want to exploit the utility model of cloud to reduce cost. Clarity on where you are heading dictates so many choices in delivery.
This clarity remains important even for mature cloud enterprises. Does your current usage, achievement and operational capabilities reflect the priorities which make a difference to your strategic goal?
For example, do you spend needlessly high amounts on DevOps tooling which marginally improves speed to deployment, but neglect the data fundamentals needed for your business to exploit the analytics capabilities they need?
Secondly, it’s important to be clear-eyed about your cloud transformation programme. Being realistic about cost and setting (and achieving) a realistic transformation ambition is important. Once migrations commence, cloud transformations are typically managed on throughput rather than quality of transformation. We believe both are important, because throughput demonstrates execution ability and quality of transformation drives value.
Thirdly, the transformation doesn’t end when the cloud programme finishes. We see companies drive most value from cloud when they incrementally improve their cloud estate – transitioning workloads to newer forms of PaaS, or adopting the latest functionality deployed in their SaaS estate.
Allowing for ongoing budget and skills to pursue these improvements over time is imperative – we call it cloud lifecycle management.
Finally, IT and the business must work together to enable business stakeholders to understand the art of the possible. Many companies are putting innovation labs in place – and our own Cloud Value Labs have helped clients redesign customer journeys, transform operations scheduling and prioritise ambulance dispatch, all using their existing cloud estates.
In parallel to securing value through these steps, enterprises constantly need to control cost on cloud. It’s not axiomatic that cloud is cheaper, and our Cloud Cost Management practice consistently finds that clients are spending 20-40% more than they need to due to inefficient use of cloud.
This is reflected in Gartner’s research which in 2018 indicated that wasted cloud spend could be as much as 45%. Whereas in a traditional data centre world, the envelope for cost performance was set in three-yearly negotiations with outsourced providers, cloud opens up thousands of configurable items which influence cost. The prices of these change dynamically, and their use is directed by cloud engineers who may be amongst the most junior in IT.
In 2017 when we started doing Cloud Cost Management reviews, we frequently saw simple utilisation inefficiency – with auto-scaling not used and redundant environments not switched off. As organisations have developed maturity around cloud the latest reviews have found less of this, but find increasing amounts of architecture inefficiency where workloads have not been fully transformed and so are using costly, older capabilities (such as older instance types).
We also see considerable savings in over-specified DevOps capabilities and sub-optimal support models.
Cloud is here and here to stay. According to the CIO Survey results, 51% of UK organisations have performed large-scale cloud implementations, ahead of the global average of 44%. Those enterprises who focus on value, and are most purposeful in resetting their own operating model to enable effective management of cost, are the ones who will drive real competitive value through cloud.
Adrian Bradley, Cloud Advisory Lead, KPMG
