I am frequently confronted with statements that the strategic direction of an organisation is “all in” Public Cloud. Now this is a common blog topic but I feel I need to add my experience to this well covered and contentious point.
Whilst I don’t disagree that Public Cloud will feature in every business, small, medium, large, new, old or even public sector, I do exercise caution when approached by the statement. To be clear, I define Public Cloud as the big 3 hyperscale providers; AWS, Azure and GCP
Let’s look at the facts:
- Public Cloud take up year on year is increasing, but not at the rate the big 3 would like, 18% increase in 2017
- Investment in the Public Cloud platforms exceeds many medium to large enterprises turnover ranging in the 10s of $Bil
- Every industry sector, including heavily regulated ones, are moving workloads to the Public Cloud – Casing point, HSBC’s use of Google Cloud Platform for all of its data analytics
But:
- 2016 was the first on record where spending for on premise infrastructure was exceeded by Public Cloud consumption
- Any business over 10 years old generally has the anchor around the neck known as Legacy Systems, we will explore this in more detail later!
- Perceived security risks still provoke a knee jerk “No!!” without any real substance or understanding of the platforms
- Its not as cheap as you may think!
So why do I care…..?
Many of the clients I speak to are well educated and well informed individuals, with one key responsibility being the strategic direction of the business and aligning IT to this. The pressures of running a P&L and everything it entails leave very little time for exploration of a vast and ever increasing technology landscape. Add to that the number of providers in the space that can or claim to be able to deliver all services under one contract structure. This leads to a minefield of choices and directions that senior strategic thinkers need to navigate. Public Cloud appears to be a utopian end point for the innovative and forward thinking organisations, and in some cases, it absolutely is, but…. It needs to be used appropriately. This is where I will explain what I mean by the term “Legacy”:
IT has always been susceptible to shifting sands of technology and trends, from the centralisation of Main Frames to the explosion of on premise server sprawl and back to the centralised model of pay per use cloud based services. Coupled with this has been the strategic direction for adopting software applications that best suit your business models. Whether these be commercial platforms, logistic platforms, compliance platforms, they have all been scrutinised against how best to consume them. The options being, in-house developed, COTS, COTS with heavy customisation or SaaS. With the availability of appropriate SaaS or COTS for many businesses being light back 10 years ago , the adoption of in-house development was high. The term legacy relates to organisations that have followed the trends at the time but not been able to innovate as the technology and services available have evolved. Hence being left with unwieldy, complex and expensive systems that are difficult to refactor.
I love Public Cloud…. and for the same reasons most of my clients do;
- Massive scale
- Flexibility
- Endless features
- Pay as you go
- Self-service portals
- Managed services
But unless these attributes are required by the workload that will be consuming them, then it’s not necessarily the best platform and certainly not the cheapest! For example; an e-commerce platform that is susceptible to surges of user activity will benefit from the burst and shrink capabilities within Public Cloud but an ERP solution that has a steady resource usage throughout the month and cannot utilise the Public Cloud managed services will not be cost effective.
I suppose my message here is simple; “listen to your application, it will tell you where it will work best and most efficiently. Cloud economics is a key tool when defining an applications best execution venue”.