CFO’s must think like rocket scientists to avoid having their heads in a cloud
Herman Singh shares the six key stages CFOs and organisations need to follow on their cloud journey.
We often fall for the hype of cloud, thinking that speed trumps planning. However, statistics have shown that between 60 and 70 percent of all cloud implementations end in failure – either the project overran on time or costs, or the benefits were not achieved.
Over the last decade, I have sat on dozens of boards of local and offshore firms where part of my role has been to provide technology insight, support, and oversight on a number of ventures, including cloud transformation.
These years of experience as a user of cloud has shown me how CFOs can ensure their firms don’t become part of the statistics for failed cloud implementations.
Think like a rocket scientist
Just as with rocket science, when implementing Cloud you first need to know where you are going and why, what the options for getting there are, what risks and costs this entails and how to mitigate them, and how to operationally achieve the outcomes that you seek.
There are six key stages in the journey to a cloud-based firm that, like a hypothetical six-stage rocket, comes with a lot of caveats and risks. It’s very important that CFOs understand their role in this and how to manage and mitigate the risks.
1. Understanding what cloud is and what benefits it conveys
Cloud computing refers to the availability of computer system resources on demand, without direct active management by the user. It refers to data centres available to many users over the internet and often has functions distributed over multiple locations from central servers.
The main reason for using cloud computing is to move from fixed costs and technologies of expensive own data centres, to a model where you have variable cost – with the vendor carrying the costs of upgrades and operation, and you only pay as you grow. You are also able to scale this seamlessly and rapidly as the business needs evolve.
Sometimes rockets blow up on the launch pad. This stage one failure can occur when CFOs’ have little knowledge of and experience with cloud.
2. Figuring out what your strategic objectives for using it are
Recent research has shown that potential cloud users are relatively evenly split between those looking for lower costs, those looking for increased speed to market for new products and services, or those looking for increased capacity on demand.
Linking this technology journey to the business need is key. It is a strategic journey to build a new capacity. Not being clear on why you are doing this and what your definition of victory is will result in compromised decision-making further down the road.
3. Assessing what type of architecture your cloud solution will have
When implementing Cloud, CFOs need to think about their cloud architecture. This means thinking about whether you are going to have one cloud or many, whether you will be in one country or many, where the data will sit given data sovereignty laws, what the impact on system speed will be, whether you will run your own cloud solution or use someone else’s, and more.
There are various different clouds that businesses can use, including Google Cloud, Amazon Cloud, and Microsoft Azure. There are also various software vendors that you can choose from that provide cloud services, like Salesforce Cloud, SAP Cloud and Microsoft Cloud.
If you choose infrastructure as a service (IaaS), it is only the infrastructure that is in the cloud, that is the servers and databases. If you choose platform as a service (PaaS), you are also sending the operating system, middleware and the run time environment into the cloud. If you choose software as a service (SaaS), everything is in the cloud, including the data and the software applications themselves.
4. Working out the costs and business case
Establishing the cost and the business case can be complex as some benefits are soft and some are hard, some are quantifiable and some are not. For example; the ability to be agile in new product releases is strategic, but tough to quantify.
The cost of today’s implementation needs to be established in detail, but this can involve dozens of cost elements, which are often left out. The “to be” costs are included in vendor submissions, but are almost guaranteed to be incomplete.
Establishing total cost of ownership is key and will include aspects like increasing network costs as you move software and data to other locations. You then need to connect them, which means you need to spend more on ongoing communications costs, but also increased latency or system delays.
There have been hundreds of cloud implementations that have had to be reversed as the costs of the migration materialised, so it is important to complete a deeper assessment early.
There are different options to consider for a migration journey, all with differing costs, timelines and risk.
If you go for a SaaS solution, you don’t need to create a migration path as the solution is cloud native. You have five options for existing code: retaining, re-architecting, remediating, re-platforming and rehosting.
Retaining means that the application is not suitable for the cloud and will continue to be run onsite in your own facilities, so basically, do nothing.
Rehosting means that it's easy to lift and shift the application with the same operating system into the cloud with a minimum of change.
These are both relatively low risk options but the next three are less so, as they involve optimising existing solutions for the cloud.
Re-platforming means that you need to make platform and service upgrades involving operating systems changes and new software frameworks.
Re-architecting involves completely rewriting the components of the application; this is also known as cloud native development.
Remediation is a combination of lift and shift and re-architecting with some smaller code remediations.
Choosing a path per application, and there could be hundreds, is quite a challenging task, and often overlooked when costs or schedules are calculated.
6. Executing and operating the end solution
Stage six is effectively mission control, when you execute the plan and operate the solution. This is heavily dependent on the skills and experience of your own staff and that of the solution partners that you employ.
The ability to migrate in a phased manner without losing control or bringing down the entire company is crucial for the survival of the CFO and the CIO. You will need new skills and tools for system management and control, as well as operations such as change control and fault finding and remediating. Some of this can be automated but that brings with it other complications.
By following these six steps, many organisations have walked the coals and emerged stronger on the other side as a cloud-enabled business, enjoying many of the promised benefits.
All firms will need to walk this path at some point in their existence, and it is crucial that they do so in a responsible and resilient manner.