When you were a kid, did people always tell you to “get your head out of the clouds?” How badly did you want to tell them off? Well now you can, because the cloud is where it’s at in today’s business and IT world. The underlying technology has matured to the point where cloud migrations are becoming standard operating procedure across industries. Whether you’re looking at a containerization project for your development teams, edge computing to speed up your data scientists’ algorithms, or something as seemingly mundane as moving file storage to the cloud for easier access from remote sites—it’s cloud all the way, right?
Slow down friend—while we’ll be first in line to tout the benefits of a cloud implementation, we’ll always add a warning to “proceed with all due caution.” There are rocks hidden under the smoothly flowing river of technology and we don’t want to see your project hit them and have to be abandoned mid-stream. The skills necessary for a successful cloud deployment are varied and less common than some seem to think. Add to that the specialized tools you may or may not have at your disposal and the time it takes to plan out and implement each stage of a migration and you start to see where these projects can go sideways, fast.
There are myriad reasons a cloud project might fail: lack of a plan for refactoring and re-architecting, lack of familiarity with necessary tools, or simply lacking the right skill set on the deployment team. By moving iteratively, ensuring the team has the right set of talented people, and ensuring they have the tools they need, you can make a strong push to avoid this outcome. In an effort to make sure you have all the information you need before proceeding, we offer our pick of 5 stats that every CIO needs to know before launching your next cloud initiative.
1) 73% of Decision Makers Say Optimization of Existing Deployments Is Priority #1
According to Flexera’s 2020 State of The Cloud Report, this is the percentage of CIOs and other decision-makers in the IT space who say that their first priority this year will be fixing and optimizing existing cloud systems. These leaders have already overseen cloud deployments and migrations, and what they saw they didn’t necessarily like. Maybe the project went over budget, yet was still underwhelming performance-wise. Or maybe it was hashed up to the point where they put a stop to the whole thing so it could be reimagined entirely.
Whatever the specifics of each situation, the decision-makers involved are dedicating the time, money, and other resources to optimizing their existing cloud implementations before starting down the road to additional deployments for a reason. They don’t want their future projects to get derailed, and they understand that they need to take this step back, fix the issues, and reassess their path forward.
The number of cases represented by this majority serves to highlight the fact that many cloud projects go off the rails and require rework to get back on track. Whether bringing in skilled consultants is required, or reskilling existing staff to better take on the coding and scripting aspects of a cloud migration (i.e. the skills that often get overlooked during planning), the outcome is the same—expensive rework and the lost revenue it entails.
2) ½ of All Cloud Migrations Will See Delays Due to Lack of Skills
This one comes to us from industry leader Gartner, who reports that fully half of all cloud migration projects started this year will end up delayed, many for as long as 2 years. These delays will be due to a simple lack of required skills on the migration team. A successful cloud ecosystem requires a combination of traditional infrastructure skills, specialized knowledge of the host system being used, and a wide assortment of programming languages. People with these skills are in high demand, leading to prolonged job searches when they need to be hired on, as most are happily employed.
The most sought-after languages come into play here, Python, Perl, and Ruby lead the charge while some standbys like Java and .NET are also still in widespread use. On top of that, DevOps need a strong basis in Linux administration (something operations may already have, depending on your current infrastructure). Round out the package with database skills in SQL and/or MongoDB and you can see why these folks are in such high demand right now.
Enter reskilling. You have an IT operations team with the skills and knowledge base required to keep your on-prem systems running smoothly, right? These folks can learn the additional skills needed for cloud projects, and many will jump at the chance to further their careers by adding to their quiver of skills like this. Partner with a service provider or consulting firm to bring in a coach/trainer to work with these people and assign them a mentor from development to answer questions from the trenches. Or, bring in an entire team of cloud specialists—these folks can get up to speed rapidly and implement the correct tactics right off the bat. An added bonus of doing things this way is that your outsourced team is likely to have seen a host of different deployment scenarios in a variety of different environments (even the best in-house specialists don’t have this kind of breadth of experience), meaning that when challenges do arise they’re poised to tackle them with a steady hand.
3) 72% of CIOs Say They Prefer a Hybrid-First Approach
The Everest Group conducted a survey of CIOs, 72% of whom say they prefer a hybrid cloud to either exclusively private or single-vendor options. A hybrid-first approach to cloud deployments means that some services may end up on one public cloud host, say AWS, while another system may stay on a private cloud for the time being. Additionally, when the time comes to move that system to the public cloud, it very well may end up on Azure rather than AWS to avoid vendor lock-in.
Vendor lock-in is what it’s called when you have all your eggs in one basket, as it were. This is a major concern for CIOs as it can be prohibitively expensive to move databases once they’re created in one ecosystem. Extensive vendor lock-in is also what often leads to monopoly situations and the inevitable antitrust suits that follow. Hybrid models allow for slower, more intentional, and iterative migrations. When you can set up new systems on the cloud while leaving mission-critical ones on-prem or on your private cloud, for the time being, you go a long way toward eliminating those costly delays discussed in item 3.
4) 63% of CIOs Say DevOps Is the Way to Go for Cloud Success
Avenga conducted a poll of CIOs, asking for their input on all things cloud. Experiences ran the full gamut from great success to dismal failure, with the common thread being lessons learned. A majority of the respondents had rave things to say about the creation of DevOps teams. They highlighted the better relationships these teams helped foster between different branches of IT as well as their broader organizations and how that aided in speedy adoption of cloud technology.
There are different opinions on the best way to implement DevOps, some of which involve hiring entirely new people who have the know-how and experience for the unique mix tools that this function requires. Job searches are costly and time-consuming, while a few months with a coach or mentor can bring your ops team up to speed (many will already have a background that includes some amount of code). By the same token, you can bring in a team of experts with the express purpose of putting the necessary systems in place to power DevOps, CI/CD, etc. in order to keep up with the competition without breaking the bank.
5) 3.9 Million USD Is the Average Cost of a Data Breach
IBM runs the numbers every year, and for 2019 the average cost of a data breach came out to a staggering 3.9 million US Dollars. There are, of course, myriad causes of data breaches. Some of the most prevalent when it comes to the cloud are human error, undiscovered backdoors, and lagging security standards on shared cloud residencies. Starting a cloud project without the right, highly-trained and skilled staff on the team opens the project to the first of those, human error. Not setting up your cloud infrastructure correctly from the beginning is asking for problems to crop up, potentially years later when those employees may not even still be on staff.
Though the modern-day cloud can offer much more robust data security protections than the on-site servers of old, misconfigurations are still a serious source of vulnerabilities. This means that you could get to the end of a cloud migration believing that your systems are secure, when in reality your security features are configured incorrectly—giving hackers or fraudsters the window they need to wreak havoc on your data. Given the state of data in 2020, it’s not worth the risk to your organization’s reputation or bottom line to jump in too fast and risk bungling something this important.
Given that 68% of CIOs report data security as their overarching #1 concern with cloud migration projects (bonus stat for you), it’s no wonder there are so many extra factors to take into account. Slow down. Take the necessary time to do your research and get the capacity and expertise you need to manage cloud migrations in a secure and cost-effective way. And if you already have infrastructure products running in the cloud, audit them to discover how well they’re truly serving your organization’s needs and optimize where necessary, again before jumping into a new project. When in doubt, find a service provider to partner with. Someone with the experience in the cloud to know which products work and which don’t, and which solutions to employ for which challenges and pitfalls. Your company’s data is worth it.