Click to learn more about author Patrick Hubbard.

Many businesses have pushed their cloud migration plans into overdrive in 2020. The emergency transition to work from home, significant changes in customer behavior, supply chain challenges, and existing pent-up digital transformation projects created a rapid-migration Voltron for many teams.

There’s been amazing delivery on the part of IT professionals and business leadership, which has accelerated the charge for all-cloud (or at least mostly-cloud) operations. But even with well-earned confidence in the IT team, the question remains: should your business go all-in on cloud now, eventually, or ever?

Cloud has been promoted as a cure-all for everything from cost management to performance, flexibility, durability, and simplicity. And for many organizations, the correct combination of these benefits can substantially improve the technology services they care about most. However, there are many caveats to achieve this vision. Here are four cloud adoption considerations for decisions you’re making now and how they may affect your organization’s “Next Normal.”

1. Know Why You’re Really Going to Cloud

Cloud is essentially a form of MSP hyper-outsourcing based on API calls instead of meetings and phone calls (hopefully). You’ll need to ensure leadership budgets are reasonable. Managing different access methods, applications, and services will require new training and support. Humans need a reason to accept change, and you’ll be exposing numerous changes to potentially everyone in your organization.

We’ve all had a moment where we needed to distill a complex project justification into human-readable form and struggled a bit to articulate it for civilians. When you can explain your team’s technical goal in terms everyone understands, it’s much easier to keep everyone on the same page and streamline execution. 2020 offered plenty of examples: “We’re transitioning everyone to Office 365^® this week to support work from home and save the business.” If you can get to equally clear explanations of where cloud will meet the business’s needs, it’s easier to achieve consensus and approval.

2. Have a Security Plan, No Exceptions

The cloud relieves many infrastructure headaches, but there’s a catch: potential access. If you make a resource available to everyone, it means everyone on Earth. With on-prem, we have a habit of starting with zero-trust access policies and then adding exceptions as needed via command line. But this natural security gatekeeping function is distributed with cloud, where IAM policies are defined by application owners rather than network engineers.

Remember, cloud security offers an advantage: it’s designed for policy-first operations. Security—along with resources, scalability, monitoring, and other details—is configured as part of the application, not assigned afterward. Make sure your team has time to learn and practice security configuration before you migrate data.

3. Invest in Skills Development

Though vendors don’t usually say it out loud, cloud is different and evolves far more quickly than on-prem tech. Just a couple of years ago, a Docker^® certification might have been a career advantage, but before mainstream IT could adopt it, Kubernetes ate Docker (and Mesos, too). This rate of change makes it difficult for managers to bolster the team with experts, and many have realized they’ll have to ramp up and maintain ongoing training for everyone working with cloud resources.

Surveys can be a good reference when planning a training budget. According to the 2020 IT Trends Report, cloud computing skills are among the top three technologies influencing organizations’ staffing needs. You may have already discovered this; perhaps you even chose to stay on-prem simply because you can get support for OpEx services but not team development.

4. Zoom Out Your Instrumentation

Regardless of what’s driving your cloud adoption, your infrastructure, applications, platforms, and user interactions will change significantly. Ensuring users are blissfully unaware of plumbing is only possible if you adapt and extend your performance monitoring to all the new elements. You’ll need to compare the performance and error rates of your apps before, during, and after migration. In addition to quality assurance and troubleshooting, leadership is usually more eager to follow IT cost optimization recommendations when you can prove ROI based on hard data. It can also help prevent a career-limiting OpEx bill, a hazard for anyone new to cloud and in a rush.

Instrumentation reveals how users are behaving in the real world. Does access occur completely outside the firewall, or are they proxied through another application, VPN, or software as a service (SaaS)? This distinction is useful for planning but critical for identifying service bottlenecks. Cloud providers are happy to let you scale any and all services to try to solve performance issues. The ability to pinpoint exactly what’s causing an issue will help you scale only what you need and avoid unnecessary costs.

Extra Credit: Consider Using a Single Provider

Nobody likes lock-in. Cloud, however, is based on lock-in. Look no further than its Hotel California-style data transport. You can upload your data any time you like, but you’ll pay every time it leaves. More than rubbing IT pros the wrong way, lock-in fears can stall cloud migration projects. It’s a reasonable concern.

However, multi-cloud (as in freely portable workloads and data you can easily shuttle between providers) can get expensive or limited in a hurry. This is primarily because each provider seeks to differentiate itself with proprietary services that don’t easily and universally translate. This means more team training because each cloud’s tools and processes are different. It means more investment in multi-cloud monitoring and management tools to provide a single, composite view of operations. And it also means a larger attack surface requiring more security resources.

With cloud—as with any MSP—you’re likely to see reduced complexity, cost, and time to deployment with a single provider. There’s no such thing as “just a little” cloud—the startup costs and learning curve are essentially the same regardless of the scale of your deployment.

You could be with AWS^® and could be a development-focused, early adopter. Or perhaps you’re primarily a Microsoft^® shop, and your experience with Microsoft technologies and their partner ecosystem makes it easy to incorporate Azure^® as extensions to your operations, not re-implementations. Either way, consider standardizing around the cloud you know best.

Keep It Real

Above all, be honest with yourself as a business about the full benefits and costs of cloud, including not only the services but the training and support you’ll need for your team. Seek peers who’ve operated in different scenarios and ask them to be candid about their experiences. Finally, be relentless in your efforts to eliminate complexity, or you’ll end up in a hybrid operations hair ball.

Fortunately, even under the tight deadlines of 2020, you have time to make great choices in your adoption of cloud. Your team has likely been discussing it for years at this point, and you’ve probably identified lots of options. Best of all, the urgency and clarity 2020 has demanded have opened new and more efficient routes of communications between the businesses needing to transform and the IT pros who’ll make it happen. That’s handy. Great cloud experiences always come down more to the people who make it go than the technologies they’re made of.