We’re going big on the cloud. All in. It’s exciting. It’s futuristic. However, it’s downright difficult. I talk with leaders at other large enterprise organizations, and everyone is grappling with the cloud. Some are grappling better than others.

Our “why” is simple. We can be more efficient, secure, resilient, agile, and sustainable in the cloud. Our cloud technology platform is an enabler to just about every other strategic initiative at the company.

The “how” is not simple. It’s taken a lot of strategic thought. I’d like to start this article by describing the patterns I’ve seen, and I’ll wrap it up with the approach we are taking.

The SaaS-only approach

There aren’t many companies left that can claim that they don’t have any cloud presence. All companies, at a minimum, have plenty of Software-as-a-Service (SaaS) products in their environment. While SaaS offerings continue to grow, most companies will be hard-pressed to empty their data centers with this strategy. There are just too many traditional applications left in the enterprise. To embrace a SaaS-only approach also means a long-term commitment to data centers for everything else.

The digital-only approach

Most companies have some sort of digital initiative that involves creating a new customer-facing digital capability. It’s not uncommon for that workload to run in the cloud. Many in this camp operate in a bimodal model. The digital team has a new cloud platform and operating model, and the rest of the enterprise operates the old-fashioned way.

While this is a common way to start, it only gets you so far. Some may call it digital transformation, but it’s just digital ancillary if it doesn’t transform the core. Sooner or later, the two modes need to be reconciled.

Both the SaaS-only and digital-only approaches are often called “hybrid cloud.” In my experience, this is way too generous of a term. This gives you credit for embracing a little bit of cloud, while simultaneously running 90% of the company the way it has always been run.

The outsource approach

Some leaders really like outsourcing IT. Those same leaders look at the cloud as a way to shift responsibility to vendors and reduce staffing. This approach will yield a small faction of the cloud benefits and all of the historical downsides of IT outsourcing we’ve seen as an industry over the past 20 years. At best, outsourcing is cheap and stable. I’ve never found it to be agile and innovative.

Cloud is many things, but one thing it is not is less work. It takes an awful lot of skilled and talented people to execute a cloud strategy. The work changes, but it doesn’t go away.

The all-in approach

I have first-hand experience with everything I’ve written about above. In many ways, these experiences prepared me for the full-blown strategy that we’re executing now. We’re all in on the cloud. Our data centers’ days are numbered.

That last sentence is an important one. If you don’t design your cloud capability to replace your data centers, then it won’t. A generation ago, Stephen Covey said, “Begin with the end in mind.” That’s still good advice. If you don’t then you’ll just have to do it over again. I’ve talked with many technology leaders who are on their third or fourth attempt.

Additionally, let me offer a few other key tenants of our strategy:

1. Start with a critical workload. Pick a big, important workload, and start there. Many pick an inconsequential workload, but I think that’s a mistake. When you pick something important, it forces you to do it right and do it well. You must design the resiliency right. It must perform. The costs must be calculated because it is material. The organizational capability needs to be built. The security must be solid.
2. Make the cloud more efficient. It can be more efficient, but that’s not a given. We can make it cheaper, but we must do the hard engineering work to achieve that outcome. So far, we’ve been able to achieve that every time.
3. Make the cloud more secure. It can be more secure, but we must be disciplined. Many ask me if the cloud is more or less secure than a data center. Inherently, it’s neither. It’s up to us to make it more secure. I believe the cloud offers more advanced security capabilities than its traditional counterparts, but it’s up to us to leverage them. Security doesn’t come by default along with the cloud bill. My team has a sign hanging on the wall: “Dance Like Nobody’s Watching. Encrypt Like Everyone Is.” – Werner Vogels
4. Make the cloud more available. Inherently, it probably already is, but we need to design our systems assuming it’s not. Any service credit we get will never soften the blow of a sustained outage. Design accordingly.
5. Reinvested efficiencies fuel the migration. Run-rate to run-rate, the cloud is cheaper for us, but it costs something to do the migration work.
6. Train the workforce. We’ve hired for some key roles, but mostly, we’ve developed the cloud skills we need in our existing workforce. Additionally, we invested in organizational change management. The shift to cloud moves peoples’ cheese big time, and they need to be prepared.
7. Partner well. We own this effort, but we aren’t doing it alone. We’ve built deep partnerships with our hyperscalers, independent consultants, and the cloud-native ISV community.
8. The cloud is a product. With modern technology comes modern methods of organizing. We run our cloud platform as a product team, complete with agile and product management practices, delivering iteratively with customer focus. Upstream teams run that way. The platform team needs to be just as agile and progressive.
9. Modernize all the things. It’s possible to run the cloud just like a data center, but we try our hardest to take the opportunity to modernize. We embrace Infrastructure-as-Code and DevOps principles as much as possible.
10. Time is running out. Modernization and migration duration are at odds with each other. On one hand, ideally, we want to fully modernize every workload that moves to the cloud. On the other hand, our data center equipment keeps aging, and we don’t want to spend more than we must to extend it. We must live with this tension and do our best to find a balance.
11. Embrace the full stack. Earlier, I mentioned SaaS and how that’s only a partial solution. We also embrace Platform-as-a-Service, Infrastructure-as-a-Service, and Function-as-a-Service. Ultimately, we look at the cloud as the new application runtime environment.
12. Go multi-cloud with intention. By now, every enterprise is multi-cloud. Most by chance, not by design. We have a specific multi-cloud strategy with a decision framework for landing workloads with the optimal hyperscaler. Ultimately, being multi-cloud is necessary. Yet, it undoubtedly stretches our limited talent capacity thin. Therefore, we must be rational and disciplined with hyperscaler sprawl.

While I’ve written about cloud several times on this blog, this is the first time that I’ve explained our comprehensive strategy. There’s nothing in this strategy that’s brand new. We’ve been operating this way for a while and are well on our way toward achieving our goals.

I thought about writing this article before, I wanted to make sure our strategy was tested with plenty of experience first. We continue to learn. We aren’t the organization we were a year ago. A year from now, we’ll be even smarter. For everyone on the cloud journey, I hope I’ve provided something beneficial for you in this article. Please let me know how it’s going in your organization. I’d love to hear from you in the comments below.

Podcast: Play in new window | Download