Cloud Transformation and the Dilemmas of Cios

Trend number one for cloud adoption

One of the meta trends of cloud adoption is that it is almost always… overestimated. We have observed this for at least 10+ years.

Year 2020 seemes to be the exception because of the rapid growth of remote work and the immediacy of allowing people to work remotely. So, there was a huge increase in Microsoft Teams, Zoom and Google Workspace licenses and a multitude of other ready to use cloud based collaboration tools. But, SaaS is just one area of the cloud adoption.

This phenomenon of rapid collaboration tools adoption is overshadowing slower cloud transformations in the area of business applications and data.

Let’s review the most common points of views and fears about the public cloud, and let’s take a look at different decision makers’ attitudes and postures. Then, it will be much easier to understand the meta trend. By the way, the predictions are even more optimistic this year about the growth and I hope it’s time to say goodbye to the metatrend and accelerate the transformation.

Everything from now on is based on my official and unofficial conversations with CIOs and CTOs about their cloud journey, and their opinions from the last few years. When there’s a lot of conversation, certain patterns emerge.  Let me share their typical attitudes in relation to cloud journey and adoption, and then Avenga Labs will attempt to predict the future.

What is cloud? Somebody else’s computer?

The cloud is just an infrastructure, so is it somebody else’s computer? Of course NOT, but this kind of mentality still exists among decision makers that perceive cloud computing as just a set of additional hardware boxes and resources. And yes, sometimes it’s easier to reach for another person’s computer than wait for the next server to arrive; preferably until it arrives and shuts down the cloud VM.

If it had happened ten years ago that would not be of any surprise, but the perception is still here that the cloud is just hardware boxes that can be rented.

It’s very easy to justify the value of Office 365 for instance, or any other collaboration suite. It’s more complex when it comes to platform services which are used mainly by developers and DevOps who use them as bricks to build advanced business applications.

The public cloud (AWS, Azure, GCP) is a set of thousands of services, both technical and higher level, and there are entire suites for data, workflows, and even business domains. Neglecting higher level cloud services is still all too common. This is a huge missed opportunity to speed up the delivery of functionalities to both employees and business partners starting from readily available integrated cloud solutions and services.

It’s not the same public cloud as it used to be ten or even two years ago, as both the availability and quality of the services are improving rapidly and delivering much more business value than ever before. This is noticed by the majority of CIOs as an undeniable fact which greatly reduces the risk of vendor lock-in, at least for many cases the risk of lock-in is simply worth it. (We’ll come back to vendor lock later).

Of course clouds are not a panacea for everything, as they differ in capabilities and the quality of functionality they deliver. But again, they are there to be used and to be combined together to create digital solutions much faster, and to be based on already existing capabilities proven in action for millions of other clients, which prevents a lot of risks and avoids having to deal with infrastructure at all.

We have our own cloud

“So many investments and so many struggles, but what a nice local infrastructure we’ve got here. We love it, we take care of it, it’s ours, and I have a feeling of full control over it. From the electrical power and  cooling, to the colors and layouts inside the room. How could it now be moved to the cloud? I’ve got a better private cloud here and I can deliver higher uptime and lower latency, and in case something breaks, our team will go and fix it.” If you hear this kind of sentence, you do not need to wonder why the predictions about cloud transformation maturity were again exaggerated.

And let’s state the obvious: it’s hard to justify the new investment in cloud opportunities when the previous program of building a local alternative was just finished and it depleted the IT budget.

Still, CIOs who started as infrastructure directors will continue to be focused on the infrastructure until they move on to digital programmes or … someone else replaces them. Of course, there might be cases when a cloud transformation or cloud migration is not the best option, however the attitude of the CIO is even more important as it’s always easier to find problems, risks, and fears, while it is much harder to plan and execute a transformation successfully.

What is important to mention here is that these local clouds are more and more often built as cloud-native solutions which may open the door to migration in the future to cloud native services in the public clouds.

Vendor lock-in or … meet your new friends

Some CIOs feel the inequality of power between their organizations and the powerful major cloud service providers. Even if their organizations are large, they often feel just like another client in a crowd of tens of thousands of others.

It makes them more anxious and careful about choosing a cloud service provider. Sometimes they prefer to delay the cloud transformation or they choose a smaller local provider of cloud services, as they believe the smaller partner will pay more attention to their needs, they will be able to negotiate better pricing and they will have access to more tailored services.

On the other hand, more and more CIOs (reluctantly) have learned to let go as the vendor lock problem is not the priority. They admit they don’t like it, but the benefits of the cloud, especially the ability to deliver business solutions faster, has made many of them ready to use components because of the improved support for the hybrid cloud and scalability. This outweighs the inconvenience of having reduced control over the infrastructure. Plus none of them are alone, as there are millions of others and cloud providers cannot afford to lose clients.

CIOs also quoted the increasing prices of the services and there is no guarantee on how long the current prices will last.  Additionally, there is the lack of flexibility, which puts everyone in the same pricing calculator in a time when products are supposed to be more and more personalized. Besides, prices always rise and they have no other choice than to pay more.

This is something that cloud providers definitely need to address sooner than later.

It is uncomfortably true that cloud providers form an oligopoly which is not the most efficient economical construct from the client perspective. On the other hand, the competition between them is tough and none of the leading cloud providers can afford to give CIOs the motivation to move their services to a competitor.

Kubernetes as the runtime of the 2020s

The trend of so-called cloud native means in reality using the Kubernetes ecosystem to build cloud agnostic solutions.  They can be (relatively) easily deployed to clusters run by local infrastructure teams and in all major clouds.

There’s even the opinion that Kubernetes is the true operating system of the 2020s, replacing bare Linux as the preferred runtime to run business applications and services.

Kubernetes has matured and now has much better support for persistent storage, monitoring, security,  as well as scalability up and down (down used to be the achilles heel of K8s until recently).

New features are planned in the near future, such as support for VMs and other OS-es other than Linux (Windows Server).

All the major cloud providers support this and there’s an entire ecosystem of tooling around Kubernetes.

The problem is that it requires legacy apps to be modernized and repackaged, and that’s possible as viable options for the majority of the applications have been built in the last ten years. There are already architectural patterns established for such transformations with known pros and cons that help to select the right cloud for a particular business and IT context.

The known disadvantage of such an approach is the lowest common denominator problem, which means using the relatively basic services of a cloud based Kubernetes orchestrator instead of a higher level PaaS services. But often, the main success criteria is the ease of movement between different clouds and on-prem, so there’s no surprise that cloud native has many dedicated fans and supporters.

Cloud-first business solution builders

This is totally a cloud first mentality, usually with a very strong belief in the capabilities and readiness of any of the available cloud services. First check if the given capability is in the cloud and reuse it, preferably in a SaaS model, if not in a PaaS as ready to use API.

Instead of asking the question “how do we prepare to move back from the cloud to a local infrastructure”, CIOs are focused on the map of available resources and services in the cloud, and how to combine them together in a viable solution for the business.

It’s getting harder and harder for digital companies to compete with these products, because they are almost the same but with the lowest prices and the highest availability possible. The area where they compete is with Customer Experience. In order to deliver great experiences faster than the competition, we need something better than traditional software development. The cloud used in a smart way may be a great accelerator to deliver those experiences faster than in traditional methods. Creating MVPs and testing ideas on different groups of customers is also much faster and safer than traditional methods.

Sometimes there’s virtually no other viable choice in sight, which may be a little bit annoying, but still, it’s the best choice for IT and business to do it in the public cloud using ready to use APIs and services.

→ Explore Feature toggles – faster digitalization by smart experimentation

There are more and more CIOs who just don’t want to deal with hardware anymore and are switching their attention to supporting the business in their digitization efforts and they simply don’t want to be distracted by local infrastructure. They usually combine multiple SaaS solutions and build API-based solutions using cloud services. Of course, the integration and orchestration of multiple services is not easy, but the time to market is excellent, plus we save a lot of time with requirements discussions and dilemmas as the ‘good enough’ solution available tomorrow is much better than the perfect solution available in two years (which probably means it will be irrelevant as everything changes too fast).

The new idea is that (whenever possible) companies should forget the idea of local infrastructures and embrace cloud services as bricks for faster digitalization as soon as possible. They just have to decide that they can let go of some of their control in order to obtain better flexibility, and then move there. For the majority of the businesses, the local infrastructure is not needed at all. This is proof that the startups and newborn companies that started with the cloud never have to leave the cloud. Obviously, there will be hybrid clouds but they are not supposed to last forever and can be a short lived stop gap solution in the majority of cases (especially in unregulated sectors of economy).

IoE / IoT companies

These types of companies often already use edge computing with or without public cloud support. Edge computing seems to be much less mature than the public cloud and there are many solution vendors and technologies which won’t last even a few years. The cloud had the same situation a few years ago, when companies such as HP, VMWare and others were losing their cloud wars and they became irrelevant as public cloud providers. It’s the natural phase of maturation.

Additionally, they require very low latency and availability with the nearest processing unit on the edge, which cannot be replaced even by the best cloud services. In their case, it’s natural to focus on devices, edge and later, in the cloud as the supporting platform, not the primary concern. Plus let me remind you, all the major public cloud providers also offer great edge solutions which can be easily combined with their public clouds.

Changing landscape. Cloud climate is warming

Even in the last two years, the temperature has definitely been rising and there’s more warm feelings towards the public cloud. In this case, cloud climate warming is a good thing.

Again, cloud solutions and services continue to improve dramatically and their offering is more sensitive to the needs of CIOs who are now embracing the inevitable hybrid environment. Cloud services are not pushing too hard or aggressively, and they are even admitting that there are other clouds and how you might combine them together efficiently. The usual fears of data protection and vendor lock-in will stay with us for a little bit longer, but in most cases they are overshadowed by the benefits. And, there are more and more examples of successful cloud transformations with positive business impact, which is the best encouragement for those who may see more risks than the others. Market research shows a success rate above 85% for cloud transformation programmes.

What we’ve noticed recently is that virtually everybody moves to the cloud, sometimes with enthusiasm and a smile, and sometimes reluctantly and as slow as they can while admitting that another business critical system is now running in the cloud and they plan to continue the transformation. The simple cloud lift and shift is slowly becoming an obsolete pattern, as we predicted, but it still can be used as an emergency solution.

This is much easier to do in startups and in unregulated markets, but even the most regulated industries are accelerating their cloud journeys.

And now, some CIOs don’t talk about the cloud anymore as it has become very natural and almost invisible, because it focuses on how to combine the services, integration and software development efforts.

There’s a huge workstream in cloud transformation ahead of us and it’s still the beginning of the road for many companies (estimated 20%-30% of cloud adoption means 5x-3x more to go). The skills for optimization of cloud architectures takes into account multiple aspects (as the cloud architects are supposed to do) including short term and long term cost effectiveness.

We are cloud enthusiasts at Avenga. We are open to helping you, as we understand different contexts of  cloud innovation mapped to business strategy and points of view; we are cloud agnostic, and are prepared for everything. Explore Avenga cloud service offering