Is Anthos the edge Google needs in enterprise cloud?

In third place in the cloud wars for years, the emergence of Google's Anthos last year was a welcome move that makes a versatile multicloud future for businesses not just a possibility, but a practical option today. But is this what Google needs to truly succeed in the cloud industry?
Written by Dion Hinchcliffe, Contributor

I've been studying the news coming out of Google Cloud Next '20: On Air, the now digital-only version of their yearly developer conference that's been spread out over 9 weeks. So far the stand-out has been the new enterprise features, such as Google Cloud's new confidential VMs. This makes sense because if Google has any serious chance to be a dominant player in cloud infrastructure for enterprises, it will be because of enterprise-class features, especially Anthos.

Unlike earlier this century, that Google is now utterly serious about the enterprise market there is little doubt. Thomas Kurian, who has become both an effective platform evangelist and division chief executive -- as well as the cloud industry executive with the most gravitas --  has proven so far to be a stellar choice. It's also a telling one, given Kurian's Oracle background, that essentially eschews Google's consumer roots while aiming squarely at the highest margin, highest value, least commoditized segment of the cloud industry today: Larger organizations and their unique and often rarified needs.

In his recent opening keynote video, Kurian spoke plainly and with great purpose, omitting much of the traditional language of the cloud, and stated their mission in terms more familiar to executives than developers, to "accelerate every organization's ability to digitally transform and re-imagine their business through data-powered innovation."


There are some key pieces to parse here. it's notable that the first part of this mission statement is much more accessible to existing enterprises with vast legacy IT investments. Most CIOs are acutely aware their aging IT landscapes are increasingly stranded in outdated and parochial data centers, while their board and stakeholders demands from them much more rapid modernization, transformation, and innovation (albeit while often providing limited additional resources to actually achieve these goals quickly.)

This is likely why the first sentence of the main Anthos product page starts with the word 'modernization.' The message is that Google isn't focusing on being the world's leading pure-play hyperscalar to host the next great $100 billion Internet startup. Instead, they are addressing the market that sustains so many of the largest software companies: Other big enterprises. As we'll see, Google is building a cloud migration path that enterprises can readily recognize and relate to.

As for the second part of the statement, that Google is pitching a data-powered vision of the future is in fact a clear nod to their heritage in taking world's Internet data and turning it into one of the most valuable companies in history. This is arguably one of their top and most valuable competencies, though it's not yet as evident in Anthos.

It is important to understand the key distinctions between Google Cloud Platform (GCP) and Anthos. GCP is actually more akin to AWS and Azure proper. It is a robust fully cloud-native platform with a full range of popular services in its own right. In contrast, Anthos is an open yet highly secure enterprise-class cloud migration, management, and operations environment. It helps move enterprise applications to the cloud in the most architecturally and process-appropriate way, then makes those applications easier to operate and oversee across every type of cloud resource.

Betting on the Enterprise in a Major Way

In fact, the focus on very specific terminology that is friendlier to business users, enterprise architects, and IT executives was in evidence throughout Kurian's Google Cloud Next talk. At its core, Kurian positioned Google Cloud as "distributed infrastructure as a service" and not the more familiar shorthand, IaaS. He is also very intentional in his description that there is one programming model and one API to run code as if Google Cloud is a single, extraordinarily large computer. Simplicity is both a core enterprise IT virtue as well as a central Google product approach, is the message.

Finally, and this is to my mind the key point and differentiator, Kurian is also clear that Google Cloud runs anywhere, period. On their cloud, in existing customer data centers, in their competitors clouds (AWS and Azure in particular), as well as edge devices (citing telecom customers.)

Simplicity and inclusion is the singular vision to enterprises from Google. If you want a cloud capability that just works anywhere, plays as well as anybody can with legacy data center infrastructure, understands enterprise needs and nuances, and is truly multicloud ready, then Google Cloud is the answer.

But what then is the actual question, if Google reasonably seeks some form of cloud supremacy, even if just in the larger enterprise? Or a future not running a distant third in the cloud industry with the real possibility of being relegated to single digit market share. This would be a precarious position where inexorable commodity pricing pressure and/or insufficient economies of scale have the potential to make Google's cloud business very hard work for marginal returns. 

This is not to say that Google Cloud is struggling exactly. Their Q2 revenue grew an impressive 43% year over year, so they're outperforming the cloud market as a whole by a good margin, which is growing at only about a 17% annual rate industry-wide. But with only $3 billion in revenue for the quarter, it's also one of the smaller divisions Google operates. Situations like this with the cloud giant has long made IT executives nervous, because of Google's well known inclination to suddenly abandon struggling products. For enterprises, most IT acquisitions are long term, carefully evaluated partnerships where product longevity is a key consideration. For now however, Google Cloud and its new enterprise offering, Anthos, seems pretty safe.

Can Anthos Put Google in the Cloud Driver Seat?

A defining question for Google Cloud today is now whether Anthos is the vehicle that will address urgent and compelling enterprise needs in the cloud in a new way that Microsoft Azure or AWS are unable or unwilling to?

To answer this, it's worth taking a look at what the top needs for the cloud enterprises actually are, based on my conversations in recent years with a good number of large enterprise CIOs:

A smooth road to cloud migration and IT modernization. At the end of the day, the vast majority (80-90%) of the IT budget goes to keeping existing infrastructure and apps operating, either maintaining them internally in a data center or outsourcing them to a managed service provider. What I've previously dubbed the IT 'legacy mountain' has to be reconciled en masse with the modern tech world over time. Ultimately, CIOs want the fewest, simplest, easiest, safest, and cheapest future operating environments that will support their needs, because a) it's more straightforward to manage, b) vendor management isn't as big a headache, and c) there are just less "throats to choke" as they say in the business. Thus, to most CIOs, the end game of IT modernization is clear: They must move everything that makes sense to the cloud, and that is the vast majority of IT. So ease of migration, low risk of migration, and cost effective migration are very much near the top of the list of enterprise needs for the cloud.

Container-ready IT. The future of IT systems, cloud operations, and IT governance lies in the ever-more-popular container model, which is now the definitive leading practice when it comes to application development and deployment. While there is still a long way to go to make most of enterprise IT container-ready, it's pretty safe to say that's where it will eventually end up. Most new applications are built to use containers today while many older applications are now being upgraded to run in containers by their vendors over time. 

It's also worth noting here that Google essentially gave rise to Kubernetes, the most popular -- and most vendor-agnostic -- container framework and open source project. The key point is that CIOs now strongly prefer to acquire or migrate to container-ready IT or container-less solutions, for all the attendant benefits they confer. This includes better management of growing cloud deployment complexity as well as ready sourcing of talent to manage the cloud itself (because of the wide popularity of Kubernetes.)

Support for hybrid cloud. There is a very significant advantage to operating IT under one primary operating model or architecture, for consistency, simplicity, and choice. CIOs are thus very interested in using cloud technologies and architecture as their default application model for all IT, even in their own data center, and mixed into their legacy systems, because it becomes more future-ready. They know that they can ultimately migrate it out to the cloud more easily and less expensively when the time is right. For new projects, even on-prem, that means using their preferred provider's cloud stack. In short, CIOs would like a cloud provider that has a platform that runs in their data center with as little compromises as possible, while offering enterprises the ability to manage all their cloud resources, private or public, with one pane of glass.

An agnostic approach to cloud and proactive support for an industrialized multicloud. This is a nuanced topic, but in short, the CIO already knows they will have too many clouds to manage easily, and that many of them may not play together well. In the big picture, the CIO wants maximum flexibility and choice to move workloads around, switch providers quickly, and mix-and-match best of breed services. For instance, the artificial intelligence (AI) services in one cloud may be a better fit, while the IoT services in another cloud may work better for a given use case. These cloud resources should be easy to use together, even if they're in separate clouds.

The cloud provider that assumes by default it's a cog in a larger enterprise IT architecture, plays the fewest favorites with its own services, and actively builds bridges to and enables other clouds will have a significant edge with IT buyers. Significant bonus points are to be had if a cloud provider can provide one management console for the commercial clouds that an organization uses. 

Unique enterprise needs in the cloud. The countless vagaries and sheer variety of on-premises IT in a large enterprise cannot be understated. The IT portfolio is typically comprised of somewhere between 500 to over 3,00 applications for most companies. Blends of many generations of IT are the norm, with applications and systems spread out across seemingly-ancient legacy hardware like mainframes and mid-range computers like AS/400 to early-generation client/server systems, distributed object applications, and old native code business applications. Many, even the majority, of IT systems in most enterprise are rather old, but they're also mature and paid for. And they tend to have many quirks and eccentricities that have emerged as the tech world has moved on around them that must be managed.

Old technologies aren't the only issue for cloud. There are also the unique aspects of industries in which enterprises are carrying out business, ranging from public sector and healthcare to financial services and insurance, which all have their own privacy, security, and other types of regulations around them that directly impact cloud deployments. Finally, there are geographical issues when it comes to large organizations, from distributed IT performance management challenges and data residency requirements to a dense thicket of privacy regulations and legal constraints when operations are spread across dozens of different countries. in short, larger enterprises are very complex. 

A successful cloud provider must adapt their offerings to these realities as much as is feasible to maximize their share of customer. Or the customer has to make the time and investment to do it. This invariably slows down or even puts off cloud migration for many systems altogether.

Multicloud Not a Cloud Future: It's The Future

The next major battleground in IT is who will be the leading player in the higher-order realm of multicloud. To this question the fortunes of some of the largest tech companies in the world are now at stake. The the biggest friction in cloud adoption that most enterprises are already starting to encounter is the sheer complexity of modern app deployment and management across multiple clouds for hundreds of applications and services, which are also increasingly integrated with each other. 

Contemporary cybersecurity theory is also contributing to this complexity. One example: The recent observation that fixed cloud infrastructures are much easier to attack has given rise to new dynamic cloud architectures that can rapidly shapeshift, disappear, and recombine in the face of cyberattacks to new location and configuration somewhere else in the cloud  -- while still functioning seamlessly -- is becoming an actual new design trend. This is an order of magnitude more complex to realize than old fixed IT architectures, and is likely the cost of doing business in the future. Thus, for reasons like this, complexity management is now a core requirement of cloud architecture and operations.

AWS and Azure have perhaps been slow to full realization of the primacy of multicloud but they are beginning to address it with offerings like AWS Outposts and the new, but still early Azure Arc

However, this is where Anthos has placed careful product bets along many of the the top enterprise cloud needs described above. Google's deliberate focus on requirements that seem unusual or unnecessary to pure-play cloud customers is a tremendous asset here, again likely to tap into customers looking for enterprise features they can't find anywhere else. Here's where Anthos seeks to meet enterprises where their most significant cloud challenges are:

  • Rapid, low friction migration paths. The Anthos migration playbook cleanly and partitions legacy apps into three different buckets, upgrading them to native cloud architectures to the extent each type is able. A focus on migrating Java business applications, which were built by enterprises around the world for over two decades to run their businesses, is especially welcome as well. Google has data that shows its migration approach for enterprise apps both simpler and faster than other methods. The Migrate for Anthos architecture is also some of the most modern and well-articulated in the industry.
  • Best-in-class, born-native container support. Anthos GKE (Google Kubernetes Engine) is Google's realization of Kubernetes and it's at the core of Google's cloud strategy. AWS was designed before modern containers were a major trend, and while it can certainly run containers, it's not baked deeply into the product architecture like it is with Anthos. Azure is in a similar boat, in that it doesn't offer containers as its fundamental operating concept. Especially with Google's Kubernetes roots, the advantage to enterprise customers is that Google offers the clearest, most-native migration and operations path for enterprise to robustly and competently containerize their legacy IT systems.
  • Hybrid and multicloud-ready, including management. Google has made it clear that Anthos can be run and managed centrally even in the most federated multicloud environments, no matter where its actual cloud resources are located: In Google's data centers, on-premises, on the edge, or in 3rd party commercial clouds like AWS and Azure. Google doesn't discriminate and they're all on an equal footing, to their extend their location allows. Yes, all of this is possible to the degree on other commercial clouds, but not with the grace, fidelity, and seamlessness evident in Anthos.
  • Support for unique enterprise needs. Google has anticipated and supports unique enterprise requirements in Anthos, such as the aforementioned Java migration path, as well as features like enterprise-grade microservices, serverless operation anywhere including on-prem, a commendable approach to true consistency in development and operations wherever Anthos resources are running, and automation of IT policy and security at scale seamlessly across multicloud environments with Anthos Config Management.

In the final analysis, Anthos is an overarching bet by Google that it can present and deliver a compelling multicloud solution for enterprises to accelerate their move to the future-ready environment that is already fully multicloud capable. In other words, it's  clearer vision for the next-generation of cloud-based IT. Anthos beats AWS and Azure as both an on-ramp for the modern cloud environments of today that is the most enterprise-friendly as well as entry to the new multicloud future.

To clear, Anthos will likely never outsell or dominate the commercial cloud computing industry, but it does have a fair prospect of improving Google's chances against the #2 player in cloud, Microsoft, who has already had virtually every enterprise IT shop as their customers in one form or other for the last 30 years. CIOs have often told me over the years that they really don't like putting all their eggs in one vendor's basket. By positioning itself as the most agnostic choice for the overall cloud strategy over all the other providers, and shifting the conversation and focus to the overall future cloud environment of the enterprise, it may help Google Cloud carve out not just a niche, but the leading strategic position for cloud in the enterprise.

Additional Reading

Editorial standards