Learning about Google Cloud's key differentiators – big data and AI
The cloud business is a real fight of titans, with the "titans” being Microsoft, Amazon, Google, Alibaba, and a few others. They are among the biggest organizations in the world in terms of network and infrastructure footprint. Competition is good for any business, of course, but for us customers, it can make it difficult for us to decide between the options we have when they are all seemingly great.
Why Google?
Google Cloud was late to the cloud business. And it didn't speak to the early adopters when it decided to focus on PaaS services primarily (as opposed to the more familiar IaaS option), something the world wasn't yet ready for in the early days of cloud computing. But things are now changing. Google Cloud offers some of the best-of-breed services for the development of modern, containerized applications, using sophisticated analytics and AI capabilities that are cost-competitive. Google is the inventor of Kubernetes and some of the cutting-edge big data technology innovations such as TensorFlow, MapReduce, and BigTable. Google knows how to handle large-scale distributed applications and very large amounts of data probably better than anyone, because this is exactly what it has been doing (and doing really well) for about two decades now. It also knows how complex these systems are, which is why it is presenting them to its customers as a friendly, serverless, and easy-to-consume service in the cloud.
The reluctance of many enterprises to consider Google Cloud is understandable. Amazon's AWS is very mature and reliable, and their client portfolio certainly speaks for the success and maturity of their platform. Microsoft's Azure is a natural best choice too if all your existing IT systems are based on Microsoft products and services, such as Active Directory, Windows Servers, and Microsoft (MS) Office applications. Microsoft has dominated the enterprise IT business for years (and probably still does, with its strong appeal to enterprise customers due to existing relationships and its level of maturity), and they are certainly making sure that a transition to the Azure cloud would be as seamless as possible. Plus, they understand the needs of enterprises deeply, their security and compliance concerns, their need for protecting users' identities, and their need for reliability and strong Service-Level Agreements. This is all very true, but an unfair assumption that's often made in that line of reasoning is that Google does not have any of that expertise or credibility themselves. People will often say Google Cloud is "cool," but it is not as safe a bet as AWS or Azure. That might have been true once, but now, that is an assumption that you, as a cloud architect, can learn to challenge.
Firstly, Google Cloud has a growing list of high-profile customers, some of which are:
- PayPal
- HSBC
- Target
- eBay
- Verizon
- Spotify
- Deutsche Bank
These are just a few. They have all reported that they have realized benefits with Google Cloud, especially around big data and AI capabilities, but not exclusively so. PayPal's CTO Sri Shivananda has stated that PayPal turned to Google as their cloud provider "because it has the fastest network in terms of throughput, bandwidth, and latency." It's not very difficult to understand why Google's global network is hard to beat: it powers latency- and bandwidth-sensitive services such as Google Search and YouTube. Twitter runs on a distributed Hadoop system that hosts more than 300 PB of data across tens of thousands of servers on Google Cloud. Parag Agrawal, CTO of Twitter, said "Google Cloud Platform provides the infrastructure that can support this and the advanced security that well serves not just our company but also our users. This peace of mind is invaluable."
Multi-cloud friendly
More and more companies are also evaluating so-called multi-cloud strategies to benefit from the best of two or more cloud platforms. Google seems very much on board with this, facilitating multi-cloud strategies and even developing products that work on multiple cloud platforms. Organizations that would like to leverage, for example, Microsoft Azure to migrate their existing Windows-based systems and Active Directory identities can still do so, while more modern applications can be migrated to Google Cloud to reap the benefits of modern capabilities. You can have yet another piece of your application in AWS too, should you consider that platform to be best suited for it. The one drawback of multi-cloud deployments is the higher management overhead and the more diverse skillset required within the organization to manage multiple platforms, which may increase the needs for internal training and/or hiring (and therefore drive costs).
Once again, however, the strategic discussions around business motivations and business outcomes become very relevant in identifying ways to actually obtain a higher ROI through a multi-cloud deployment model, by aiming to explore the best of what each cloud provider has to offer. With the right strategy and the right execution, multi-cloud certainly pays off.
Big data and AI
One important consideration when it comes to making the case for Google Cloud is to adopt an innovation-led development mindset and understand industry trends. We may be in a moment now where big data and AI are somewhat what cloud computing itself used to be in its early days: a shiny new object, a cool-but-maybe-not-for-me-yet technology. Now, nearly everyone is in a rush to get to the cloud, and some probably regret not being among the early adopters. The same could happen with AI and big data: this may likely be the right window to get in, and should you decide to do so, it is hard to dispute the fact that Google Cloud is leading the way and offering the very best you can get in that arena. Google firmly believes in the saying that "every company is a data company," or at least that this will eventually be true, because it seems to be seeding the ground for a future where the best big data capabilities will be the deciding factor on who's going to win the war for cloud market share.
In a 2019 PwC study, it was estimated that AI will provide a potential contribution of $15.7 trillion to the global economy by 2030, with high potential use cases across several different industries. What is already a reality in 2020, however, is that companies are investing heavily in cognitive software capabilities to analyze data and inform decision-making or to build more personalized and profitable products and campaigns. Other areas of focus of AI currently include supply chain optimizations, mobile features (such as real-time speech translation), enhancing customer experience (by providing, for example, always-on chat bots or contact center AI), and improved security (with features such as fraud and threat detection).
The common theme underlying these use cases is not old technologies being replaced (or people, at least not for the time being), but rather the improvement, enhancement, or augmentation of existing ones to provide more value and profitability. The field of AI has many aspirations and ambitions for the future, and those are the ones we tend to hear about the most in the press, which gives us the impression that AI still only belongs to the realm of imagination and research. But it is very real already, and companies have been reaping its benefits for years. Jumping on the AI and big data bandwagon is likely one of the safest bets tech organizations can place today, especially considering the entry barrier has never been this low, with easy-to-consume, pay-as-you-go AI services that are on the bleeding edge of the field.
And it gets better. You don't need to lock yourself in with Google Cloud products if you want to leverage these capabilities – you can simply build your applications on open source tools you might already be familiar with.
Open source
If you work with an organization involved in the open source world (be it as a contributor, or purely as a consumer), Google's heavy involvement and strategic partnerships with leading open source companies may appeal to them. These partnerships have allowed Google to integrate open source tooling and services into Google Cloud, making it easy for enterprise customers to build and use apps by using those services in ways that feel very similar to the experience of using cloud-native resources. If vendor lock-in is a concern in your organization, you can see how Google Cloud embraces not only the open software culture, but also the multi-cloud model, by allowing several of these services to seamless integrate with systems that are hosted in other public clouds or on-premises data centers (in a hybrid cloud model). And because open source software does not rely on proprietary libraries and technologies, they are easier to port to another cloud provider, should you decide to do so. One of the best examples is Kubernetes, an open source container orchestration platform where you can run service-oriented applications completely agnostic to the underlying infrastructure. In fact, one of Google's latest developments is Anthos, a new and also open platform service that lets you run applications anywhere inside Kubernetes clusters, unmodified, across different clouds and/or private on-premises environments. It is a fully fledged hybrid- and multi-cloud solution for modern application development.
You have now gotten a better understanding of the many strengths of Google Cloud and some reasons why organizations should consider it, as well as how you can make a convincing case for it. Next, let's dive into Google Cloud and its technologies.
