I was fortunate enough to attend an invite-only Google event to get briefed on numerous announcements pertaining to Google’s cloud services. The announcements included updates on products ranging from Google Docs to Google’s public cloud offering. Additional information was shared on Google’s go-to-market strategy and staffing ambitions as it gears up to gain ground on AWS and Azure over the next few years.
The Google Cloud Platform blog did a nice job of summarizing all of the announcements from the event. My focus was entirely on the Google Cloud Platform (GCP), so I will briefly cover some of the non-GCP announcements and then dive into GCP.
The first announcement of the day was that Google is changing the name of its enterprise business suite of tools from “Google for Work” to “Google Cloud.” The second branding change is the renaming of its suite of collaboration tools—Google Docs, Sheets, Slides, Gmail, Gchat, Hangouts, etc.—to G Suite.
The G Suite announcement was more than a branding exercise. Google added its own machine learning technologies to create a more intelligent set of tools that learns how you use the tools and helps you collaborate more effectively. For example, Google can recommend content, templates, images, and more based on what it has learned from your styles and preferences over time. Search on G Drive is much more intelligent now, and Sheets takes the complexity out of creating formulas by introducing speech recognition so that users can say what formulas or calculations they want. The G Suite stuff was cool, but that was not what I was there for.
Diane Greene, Senior VP of Google’s cloud business, shared how Google is aggressively going after enterprise customers. Here are some of the highlights:
- Hiring 1000+ customer-facing resources
- Drastically increasing the number of both technology and services partners
- Increasing investments in enterprise security features
Diane claimed that the above-mentioned investments have resulted in a doubling of the number of new enterprise customers and an uptick in compliance-focused companies (finance, healthcare, etc.)
Google invested $9.9B CapEx in GCP in 2015, and it will invest at least that much again by the end of 2016. Google has seen an impressive year over year growth of “10x storage, 4x compute, 25x containers.” It has over 10,000 engineers working on the platform and is iterating at a rapid pace. Keep in mind that Google first released its IAM services at the beginning of this year and has already increased its feature set and robustness substantially in just eight months.
Google believes that GCP’s four key differentiators are security, data/analytics/machine learning, openness, and performance. I personally agree with the last three. The jury is still out on security, which can only be determined by the opinions of enterprise customers. I think Google has a very robust security feature set. Its challenge is more on the marketing side. Not all enterprises are aware of how much GCP’s security capabilities have matured over the last year.
Google announced that new regions in Northern Virginia, Singapore, Australia, United Kingdom, Germany, Brazil and India will roll out in 2017. The Germany region is a big deal. I have worked with many clients who have run into challenges with their public cloud initiatives because of the strict German privacy laws. With a data center in Germany, the issue of data leaving the country goes away. There may still be issues with data running on an American-based company’s infrastructure, but I will assume that in many cases, a Google Germany region will help enterprises get over the privacy regulation hurdle.
Google Cloud Machine Learning is now publicly available in beta. ML allows businesses to easily train machine learning models at an amazingly fast rate, with no previous knowledge required of the actual models. Google has abstracted the complexities of training models so that developers can start writing code leveraging robust models without waiting for the long and rigorous processes that were required in the past to complete.
Cloud Machine Learning is a fully managed service that can scale on demand and creates a rich environment across TensorFlow and cloud computing tools such as BigQuery, Cloud Storage, and Cloud Datalab. Google also announced its new Machine Learning Advanced Solutions Lab, which allows businesses to engage with Google Machine Learning engineers to help them solve their complex problems. In addition, it has created the Cloud Start program, which offers educational workshops for businesses to learn about the fundamentals of the public cloud and how to identify opportunities with advancements such as machine learning. Google is dedicating resources to help enterprises conquer the cloud learning curve at a faster rate.
BigQuery for Enterprises
In my opinion, BigQuery’s capabilities and performance were already unmatched by any cloud provider. The only knock on BigQuery was that developers had to learn a proprietary query language to use it. Google announced that BigQuery now supports standard SQL and new ODBC drivers, which makes it possible to use BigQuery with a number of tools, ranging from Microsoft Excel to traditional business intelligence systems such as MicroStrategy, Qlik, and Alteryx.
Enterprises want control and visibility into the cloud services they consume. Google answered the call with these announcements:
- Monitoring through Google Stackdriver to track workload performance and usage
- The ability to update, delete, and insert rows and columns in BigQuery datasets using Standard SQL
- Query sharing via links, to foster knowledge sharing and collaboration within organizations
- Integration with Cloud Identity and Access Management to manage fine-grained security policies for BigQuery users and resources
BigQuery’s previous pricing model was a pay-as-you-go model based on resultset rows returned. This was an awesome pricing model if you were good at creating efficient queries, but expensive if you were not good at it. So, Google now offers monthly flat-rate pricing that pairs unlimited queries with predictable data storage costs. With flat-rate pricing, you perform all the queries you want, delivering the power and scale of BigQuery with a simple, predictable bill. In this model, storage is priced separately, allowing you to economically keep as much data as you want.
Google also announced the release of Kubernetes 1.4, which you can read on the Kubernetes blog.
Google is all in on going after enterprise customers. It is ramping up staff in all areas from engineering, marketing, and sales to professional services and training. It is making huge investments in infrastructure. In fact, Google claims it is investing almost as much as AWS and Azure combined. It is adding more regions across the globe and rapidly adding features in areas of security and governance that are required by many enterprise customers. It has had 339 product releases since January.
An amazing statistic is that only 5% of the $1T cloud market has been won so far. Google’s infrastructure, especially around networking, is second to none. Although Google might be a distant third behind AWS and Azure today, the rate at which it is delivering, coupled with the new innovations it is bringing to market, puts it in a great place to get a much larger chunk of that $1T pie in the next five years.
Another key factor is the NoOps direction. NoOps is a loaded term that freaks out DevOps people and sys admins. In this context, NoOps really means higher levels of abstraction. It does not mean no DevOps or no operations. It means more abstraction, fewer operations required, and more time spent creating business value. Just look at Google services like Cloud Machine Learning, BigQuery, Cloud Pub/Sub, and Genomics. These services allow developers to create amazing solutions in days in what used to take weeks or months. All of these services autoscale and require minimal development to operate. The speed at which developers can build on the Google Cloud Platform is a key differentiator in my book.
Google is already known by many to be a top choice for running analytics workloads in the cloud. Soon we will see if enterprises will view it as the place to run most other workloads as well. If it succeeds in its marketing efforts to convince customers that it is enterprise ready, the three-horse cloud race could get real interesting down the final stretch.