Google's Next cloud move - yet another stab at the enterprise

Kurt Marko Profile picture for user kmarko March 13, 2017
Is Google finally getting serious about enterprise cloud? We'd like to think so but then we've been here many times before.

Eric Schmidt, Executive Chairman, Google

Should anyone have doubts about Google's commitment to cloud services, the company tried to silence the skeptics, particularly those among the media and analyst community seeing the cloud business as Amazon's to lose, by unleashing a bevy of product, partner and customer announcements at its Cloud Next '17 event last week. But will Google (finally) get serous about he enterprise?

As has become standard practice for tech giants seeking to maximize their exposure and marketing effect, Google queued up an array of new services, features, apps and hardware that spanned its IaaS and PaaS Google Cloud Platform and SaaS G Suite businesses.

Indeed, the seeming non-stop barrage was such that most analysts that arrived early for a pre-conference analyst day event were shellshocked when an evening welcoming reception turned into a 90-minute litany of announcement summaries that Google had queued up for the rest of the week.

The overarching themes are that Google sees the cloud market as nascent and ripe for competition, that the company is committed, to the tune of $30 billion in infrastructure investments over the past three years, and ready to aggressively court enterprise customers through a combination of direct outreach, partners and new migration and onboarding services. As Google's Executive Chairman Eric Schmidt put it in his day two keynote,

This is an incredibly serious mission; something I've wanted to do since I joined the company 17 years ago. … The company has both the money, the means and the commitment to pull off a new platform of computation, globally, for everybody who needs it.

News highlights

Google's cloud business includes both infrastructure and applications, so in this column, I'll focus on the former, news about Google Cloud Platform (GCP), with a follow-up to review enterprise developments in the G Suite productivity and collaboration portfolio. The GCP news falls into four categories:

  • Core infrastructure, security and cloud operations
  • Enterprise application development
  • Data management, analytics and machine learning (ML)
  • Enterprise partnerships

Google's primary selling points are the scale, sophistication, reliability and security of its infrastructure, upon which it builds a host of application development, data analysis and AI services. As the $30 billion figure indicates, Google has been busy building out perhaps the most sophisticated cloud infrastructure and network in the world.

In a private briefing, some of Google's network and security engineers indicated that the traffic across Google's network rivals, if not exceeds that of the largest telecom carriers.

At Next, the company announced three new regions, in California, Montreal and the Netherlands, each of which will feature at least three zones, a new pricing discount based on aggregated committed usage (rather than fixed-term leases like AWS reserved instances) and improvements to its VPN service to support cross-project networking (XPN) that allows sharing virtual networks among different enterprise organization or projects.

Google emphasized its expertise and investment in security technology. These are top of mind enterprise topics which Google supported with several pertinent announcements. Recognizing the increasing vulnerability of hardware OEM'd from foreign suppliers, Google has developed a custom security chip that embeds crypto certificates to establish an immutable hardware root of trust for servers, network equipment and peripherals used in its data centers and that allows it "to more securely identify and authenticate legitimate access at the hardware level."

Google also introduced several new security services for key management, hardware security key enforcement, data loss prevention and an identity-aware proxy that provides more granular control of application access than a client VPN.

The DLP API is particularly innovative, as this evidenced by this demo in one of the keynotes. The service uses Google's extensive expertise in search, textual analysis and pattern recognition to enable applications to automatically scan and redact sensitive, Personally identifiable information (PII) such as names, credit card numbers and taxpayer IDs from both text documents and images.

App development and data analysis

It's not surprising that a company built upon selling targeted ads based on user demographics and behavior should have an impressive portfolio of data storage, management and analysis services. Indeed, Schmidt sees big data as such a strategic technology for both business and national security that "nation states will fight over it."

Although AWS and Azure also check the right boxes with the requisite SQL, NoSQL, Hadoop and analytics features, no one can provide a global-scale, distributed SQL database like Google Cloud Spanner. Introduced a month ago, Spanner was a highlight of several Next keynote presentations and technical sessions since it illustrates Google's technical expertise in building highly scalable distributed systems.

Although Spanner usage scenarios and customer adoption are still unclear, it provides the technology to address transaction processing applications with unprecedented scale while providing ACID reliability and full SQL compatibility.

AI and ML are areas where Google excels and sees as providing a competitive advantage in the cloud market. Higher value service abstractions are critical to what Google's head of AI and Chief Scientist Fei-Fei Li called the "democratization of AI" by allowing non-specialists in business to tap into sophisticated models and techniques that were formerly limited to a high-priesthood of ML researchers.

In expanding its AI portfolio, Google announced the general availability of the Cloud Machine Learning Engine that simplifies the training and deployment of custom ML models using the TensorFlow platform and other GCP services like BigQuery (data warehouse), Cloud Dataflow (ETL) and Cloud Datalab (data visualization and transformation).

Google also announced a limited beta of Dataprep, a data transfer and preparation service that allows users to visually inspect, rearrange and filter data structures before ingesting them into a database. GCP's new Vision and Video Intelligence APIs add support for video, not just static image recognition, and powered one of the most impressive demos at Next '17 in which an app highlighted video snippets from a longer clip that corresponded to specific keywords, for example, "baseball" in an hourlong sports highlight show.

Cloud partnerships

A clear goal of Next was to show Google as a safe choice for enterprises, which meant a parade of major customers and new partner announcements was inevitable, however none was more important that its work with SAP to port HANA to its platform. As Jon Reed and Brian Sommer said  in their analysis:

These announcements are in tune with the choice today’s customers expect. It’s refreshing to see SAP acting like the cloud company in their tagline, talking openly about containerization, customer choice, and empowering developers. This is a striking contrast to the perception of SAP as data auditors from the recent Diageo verdict.

Google also emphasized the importance of service partners and system integrators in engaging enterprise customers and delivering tailored services and applications using GCP.

According to Google Cloud President and head of customer engagement, Tariq Shaukat, it has 13,000 partners with collectively more than a million customers, although many of these are undoubtedly using G Suite, not Cloud Platform. According to the head of Google Cloud, Diane Greene, the company closes well more than half of its enterprise sales engagements.

My take

Google has an annoying history of making big announcements and then failing to deliver. It is for example doubtful whether Google Apps beyond Gmail have made any demonstrable inroads into the Microsoft franchise. A big part of that problem stems from Google's tendency to focus its appeal towards the technical audience rather than the business buyer.

You can certainly make the argument that is what happened in the tsunami of announcements I've briefly touched upon. You can also argue that even holding hands with SAP while eyebrow raising is nothing new. These two companies have been dancing around one another for years but with no discernible outcome.

Even so, at Next '17, Google worked hard to dispel the perception that GCP only appeals to companies with cloud-savvy propellerheads like, well Google.

For example, Cloud Platform VP Brian Stevens discussed its support for various stages of cloud adoption by highlighting a new VM workload migration tool, better Windows application support and better pricing (including the aforementioned committed use discounts) that are designed to entice enterprises to move workloads from legacy data centers to GCP. He recited a mantra heard from several execs that Google "seeks to meet customers where they're at on their journey to the cloud."

Stevens and other GCP leaders are correct in seeing workload migration onto a VM platform like Cloud Engine as the first step in cloud maturity that can evolve into greater usage of Google services as enterprises refactor applications to use containers and migrate corporate data into Google storage to use its data services for sophisticated analytics.

Ultimately, these customers will build new applications using cloud-native services using higher level abstractions such as Google App Engine (PaaS), Functions (serverless) and AI and ML APIs.

AWS clearly dominates the cloud business, but the cloud market is not static and still rapidly growing. It's impossible to argue with this assessment from the WSJ that the short-term advantage in the cloud goes to AWS and Microsoft Azure, but those assuming Amazon owns the cloud in perpetuity are making a big mistake, as I discussed here and here.

You can argue with the exact number, but I agree with Greene when she says that the cloud market has reached perhaps 5% penetration: there's far more cloud business to be won than to be protected. As it has demonstrated in so many other areas such as mobile devices, fiber infrastructure and autonomous vehicles, once things leave the labs and become a business unit, Google plays a long-term game.

Cloud services are no hobby like Glass; Google is serious and wooing customers and enterprises would be wise to listen. Competitors, customers and analysts underestimate Google at their peril.

A grey colored placeholder image