Google first announced its plan to become a top cloud provider for the enterprise in June 2012. But turning an inward-focused, engineering-driven company inside out to cater to enterprise customers has been a struggle. By most estimates, Google Cloud remains a distant No. 3 behind AWS and Microsoft Azure.
Last week’s Google Cloud Next conference may mark a turning point. At 10,000 attendees, the three-day event was more than four times the size of last year’s conference. A change in tone emerged: Google spent more time actively reaching out to enterprises than it did flogging its technical superiority.
Instead of Snapchat or Evernote, real enterprise customers waltzed across the stage, including Colgate, Disney, HSBC, Schlumberger, and Verizon. Plus, Google announced a partnership with the fusty enterprise software vendor SAP, which will run its in-memory HANA analytics database on Google Cloud.
A slew of announcements fed into Google Cloud’s strengths: machine learning, security, G Suite, and that gargantuan Google infrastructure. The real surprise, though, was the level of effort to engage with enterprises looking to migrate workloads and build new applications.
Teeing up for a smarter future
If you were to boil down Google Cloud to one differentiation, you’d have to single out its AI and machine learning capabilities. In Martin Heller’s comparative review of machine learning clouds, Google’s TensorFlow library came out on top, a reflection of the company’s ongoing efforts to package deep learning offerings originating from its core search business.
At Google Cloud Next, the recently hired Dr. Fei-Fei Li — previously director of the Stanford Artificial Intelligence Laboratory and now Google’s Chief Scientist for AI — announced her intent to “democratize” AI, which she characterized as “lowering the barriers to entry and making it available to the largest possible community of developers, users, and enterprises.”
Li’s big announcement was the acquisition of Kaggle, the influential crowdsourcing platform for analytics and machine learning solutions. Kaggle adds to the outreach represented by the Machine Learning Advanced Solutions Lab that Google launched last fall. Other announcements included a new Cloud Video Intelligence API to enable search for specific content along with the general availability of Cloud Machine Learning Engine.
The most impressive new offering, however, was Google Cloud Dataprep, which both helps customers prepare data for analytics and machine learning purposes and uses machine learning to make the prep process easier. Of all the Cloud Next announcements, Dataprep may be the most significant, because it reflects a real-world understanding that customers spend far more time prepping data than actually crunching it.
Securing the cloud
As Google Cloud Senior Vice President Diane Greene told me last year, Google’s exceptional security is a key part of its value proposition to enterprises. In reference to potential enterprise customers, she said, “We have 600 security engineers. They can’t keep up with that no matter how big a company they are.”
But by now, most companies understand that the major public clouds can withstand attack better than their datacenters. The key is in managing cloud security. As InfoWorld's Fahmida Rashid noted in her analysis of Google’s security announcements, Google’s new Key Management System simply plays catch-up with AWS, but two new security services in beta — the Data Leak Prevention API and the Identity-Aware Proxy — break new ground.
The Data Leak Prevention API enables developers to add the ability to redact sensitive information in any application they build in the Google Cloud, opening up the previously announced DLP features for Gmail and Google Drive. The Identity-Aware Proxy offers granular control over access to cloud applications without a VPN, which Fahmida cites as an effort to differentiate Google’s cloud security at the application, not only infrastructure, layer.
Reaching out to the enterprise
There were other big announcements, such as the ability to stream files from Google Drive and new details on the forthcoming CloudSpanner, a massively scalable SQL database announced a month ago that will complement Cloud SQL.
Yet the new levels of customer support stood out. Take the new Engineering Support program, which Google describes as “a role-based subscription model that allows us to match engineer to engineer, so we can meet you where your business is, no matter what stage of development you’re in.” Compute Engine’s new Committed Use Discounts are also impressive, enabling customers to save 57 percent if they commit to memory and virtual CPU usage for a period of one to three years.
Last year’s Google Cloud Next signaled Google’s intent to reach out to enterprises. This year, they backed it up with new programs and capabilities that could make a real difference in the pace of Google Cloud adoption.