Google Cloud Platform’s secret sauce: Its time is now
- 21 December, 2017 22:00
2017 was a good year for Google Cloud Platform. But 2018 promises to be even better. Although Google still sits well behind Amazon Web Services and Microsoft Azure in cloud revenue, with just 2.3 percent market share against AWS’s 44.1 percent and Azure’s 7.1 percent, according to Gartner data, 2018 could well prove Google’s breakout year.
That is, if it can just deliver on one, eensie weensie thing: helping enterprises operate more like Google.
In the past, this has seemed like a fairy tale. Mainstream enterprises, after all, lack the DNA necessary to push thousands of changes a day to their systems in the way a Google routinely does. Early indicators, however, suggest that maybe this isn’t as true as we once thought and that, in fact, “run like Google” just might be an attainable goal for the typical enterprises. If so, that be what gets Google Cloud seriously into the enterprise mix.
Maybe enterprise can “run like Google” after all
Sure, AWS dominates, but much of that dominance comes down to a seven-year head start gifted to it by legacy infrastructure vendors and the current crop of cloud contenders. The legacy vendors are, in fact, doomed (though “doom” takes decades to occur in the enterprise). But cloud contenders like Microsoft and Google? Game on!
Microsoft has managed to accelerate adoption of Azure with a combination of enterprise bona fides (CIOs that do a lot of business with Microsoft trust the vendor to help cloudify their infrastructure) and next-generation AI smarts.
Google is growing market share even faster than Microsoft (100 percent compared to 61 percent, by Gartner’s estimates), and primarily through a combination of smart open source bets (such as Kubernetes and TensorFlow) and industrial-strength AI offerings.
This latter advantage has almost appeared to be a liability at times, with Google seemingly saying that enterprises could run a lot more like Google if they weren’t so stubbornly like themselves. However, it’s becoming clear that the “run like Google” tagline is an attainable goal, one that even the most staid of enterprises can realize.
Some of Google’s early customer wins may have served to solidify the “no, real enterprises can’t run like Google” argument. For example, when it was learned that Snap was paying Google $400 million a year for the “vast majority” of its “computing, storage, bandwidth, and other services, it was easy to respond, “Well, of course a ‘new school’ company like Snap would do that.”
More recently, Spotify announced its intention to build on Google Cloud. Although Spotify is also a new-school enterprise, its reasons for dumping not-that-old Hadoop, Hive, and other such data infrastructure for Google’s BigQuery, Dataflow, Dataproc, and Pub/Sub suggested that “run like Google” is within reach of mainstream enterprises. Although Spotify’s vice president of engineering and infrastructure, Nicholas Harteau, cited his belief that Google’s infrastructure is (for the moment) superior to others’ offerings, ultimately the decision came down to data:
What really tipped the scales towards Google for us, however, has been our experience with Google’s data platform and tools. Good infrastructure isn’t just about keeping things up and running, it’s about making all of our teams more efficient and more effective, and Google’s data stack does that for us in spades.
Harteau went on to laud the “sophistication and quality of [Google’s] data offerings”:
From traditional batch processing with Dataproc, to rock-solid event delivery with Pub/Sub to the nearly magical abilities of BigQuery, building on Google’s data infrastructure provides us with a significant advantage where it matters the most.
This advantage will only increase as Google’s data infrastructure improves. Already best in class, Google hasn’t been resting on its laurels, as this BigQuery example demonstrates.
Software may be eating the world, but it’s data that fuels its appetite. Any data advantage, in other words, has the potential to change an enterprise and, by extension, entire industries. Spotify isn’t “new school” because it streams music but because of how data-driven that data streaming has become, something that it expects Google Cloud to significantly advance.
Making enterprises less dull and more like Google
This wouldn’t be very interesting if only the Snaps and Spotifys could benefit, but it’s not just them. Old-school enterprises are climbing about Google’s data-friendly cloud, too. Take nonagenarian Schlumberger, for example. Hardly a cloud native, Schlumberger’s decades-old business has traditionally dealt with oil and gas, not ones and zeroes. This is changing, however, and Google Cloud has enabled that change.
Part of Schlumberger’s Google story begins with scale: Google Cloud enables “Schlumberger Omega geophysical data processing platform to run at a scale not possible in traditional data center environments,” the company says. Of course, Schlumberger could have done the same with AWS or Azure. But it’s data that made the difference, and got Schlumberger on Google Cloud:
- “Schlumberger launched the Delfi cognitive E&P [exploration and production] environment and the deployment of an E&P Data Lake based on Google BigQuery, Cloud Spanner, and Cloud Datastore, with more than 100 million data items comprised of over 30TB of petrotechnical data.”
- “Schlumberger leverages TensorFlow for complex petrotechnical interpretation of seismic and wellbore data, as well as automation of well-log quality control and 3D seismic interpretation.”
This focus on innovative data processing/analysis at scale, coupled with strength in AI, has pushed other mainstream enterprises to embrace Google Cloud, including:
- Telegraph Media Group: “As the solution we use to power our enterprise data warehouse, Google Cloud Platform is miles ahead of the competition in terms of delivering cost-effective speed, flexibility, and accuracy. … The ability to iterate rapidly over multiple terabytes of data across user interactions comprehensively has dramatically improved our audience intelligence. Google Cloud Platform makes this all possible.”
- Airbus: “In our tests, Google Cloud Machine Learning enabled us to improve the accuracy and speed at which we analyze the images captured from our satellites. It solved a problem that has existed for decades.”
- Philips: “With the help of Google Cloud Platform, we are changing the fundamental business model of selling lighting to consumers. We’re providing an entirely new set of value-based services that transform the home, rather than just replacing bulbs that burn out.”
Add to these BNP Paribas, Coca-Cola, Best Buy, and more, and it’s clear that plenty of boring, old-school enterprises are learning to “run like Google” with Google’s help.
Is Google afraid of its own message?
And yet Google has shied away from arguably its greatest assets. To the question “Why Google?” the company responds, “Easy to set up. Easy to run. Security built in.” Yes, lower on that same page the company mentions data and analytics and datacenter innovation, but these are stuck alongside security and pricing, as if they’re equivalent differentiators. They’re not.
The thing that Google offers, which no other cloud vendor can, is access to Google’s secret sauce for being Google. To the extent that Google can democratize the technology and processes that let it run at massive scale, all while processing and analyzing data to fuel its advertising business, Google will reap huge rewards. Yes, the company needs to continue to cater to enterprises, and things like data governance and security are a big deal to CIOs.
CIOs, however, aren’t building the future. Their developers are.
Google needs to light up possibilities for those developers, and the lines of business they serve, to do more, to be different, to run like Google. A few companies like Schlumberger have figured it out without assistance from Google’s marketing department, but a key to Google winning big in 2018 and beyond is selling the reality that it uniquely has the cloud to reshape the enterprise in its image.