Google Cloud Next postmortem: Thomas Kurian wants the right people

With the hiring of Tom Kurian, Google Cloud Platform gains an executive highly experienced with running an enterprise computing business. Among his earliest priorities for GCP are hiring the right people – and lots of them. They will be needed in more places than you think.
Written by Tony Baer (dbInsight), Contributor

Last week's Google NEXT conference could be dubbed CEO Thomas Kurian's coming out party. With the first major changing of the guard at the CEO suite, the baton has passed from Diane Greene, whose contribution could be summarized as the enterprise whisperer – the one who had the ear of CEOs and CFOs who taught the Google Cloud organization about concerns of the enterprise.

Now it's time for the next step: build the right organization. Barely three months ago, the baton passed to Kurian. With his track record at Oracle, he knows what it takes to build an enterprise computing business, and he has carefully prepared bulleted points for doing so.

At the top of Kurian's list is having the right people in place. The Google NoOps approach that distinguishes its technology infrastructure may be great for automating what it takes to deploy and run cloud services such as a database, application hub, or machine learning service, but when it comes to entering long-term agreements with enterprise, you need empathetic, thinking, and knowledgeable human minds. The flipside of the challenge is evolving an engineering-driven organization into a business-driven one. Some, like Microsoft and Oracle, have made the leap; others, like Sun, did not and wound up relegated to the dustbins of history.

The same applies to enumerating the terms of engagement and delivering the front-line support. While Google has been best known for the exciting stuff, like innovations such as TensorFlow or Kubernetes, making the transition from technology innovator to reliable business partner requires a lot more mundane blocking and tackling. And so, among Kurian's year one goal is bulking up enterprise go to market and front-line support, not by some double-digit percentage, but by geometric scale. Finding the right people doesn't come overnight. Another goal is hammering out contracts designed for CFOs and corporate legal departments, as opposed to DevOps engineers. And by the way, GCP must pull off each of these fetes while avoiding distracting itself from the continued tasks of filling out its portfolio and cultivating partnerships.

Partnerships are key to Google boosting its penetration of the enterprise cloud market from the few percentage points to the double digits to make it a credible competitor for AWS and Azure. A thriving partner ecosystem will be key to Kurian's goals for growing GCP's solutions business, that is still at the early stages for horizontal offerings, to going vertical.

GCP kicked this off last year with the Contact Center AI solution that packaged its AI capabilities for a common business challenge across industry sectors: engaging the customer and preventing churn. At NEXT, GCP announced a Retail solution that encompasses e-commerce hosting; peak seasons operation support for surviving Black Friday; real-time inventory management; vision product search; and collaboration capabilities from G Suite; and other capabilities. These capabilities provide the core foundation, but for GCP's retail solution to gain traction, it will require a rich partner ecosystem to fill specific niches requiring higher levels of domain knowledge, or for targeting niche subsegments such as specialty retailers.

Building the ecosystem was also very much the driver for GCP's open source hub strategy announced last week. The initiative is all about GCP partnering – and sharing revenue – with popular open source platforms. The ones included in this tranche are Confluent, DataStax, Elastic, InfluxData, MongoDB, Neo4j, Redis Labs. It's hardly coincidence that these are the very companies that also devised their own iconoclastic open source licenses over the past year, which we commented on a few weeks back.

Big on Data bro George Anadiotis gave a detailed discussion last week on whether this strategy could, in his words, save open source.

While the initial impression might seem like a manifestation of Google's original "Do no evil" strategy to paint a contrast to AWS, the reality is that this is more about GCP building customer loyalty as the preferred destination of the open source platform installed base who want the best of both worlds: the full open source or open core stack of these platforms and the flexibility of having a choice of cloud providers. It's another way to overcome the syndrome of enterprise customers about nobody getting fired for buying from the usual suspects.


With his track record at Oracle, Kurian knows what it takes to build an enterprise computing business.

The good news is that this is not just a Kumbaya moment with open source providers who are undergoing their share of competitive angst. Google Cloud is taking the first steps to make these managed cloud services first-class citizens on its cloud platform, not just by listing them in its marketplace, but offering unified billing and management through the common GCP console.

But to make this compelling, GCP needs to deliver a 1 + 1 = 3 value proposition for each of these services. The cold hard reality is that, as independent players with a tiny minutia of Alphabet's market capital, the Mongos, Confluents or Neo4js of the world will never be able to match the economies of scale in their own managed clouds that the household names could. That puts limits on their abilities to globally expand, support a wide range of instance types, or go head-to-head in meeting or undercutting the pricing of their deeper pocketed third-party rivals.

Going native with GCP opens the possibility to get that global footprint and expand their offerings by leveraging the growing portfolio of products and services that it already offers, all under the umbrella of joint support with GCP. For instance, there could be native integrations with Cloud Dataflow (or Confluent Cloud Kafka) that could make these platforms ready targets for live streamed data. There could be integrations with BigQuery or Google Sheets for analytics and reporting, or integration with Google's portfolio of machine learning services.

We're not just talking idly: Microsoft Azure and Databricks have already built the gold standard for exactly this type of business model with Azure Databricks. Offered as an OEM service, Azure Databricks delivers capabilities that go beyond its original offering hosted on AWS. For instance, it has optimized connectors to Azure Blob Storage and Azure Data Lake Storage; the ability to pull data from Azure SQL Database and Azure SQL Data Warehouse; query access from Power BI; and access control through Azure Active Directory.

Keep in mind that for Azure, we are speaking of just a single product; for GCP, there are seven platforms to support and integrate. So, the task of building such offerings won't be trivial – it would be the equivalent of launching seven new SKUs for the portfolio. GCP is hardly short of cash to pull this off – it is investing $13 billion in building new data centers in the U.S, alone over the next year. But, guess what? If you read this post from the top, the challenge will sound rather familiar: it's about mobilizing, recruiting, and prioritizing armies of people to get the jobs done.

Editorial standards