There’s no doubt that Google exists to make money. They make money by getting companies to buy their services. When it comes to selling ads on search engines, Google is number one. When it comes to their cloud business, Google is… well, number three.
I’m guessing that irks them a bit especially since they sit behind a company whose main business is selling whatever stuff people want to sell and a company that made its name in the first wave of PCs. Basically, a department store and a dinosaur are beating them at what should be their game.
Into this environment walks Thomas Kurian, late of Oracle, to transform their cloud business into a real business. Kurian’s marks are all over Google Next this year (2019). Take, for example, the endless customer testimonials. Something common at Oracle OpenWorld in the past, Google paraded out a large number of Fortune 500 executives to extol the virtues of the Google Cloud. They all discussed their “journey to the cloud” which were remarkably similar, to the point of it being a repetitive trope. I was thankful that Oracle dialed this back at the latest Oracle OpenWorld and hope that Google will next year.
Aside from Anthos, an addition to the Google Cloud Platform with a fuzzy message (What is it really?), there were few major announcements. There were a variety of product extensions, especially to the analytics and AI portfolio, that seemed like they should have been announced a year ago. They were interesting, nonetheless, helping to make AI, machine learning, and big data more accessible. Nothing earth-shattering but useful.
The same was true for their announcement of SQLServer as a native service. In the past, you would have had to instantiate a VM, and then load and configure a SQLServer database within it. Most Windows applications depend on SQLServer and removing the extra work makes GCP more attractive as a platform for those workloads. Again, this is a convenience but not industry changing like the AI portfolio was in the past.
There were also several enhancements to billing and support for Google Cloud Platform (GCP). The most interesting was the unified billing and support for native Google cloud services and partner products. Now, products from partners can be created just like any Google service and made part of the GCP billing. This is probably a more a boon to partners but, like native SQLServer, convenient for customers.
The caravan of big customers and few blockbuster announcements are indications of the strategy that Kurian is enacting at Google.
Google always catered to DevOps teams. They created software, such as Kubernetes and Istio, that are being widely adopted by the open source community and corporate IT alike. This has been their strength. They talked to the people who really make technology and services decisions in companies – system architects, operations managers, and developers.
Kurian is trying to get Google to act like Oracle used to – speaking to the upper management layers, especially CIOs and CTOs. Clearly, it is hoped that this strategy will generate wider adoption of GCP in major corporations, the type that run up big bills for IT services.
When Kurian was at Oracle, this strategy appeared to make sense since they were starting from Oracle’s perceived strength. Their flagship database underpins the majority of the mission-critical applications and much of their revenue comes from selling ERP, CRM, etc. to big companies. It was, however, a strategy that put them into the bottom half of cloud service providers and was quickly abandoned the minute Kurian left for Google.
The cloud business is different than the on-premises software business. Costs are spread out over time and, hence, less risky than buying millions of dollars’ worth of software, deployed to an equally expensive server farm. This is why most cloud vendors have engaged with DevOps teams more than the top bosses. Decisions are being made on the technical merits of the services. Developers are not going to be swayed by six CIOs discussing their company’s “journey to the cloud.”
When individual software purchases were massive capital expenditures, each purchase required top-level oversight. The cloud shifted this CAPEX to OPEX and into the budget of individual teams. Individual DevOps teams can now make these purchases themselves, choosing what works best for their projects. This is precisely why multi-cloud is happening – individual teams are purchasing resources from different cloud service providers to meet the needs of their teams.
Purchases are no longer top-down affairs. It is also what makes Google’s strategy baffling. They have the mindshare of the DevOps teams that are buying cloud services. They should be building on their strengths in AI, machine learning, and, especially, Kubernetes. Instead, they are adopting a worn-out strategy of wooing top-level executives. It didn’t work for Oracle, so they changed. It won’t work for Google.
For Google customers, there is the danger that innovation will slow. With more focus directed at enterprise deals and the features that are desired by the Fortune 500 crowd, it’s unclear if Google will continue to invest in the advanced products it’s known for. The keynotes at Google Next 2019 suggest this is a real concern. The spotlight was on big enterprise deals and not awesome new services that allow forward-thinking customers to make amazing software. Unfortunately, this is more likely to cause Google to slip in popularity with its loyal base. All we can do is hope that Thomas Kurian will do what he wasn’t able to do at Oracle: bring in enough massive deals to drive up revenue.