Google Cloud Next 2017: Slow and steady race to greater enterprise public cloud adoption
Google recently held its Cloud Next 2017 user conference in San Francisco, with 10,000 in attendance. A late arrival to the enterprise cloud computing market, Google showcased its attractive capabilities for migrating traditional enterprise development and workloads to the Google Cloud Platform, investment in professional services, new partnerships, and geographic expansion plans in an attempt to compete with other dominant hyperscalers, such as AWS and Microsoft Azure.
The 451 Take
Google needs to penetrate enterprise IT organizations if it wants to gain a significant foothold in the public cloud sector. It has compelling technology, and is rapidly building out the enterprise-ready features needed to get business customers in the door. It will then look to retain enterprise customers based on innovation and partner-driven digital transformation offerings that incorporate Google's company-wide capabilities and financial resources. However, its major competitors are not slowing the rollout of their new features and functionalities. The initial signs for Google's success among enterprises are promising, but the company's ultimate success is by no means assured.
Google is serious about winning enterprise business in the public cloud space. This includes cloud native workloads and legacy workloads. Per the company, its Google Cloud organization, which includes its datacenters, cloud infrastructure, G Suite, Analytics and API platform, spent $29bn in capex over the last three years, and added more staff than any other part of Alphabet. It invested in go-to-market teams, professional services, training, certifications and its partner ecosystem in order to win over enterprises. Google is also investing in migration tools and industry offerings while increasing its focus on developers. It claims to have had the highest uptime availability in 2016, and that its longest outage was 18 minutes a few years ago – a clear swipe at AWS and its recent S3 outage.
In 2016 Google paraded its enterprise cloud wins from Coca-Cola, Disney Interactive, Snapchat and Spotify in front of attendees. This year multinational corporate customers, including HSBC, Colgate-Palmolive, Home Depot, The New York Times, Schlumberger, SAP and Verizon, were prominently on display at the conference. Evernote moved from an on-premises offering to Google Cloud in 70 days, and Carousell, a fast-growing mobile marketplace app in Asia, moved from a competing public cloud in 30 days. EBay now has a parallel environment on Google Cloud for its live catalog of one billion listings, and integrates with Google Home.
According to company executives, 2017 will be a period of 'accelerated growth.' Across the Google Cloud Platform there are one billion end users and 100 billion daily app engine requests. Google says it is winning 50% of deals against competitors (an upward trend). It further claims a threefold year-over-year increase in Google Cloud Compute usage, and says data analytics is often the first workload for enterprises on Google Cloud. While thriving in the retail vertical, it admits an uncomfortableness competing in regulated verticals like banking and healthcare. The company claims that one reason enterprises choose its cloud platform is that it is a 'pure' cloud company without legacy baggage – a clear swipe at Microsoft. Reasons for its losses include people questioning the company's seriousness of being enterprise-grade, siloed sales teams, channel conflict and limited geographic footprint. At the conference, Google articulated a plan to address these concerns, and we detail some of those in this report.
A major pain point for enterprises considering a broad shift to cloud is the risk of moving functioning applications, data and workloads onto a new platform. During the past year, Google has put into place the security features and high-touch engagement model that enterprise customers expect. At its conference, the company revealed a number of new initiatives to smooth the migration process at the data, application, infrastructure and support levels. Among the announcements were:
- Cloud Dataprep (beta) – an intelligent data preparation tool that is a Google variant of Trifacta, a data-wrangling tool that uses a point-and-click interface to explore, cleanse and convert data in preparation for processing.
- BigQuery Data Transfer Service (beta) – a fully managed data import capability to automate the transfer of data from SaaS applications into BigQuery; it is pre-integrated with DoubleClick and Adwords for faster time to insight.
- SAP partnership – HANA is now GA-certified on Google Cloud Platform, and HANA Express Edition is now available in Google Cloud Launcher.
- The expanded machine-learning portfolio.
- Cloud functions (beta) – serverless, event-driven microservices.
- Windows support (GA) – support of SQL Server Enterprise Edition on Google Cloud Platform (GCP), including prebuilt images and bring-your-own-license support, claiming that database workloads can run faster on Google's infrastructure than on Microsoft's own Azure cloud.
- Windows Partner Program – increasing its engagement in the .NET ecosystem in partnership with a handful of global systems integrators, including Capgemini, CI&T and SADA Systems.
- Free VM transfer service – DIY migration of VMs running supported Windows and Linux operating systems from either on-premises environments or other public clouds.
- Engineer-to-engineer support – a per user fee per month for access to Google's engineers and CRE partners. Coming in Q317.
- Rackspace managed services – useful for those enterprises looking for a partner to co-manage a GCP environment.
Note: A number of other Google announcements were made at Next 2017, and will be covered in other 451 Research reports in detail, including G Suite, Data Analytics and Machine Learning.
Google believes that a hybrid architecture will persist in the coming years as enterprises continue to migrate workloads to various clouds. Its hybrid cloud architecture revolves around its virtual private cloud. Google VPC is an instantiation of GCP that can dedicate compute, storage and network resources to an enterprise. It is built upon Google's proprietary private global network designed for high reliability, low latency and hardened security. Kubernetes acts as the orchestration and operational backplane for hybrid implementations. Elasticity and scale are achieved by linking to Google public cloud services.
Connectivity to VPC is enabled through IPsec tunneling, or a direct connect if the user is the vicinity of one of Google's 180 global points of presence. Other means for integration between on-premises resources, Google VPC and its public cloud services include its new 350-customer strong Apigee Hybrid Gateway. Google also acquired integration vendor AppBridge to link on-premises resources to Google's enhanced G Suite office productivity tools and other SaaS offerings.
Investment in sales and channel
Wooing enterprise CIOs rather than developers is not a simple pivot, and requires a major global investment in sales, professional services and channel marketing. There are now 1,000 Google employees in the field. Salespeople and customer engineers are paired on sales engagements. Workforces for G Suite and GCP sales are now merged in an effort to have more strategic conversations with CIOs and IT decision-makers. There is now one territory and quota shared by inside sales and the field sales reps.
Google also has revamped its channel program and tech partners. Now its own sales people will get paid to work with partners and customers, and no longer have to worry about channel conflicts in the agnostic model. Currently, Google says it has 13,000 partners and one million paying customers through partners. In fact, it is closing three times more enterprise customers through partners than in the past. Its partners are categorized as global SIs (e.g., Accenture, Capgemini, PWC), regional and cloud-first SIs (e.g., Cloud Sherpas), online resellers (e.g., CDW, SoftBank), and telco/scaled (e.g., Endurance International, Verizon, Squarespace).
Professional services also represent an important investment focus for Google. To date it has trained 66,000 'students' in 21 countries. It is a team of 250 customer engineering specialists, 'solutions architects' and customer reliability engineers (CREs) that are now organized into regional pools. Proofs of concept are a big focus, but engagements also include deployment advisory, implementation support and thought leadership. CREs, invented by Google, are part of a customer's on-site team, but have access to Google internal tools and teams. At the conference, Pivotal was announced as a CRE partner.
At Google, a region is a specific geographic location where customers can run resources, and each region has one or more zones. Google currently operates six regions (Oregon, Iowa, South Carolina, Belgium, Taiwan and Tokyo). At the 2017 conference, the company announced a few new datacenter builds, and the current list of all pending regions now includes: Sydney, Sao Paulo, Frankfurt, Mumbai, Singapore, London, Hamina (Finland), Montreal, Northern Virginia, California and the Netherlands.
Azure now has 34 global regions around the world, with France (pending), Korea, China and US government regions being notable differences. AWS offers 42 separate availability zones across 16 global regions in many of the locations where Google has made plans. Having a local presence is important for enterprises with data sovereignty requirements, especially in Europe and Asia, and the reason for the rapid expansion by all providers.
According to 451 Research's Market Monitor service, AWS has a run rate of $14bn. It is the provider that nearly every enterprise will evaluate when considering a workload shift to the public cloud. While Google remains a distant third in the public cloud race, momentum is building around Google Cloud – according to 451 Research's Voice of the Enterprise Cloud Transformation Vendor research, adoption of Google Cloud platform rose to nearly 16% in 2016, compared with 8.8% at the beginning of 2016, and GCP received particularly strong marks for service reliability and value for money/cost.
Google's increased pace of iteration is key to its competitiveness in the IaaS space. AWS and Microsoft release hundreds of updates and new services every year, with development times measured in weeks at AWS, and somewhat longer at Microsoft. Both of these two main rivals have mass consumer outreach combined with a global presence. Competition specifically for enterprise customers is intense, with IBM, VMware and Oracle also looking to sell software and services to enterprises.
Google has highly advanced and efficient datacenter operations at a scale that few can match, and its software services are equally innovative. It also has practically unlimited freedom to invest in its platform efforts and hire at will.
The company operates a consumer-driven business, and enterprises have different needs and requirements to address.
The opportunity for cloud infrastructure is growing fast – it is already in the tens of billions of dollars.
Google's major threat in IaaS is being too little, too late with its enterprise-friendly outreach efforts.