Next '19: 451 Research analyst key insight roundup
April 23 2019
by Matt Aslett, Owen Rogers, Nick Patience, Chris Marsh, Fernando Montenegro, Scott Crawford, Christian Renaud, Jeremy Korn
Google's Cloud Next customer event has provided us with regular opportunities in recent years to ponder whether and how the company would be able to translate its popularity with developers into enterprise adoption. Indeed, this has been a recurring theme at the event for Google itself, and this year, there was an increased focus on the company's enterprise sales strategy following the appointment of former Oracle executive Thomas Kurian as CEO in November 2018. As expected, Kurian emphasized the changes he is making to Google Cloud's culture and sales approach to better engage with enterprise customers. However, since Kurian has only been at the helm for a few months, the event also provided evidence of strategic changes that were already underway prior to his appointment in terms of the company's positioning of the various Google Cloud Platform services.
Among the 30,000-plus attendees were 451 Research analysts across multiple channel disciplines who view Google from their own perspectives. This report provides a high-level take on what they think are the most interesting developments from the conference across hybrid- and multi-cloud, AI and machine learning, data and analytics, security, workforce productivity, and IoT.
The 451 Take
Having attended all three of the Google Cloud Next events (as well as the preceding GCP NEXT event in 2016), we have found it interesting to watch the event become increasingly enterprise-focused while the Google Cloud business attempts to maintain its engagement with developers and startups. While the latter are still able to experiment to their heart's content using Google Cloud services, there is a clear strategy shift from the company to focus its sales teams on major deployments that will be taken into production at scale. This will increasingly involve Google's internal engineers engaging more directly with enterprise customers, particularly for AI projects, to build playbooks for repeatable, transformational use cases. To do so will likely require the company to engage more fully with consulting and service providers, as well as build up its own professional services organization.
Key analyst takeaways
Overall enterprise strategy
While Google Cloud undoubtedly has had some success with enterprise customers over the years, the flagship customers, such as Spotify, have tended to be digital-native. Deliberate attempts to refocus its sales strategy landed more traditional flagship accounts including HSBC and Airbus, but it is fair to say that engaging with enterprise customers while maintaining Google's relationship with developers has not been an easy balancing act. New CEO Thomas Kurian outlined how the company is hiring more enterprise sales staff, but that is only one aspect of how Google is changing – and had already begun to change prior to Kurian's appointment – to address enterprise customers. For example, the company is now focused on five key priorities (infrastructure modernization, data management, application development, smart analytics and productivity/collaboration) and six industry verticals (financial services, healthcare, retail, manufacturing, government, and communications, media and entertainment).
Another notable change is a more empathetic approach to meet customers where they are – accepting that while many potential customers might like to – run like Google,' legacy on-premises investments combined with adoption of other cloud services often don't make that possible. The launch of the Anthos hybrid cloud platform for on-premises and multi-cloud application development and management (more on this below) is a good indication of that strategy change, while there has also been a detectable change of emphasis toward describing Google Cloud Platform less as a destination and more as an engine for digital transformation. There is also a greater focus on use cases and solutions' rather than stand-alone products/services. This should serve the company well, particularly in relation to AI and ML, although it will also likely require an increased investment in consulting and professional services.
– Matt Aslett, Research Vice President, Data, AI and Analytics
Hybrid- and multi-cloud
Anthos – the new name for Google's Cloud Service Platform – has the potential to be a big differentiator for Google. Anthos is a hybrid cloud management service based on Kubernetes, the technology at the very heart of Google's cloud proposition. At the keynote, Google boldly demonstrated its attitude to openness with a demo that showed containers being managed on AWS, its big competitor, and it addressed enterprise credibility with partners including VMware, Dell, Cisco, HPE and Lenovo all wanting to be part of the Anthos story. In fact, Kurian was up front about the fact that it is hiring new sales engineers to target enterprise requirements. Shortly after the announcement, Google released documentation, architectural diagrams and an FAQ that show how Google Kubernetes Engine (GKE), Config Management, Stackdriver, Cloud Build and other services together can deliver a hybrid experience to enterprise customers. But the big missing component from the website is the software itself. Potential clients will need to engage with Google to access the software and, once in operation, be billed according to the number of vCPUs managed by Anthos, consumed in blocks of 100. For a company that traditionally sells products as consumable services, supporting on-premises software and billing on a license model is a significant cultural shift that will require quality sales and support personnel, along with robust internal processes, to pull off. As such, while Anthos is certainly something enterprises should keep an eye on, we wouldn't recommend jumping in head first.
– Owen Rogers, Research Director, Digital Economics Unit
AI and ML
In terms of its ever-expanding AI and machine learning portfolio, Google Cloud released several new products including AutoML Tables, AutoML Video and AutoML Edge Vision. These automated machine learning tools should empower non-experts to, respectively, build predictive models from structured datasets, construct video classification models and perform image analysis in edge environments. Google Cloud also announced AI Platform, a data science development environment that unifies Google Cloud's suite of AI and machine learning tools with additional features and integrations for accelerating the deployment of AI projects into production. Above the platform sits the newly released AI Hub, an enterprise-grade collaboration tool through which developers can access plug-and-play AI components or share projects or pipeline components across their organization. Taken together, these tools and platforms should allow Google Cloud to serve the AI and machine learning needs of both individual developers and data scientists working on complex projects across the enterprise. Finally, in line with the aforementioned shift of Google Cloud's strategy toward specific industries, Google Cloud announced a series of horizontal and vertical AI software offerings. These off-the-shelf products include Contact Center AI, Document Understanding, and two retail offerings: Recommendation AI and Visual Product Search. They represent how Google Cloud plans to roll up its AI and machine learning tools as ready-made packages with easily articulated value that will catch the eyes of business users and C-level executives looking to digitally transform their businesses.
– Nick Patience, Founder & Research Vice President, Software; Jeremy Korn, Senior Research Associate
Data management and analytics
Data processing and analytics is a key differentiator for Google Cloud, thanks to services such as Google Cloud Spanner and Google BigQuery, and it is notable that data management and analytics constitute two of the company's five priorities. Data and analytics were, therefore, well represented in the announcements at Google Cloud Next '19, including the launch of Google Cloud Fusion, Google Cloud Data Catalog and Google BigQuery BI Engine, all of which will be addressed in more detail in a subsequent report. In relation to Google's strategy to meet customers where they are, another significant announcement was the expansion of the Google Cloud SQL managed database service to support Microsoft SQL Server, as well as the accompanying managed service for Microsoft Active Directory. Google Cloud also announced several strategic partnerships with leading open-source-centric database companies – Confluent, DataStax, Elastic, InfluxData, MongoDB, Neo4j and Redis Labs – to offer managed services that are tightly integrated into GCP with unified billing, management and support. There was also a hint of interesting announcements to come, with Kurian acknowledging that the new approach to hybrid- and multi-cloud involved the evaluation of whether it might make sense to make some Google-only database services more broadly available.
– Matt Aslett, Research Vice President, Data, AI and Analytics
The G Suite team was more bullish this year in putting out a vision of investment areas and its long-term vision for enabling the future of work. Several of the product managers said that the company leadership team is openly encouraging the G Suite, as one of Google's smaller teams, to articulate its voice and vision more publicly. Its portfolio story has grown stronger, drawing on a series of important announcements including the general availability of its cloud telephony product, Google Voice, which rounds out its Unified Communications offering, and the launch of connected sheets, which hooks into BigQuery and which the company is pitching as a way to democratize the complex analysis of data. Google is creating a more seamless experience across G Suite with Hangouts Chat integrated with Gmail, Google Assistant integrated with Calendar, and its new(ish) Cloud Search third-party connectivity available for eligible G Suite enterprise customers. It has also repositioned what was Google+ as Currents, a kind of knowledge-sharing capability, and announced a series of what it calls assistive tools' supporting the meeting experience. ML was a big part of the overall G Suite story as it looks to iteratively layer in intelligent capabilities to reduce the still very pervasive friction across daily work scenarios.
– Chris Marsh, Research Director, Workforce Productivity and Compliance
More than 30 security-related announcements were made at this year's Google Cloud Next, but one of the more provocative strategically was not a security announcement per se, nor, in fact, was it an announcement of new functionality so much as a rebranding to emphasize its role in Google's cloud strategy going forward. Anthos is the new brand given this week to what had formerly been Google's Cloud Services Platform. Powered by Kubernetes (which arose as a Google innovation), Anthos represents Google's strategy for enabling customers to deploy workloads across any compatible environment and is, thus, a key aspect of Google's multi-cloud strategy. Policy definition and enforcement and the concept of security guardrails' are central to Anthos and reflect values similar to AWS ControlTower. These offerings are key indicators of a trend we expect to increasingly dominate the security market: the disruptive role cloud hyperscalers are playing in defining security integral to emerging architectures, potentially threatening the separate, product-oriented strategies of security's legacy leaders.
– Scott Crawford, Research Director, Information Security
Throughout the conference, Google consistently placed security as a key focus area and, in some cases, a key differentiator, and followed it up with roughly 30 announcements related to security, covering both GCP and G Suite. In many cases, the announcements covered not so much new functionality as the general availability of existing offerings, such as the Cloud Security Command Center, VPC Service Controls and Shielded VMs. Still, there were interesting new developments, such as access approval workflows for GCP back-end access by Google employees, support for Android phones as security keys, machine learning for IAM access control recommendations, and leveraging Google's SafeBrowsing initiative for external customers.
– Fernando Montenegro, Senior Analyst, Information Security
IoT was noticeably absent for most of Google Cloud Next, although the company did showcase its renewed enterprise focus on industry verticals, which from an IoT context includes smart cities, manufacturing, transportation, energy, retail and education. Because of the cloud-first focus of Next '19, edge computing was, not surprisingly, also given very little airtime (aside from the aforementioned AutoML Edge Vision). The IoT security message at the show was 'IoT is another application among many, so the broader Google security architecture components, including behavioral analytics of device traffic, encryption and device agent updating, apply equally to IoT devices.' This approach paints a picture of IoT devices being natively connected and not compute-constrained, which will eventually be true but is not yet the case in brownfield/legacy device-heavy IoT verticals such as manufacturing, energy and transportation.
– Christian Renaud, Research Vice President, Internet of Things