Trip report: Google Cloud APAC analyst summit (November 2026)
Note: Virtified research is written by humans and 100% independent. Trip reports are not sponsored content. Disclosure: Virtified did travel to Singapore at Google Cloud's expense, which is a typical custom for IT industry analysts.
AI has taken the spotlight from hyperscale cloud, and the impact on hyperscalers is evident. Each is now competing to win in the AI era and each is taking a different approach. Enter Google Cloud and the chance for the APAC analyst community to hear directly from their APAC leadership. The event took place on November 3.
The story so far
During the cloud wars of the last decade, Google Cloud struggled to beat rivals AWS and Microsoft—for the enterprise IT market. There were 3 key reasons.
In short summary, Google Cloud:
Took longer to establish its footprint relative to AWS and Microsoft. This issue is now resolved, for the most part.
Spent too much time talking about its technology and less time understanding enterprise needs. A cloud native engineering mindset meant it was enamoured with its own brilliance (at the expense of understanding customer requirements).
Engaged with enterprises in a relatively immature manner that often left prospects feeling concerned. Partnerships and alliances were fewer, less established and hamstrung by commercial dealings that were less polished.
Although none of these issues is completely resolved, Google Cloud had a more fundamental problem: it was also fighting its competitor's fight.
Success as a hyperscaler does not guarantee success as an AI vendor
Key takeaways
Google Cloud's cloud-only (a) strategy leaves it with a lesser attack surface with greater control. Relative to Microsoft, this is a major difference. As example, securing an on-premises Copilot-enabled SharePoint deployment running on Windows Server 2019 is a much more complex challenge than Gemini with Google Drive. Anthropic and OpenAI can be considered similarly, for good and bad. Although they don't struggle with legacy, deploying to the target device adds functionality but adding an on-device agent introduces risk. Virtified asserts that agent isolation is far from mature and will remain so through at least 2029.
Google is the only (Western) hyperscale cloud vendor that can confidently claim to be at the forefront of frontier model development. AWS and Microsoft will claim that their approach enables choice, and it does, but Google Cloud can also do this. They also have a long history of investing in (pre-generative) AI. Moreover, the expertise acquired from DeepMind is an ongoing strategic advantage compared to its IaaS/PaaS-centric competitors.
As a result of (2), alongside the existing cloud services in its portfolio, Google Cloud is the only (Western) hyperscaler to offer a 'full stack' (b). This improves Google Cloud's relevance for enterprise needs in the AI era.
Google Cloud's 3Q26 earnings are solid evidence of this evolution. Increasing revenue provides a 'tailwind' and Alphabet will need this to continue. It knows it must adjust its business model away from being dependent on advertising revenue and an attractive target for capital investment (as announced in its June 2026 announcement of an $80B equity capital raise).
Footnotes:
a. Virtified chooses not to quibble about Antigravity in this post.
b. This assumes the 'stack' stops before the application layer. If that's included, Oracle is best-positioned to make the claim.
