1) The IAAS game has become table stakes. It's no more the differentiator. AWS leads here given the headstart they had in the cloud game. Their VM's are stabler with more variety than GCP and Azure. The latter two are also improving q-o-q (have had first-hand experience with both AWS and Azure, anecdotal evidence for GCP).
2) In the managed services market, all three have their areas of expertise: 1) GCP for ML/AI/Data Processing and the control planes (GKE et al). They also lead in running workloads requiring large amounts of cross-region data sync due to the petabit network lines they own (I believe Facebook is the only other company that owns a dedicated line, AWS is planning to build one). 2) Azure in the data analytics space given the flagship products like SQLServer, PowerBI. Retailers just love these products :) 3) AWS has the edge in newer paradigms like serverless (AWS Lambda, Step Functions). GCP and Azure are catching up.
3) Sales - Azure is by far the leader here, given Microsoft's good handle on enterprise sales. It's evident from the large deals they have been able to close in the last few quarters. AWS is catching up here. GCP is still clueless. This stems largely from the fact that Google has been for a long time an engineering-driven org rather than business-led. They have always been thinking "on behalf" of the customer instead of "for" the customer. It's a big mindset change for them.
4) Customer Support - Amazon leads here by far. Bezos' customer-obsession runs deep in their veins. Azure is catching up quick though. They still run in the mode of making the customer believe that they are working to resolve the problems rather than providing actual solutions. But I think this will be sorted soon. GCP is yet to figure this out.
Regardless, the next decade is going to be really exciting in this space, given AWS has started to lose the first-mover advantage they had until a few years ago. With the massive war-chests Google and Microsoft have deployed, it will be critical for AWS to innovate faster than before.
Nice read ! Matches quite a bit with my thoughts. Another advancement to watch out for is the whole edge cloud. Mobile edge computing (Partnership between cloud providers and telcos, AWS Wavelength) will take off with the adoption of 5G. https://www.sdxcentral.com/articles/news/south-korean-operators-tout-skyrocketing-5g-subscriptions/2020/02/. At present, it is gearing towards low latency deployments with cloud computing at the telco premises. My personal view is that, in future, this will extend towards private edge computing in the context of enterprises. If that happens, the hybrid cloud, as we know today, might get changed significantly before it matures. Application architectures, even enterprise on prem ones, will have to evolve.
Super insightful, Krishna. What do you think are the real life drivers to think “multi cloud” for enterprises apart from vendor lock-in ? Specifically for B2B SaaS companies.
Each vendor also has different strengths. E.g. GCP for AI/ML and Azure for the windows ecosystem. Sometimes you may need to choose the best of breed for specific workloads.
Nicely summarised KK! Few thoughts:
1) The IAAS game has become table stakes. It's no more the differentiator. AWS leads here given the headstart they had in the cloud game. Their VM's are stabler with more variety than GCP and Azure. The latter two are also improving q-o-q (have had first-hand experience with both AWS and Azure, anecdotal evidence for GCP).
2) In the managed services market, all three have their areas of expertise: 1) GCP for ML/AI/Data Processing and the control planes (GKE et al). They also lead in running workloads requiring large amounts of cross-region data sync due to the petabit network lines they own (I believe Facebook is the only other company that owns a dedicated line, AWS is planning to build one). 2) Azure in the data analytics space given the flagship products like SQLServer, PowerBI. Retailers just love these products :) 3) AWS has the edge in newer paradigms like serverless (AWS Lambda, Step Functions). GCP and Azure are catching up.
3) Sales - Azure is by far the leader here, given Microsoft's good handle on enterprise sales. It's evident from the large deals they have been able to close in the last few quarters. AWS is catching up here. GCP is still clueless. This stems largely from the fact that Google has been for a long time an engineering-driven org rather than business-led. They have always been thinking "on behalf" of the customer instead of "for" the customer. It's a big mindset change for them.
4) Customer Support - Amazon leads here by far. Bezos' customer-obsession runs deep in their veins. Azure is catching up quick though. They still run in the mode of making the customer believe that they are working to resolve the problems rather than providing actual solutions. But I think this will be sorted soon. GCP is yet to figure this out.
Regardless, the next decade is going to be really exciting in this space, given AWS has started to lose the first-mover advantage they had until a few years ago. With the massive war-chests Google and Microsoft have deployed, it will be critical for AWS to innovate faster than before.
Nice read ! Matches quite a bit with my thoughts. Another advancement to watch out for is the whole edge cloud. Mobile edge computing (Partnership between cloud providers and telcos, AWS Wavelength) will take off with the adoption of 5G. https://www.sdxcentral.com/articles/news/south-korean-operators-tout-skyrocketing-5g-subscriptions/2020/02/. At present, it is gearing towards low latency deployments with cloud computing at the telco premises. My personal view is that, in future, this will extend towards private edge computing in the context of enterprises. If that happens, the hybrid cloud, as we know today, might get changed significantly before it matures. Application architectures, even enterprise on prem ones, will have to evolve.
Super insightful, Krishna. What do you think are the real life drivers to think “multi cloud” for enterprises apart from vendor lock-in ? Specifically for B2B SaaS companies.
Each vendor also has different strengths. E.g. GCP for AI/ML and Azure for the windows ecosystem. Sometimes you may need to choose the best of breed for specific workloads.