[manyhats]

Suggest that readers look at cloud revenue from a business model perspective, not a technology one.

Cloud revenues between Amazon and MS are comparable from an investment point-of-view:

- sticky (unless something bad happens, if you're a paying customer in month 1, you're probably still a paying customer in month 2)

- service delivered "in the cloud" (neither vendor needs a local brick/mortar storefront)

- fixed cost: data centers, variable cost: "things" in data centers

- for enterprise accounts: high-touch salesforce and dedicated contacts

- for end-user / SMB accounts: primarily self-service via online tools

Yes, one of them gets more revenue from IaaS and the other from SaaS. Each of these, and the specific submarkets that the two of them play in, will have different growth rates and potential and affect their respective valuation.

Still, from an investment standpoint, based on the above characteristics, I'd feel reasonably good about comparing the two of them on financial and valuation metrics for the "cloud" parts of their respective businesses.

[timsneath]

This an insightful post -- but I think there's one key characteristic missing that Wall Street cares about still more, and that's growth.

Traditional products like Office 365 may be sold as a subscription, but the market is largely saturated. Its growth is primarily dependent on productivity device growth. "True" cloud products, whether serverless or lift and shift of existing workloads, have much higher growth potential since a large percentage of the attachable market is still unreached.

That's why it's a little disingenuous for Microsoft to claim products like Office 365 and Dynamics 365 as cloud revenue. For many years before the cloud was a thing, they earned that revenue in the form of Enterprise Agreements: which are essentially three-year subscriptions for the same products.