[_8j50]

They don't need to compete with amazon they just need to be profitable. Their AAD integrations are nice and their portal+ux(including api+powershell) and Defender ATP for E5 customers make them much better than GCP and in some cases AWS. The big three are here to stay.

[cjbgkagh]

They need large amounts of exponential growth, it’s baked into their current value which they need to justify their investments. Microsoft has reoriented itself around Azure and if Azure was merely just profitable Microsoft would have to jettison a lot of what they do. Without another money printer it would be a blood bath.

I think that like Auditing companies there might be a bit of a rotation, or risk spreading among the big providers. Maybe it could help get a discount from Amazon if you could show that you could move to Azure for a little while.

My main worry that their sales will run out of wealthy suckers, and efficiency gains from CPUs may slow down a bit in time.

[_8j50]

My point was Azure can't compete with aws on small/medium size orgs but for large corps who can afford E5 and more it has tremenduous value. Existing windows IT likes azure better. Their main income as a company is still licenses and support which extends to Azure.

[cjbgkagh]

Their prices are targeted towards large corps, I think that by leaving room at the small/medium orgs you're allowing competition to improve to a point where they will go after your large corp customers as well. It also becomes difficult to lower prices because you'd lose more from your existing customers than you'd gain from new customers, so you kind of get stuck at a high price point and slowly bleed out customers who are on the margins. You can make lots of the money in the process and doing this could be a perfectly legitimate strategy. The value adds may be enough to prevent this. I don't know. My datapoint was that it seemed that initial purchases were based on vast overestimates of provisioning and renewals tend to come with significant reductions. This is something that can stay hidden during a period of high grown where there is a lot of onboarding but I would expect to show up in the numbers quite quickly if growth slows with a ~3 year lag.