My experience leads me to define four reasons:
One.) Cloud is viewed as an IT task and not a cross-functional business driver
Two) There’s still poor understanding of the one-to-one link between architectural choices and runtime costs which leads to the deployment of lots of technically interesting, but financially, um, dramatic designs
Three) The concept of unit economics isn’t widely understood outside of finance and value-chain circles; what does each customer cost you and what’s the ROI? To put this into semi-technical terms, what does it cost you, per customer or service offered, to maintain the Azure web app or IaaS server farm you’re running to meet demand?
Four) To the extent the need for cost management is acknowledged, it’s viewed as purely an accounting-esque activity managed via dashboard tools without insight into cost drivers and contextual factors (full circle back to it being an IT task).
If your Azure cost management efforts remain stuck in the lands of ‘turn it off’ and dashboard gazing alone, you will never get ahead of things…i.e., you won’t be able to forecast which is exactly what everyone wants.
A good guide to the factors you should keep in mind is Duckbill’s “The Unconventional Guide to AWS Cost Management” which, although, it (of course) lists AWS specific opportunities, starts with universally true principles that absolutely apply to getting a handle on your Azure costs
Another excellent place to start is Microsoft’s doc, “How to optimize your cloud investment with Azure Cost Management“.
And of course, keep watching this space as I continue to learn, explain and share what I’ve witnessed ‘in the wild’ about managing Azure costs.
Oh and check out my book on Azure Cost Management.