To add to this - what they probably did was run an nslookup or dig against the domains themselves. And whatever IP resolved for those domains, they checked who owned that IP space. With a CDN, it won't show AWS, it'll show the CDN.
Also, I'm assuming Datanyze used the publicly facing domain of each of these websites (very similar to what I've done). What you'll notice is a lot of these startups have APIs or apps that have very different domains than the ones being used for their landing pages. I spoke to the folks at Bizspark about this and Azure adoption is pretty high (I believe these are private numbers).
I ran a similar study a while back - I was curious to see if the YC incentives for startups (Azure credits, Amazon credits, Digital Ocean credits, etc.) were significantly swaying founders to use a technology stack.
That's correct, but it's very difficult to determine the hosting environment if you're using a CDN. I have a write up on my blog at https://www.gra.pe/?p=36. But the idea is that if you know the hosting IP, you now where the IP sits (ASN). If you determine the ASN, you know which provider is being used.
You're absolutely right. I haven't really read through many of the DO use cases, so maybe I can start there. But, Microsoft also has a separate program for YC ($500K instead of $60k I believe for startups outside of YC). And I know Amazon and Google both have programs that are not exclusive to YC ($100k both). I'm very curious to see how hardware companies are using cloud, and if these incentives are even working.
YC and many other incubators provide startups technology incentive programs (Digital Ocean providing $250k credits, Azure providing $500k in credits, etc.). Are any of the hardware startups here utilizing any of the incentive programs? If so, how so? Very interested to hear about cloud strategies, especially as they relate to hardware companies (APIs, IOTs, etc.).