He's saying that cloud technology is only just beginning and the bazillion dollars that cloud and cloud related services are currently worth is just the start.
I think that "operational analytics" covers a portion of what the OP meant: allowing data analytics to help run and make business decisions at a day to day level.
> Using amazon or azure to host your service?
Yes, the the cloud helps businesses understand, manage costs, and scale without the headaches of haggling with Oracle or enterprise providers. Also, less in-house infrastructure labor needed. Cloud is even easier if a data service provider is managing it for you.
> Charging by subscription rather than upfront?
Yes, SaaS margins and cash flow structure are something investors love about the data services space. SaaS is not going away.
> Mining your customer's data for fun and profit?
Well, enabling your customer companies to mine their own data for their fun and profit, with less engineering labor and more operable outputs.
I'm not sure there's any substance to those beliefs. The push for cloud-agnostic solutions is also a thinly veilled push for self-hosting services, or at least parts of it. Cloud services charge a huge markup for stuff that comes for free somewhere else, like certificates or bandwidth or running your services in dedicated hardware.
It's not possible to ignore the fact that "cloud" is a marketing term for renting someone else's hardware, and beyond a certain scale it's always better to run our code in our hardware.
> beyond a certain scale it's always better to run our code in our hardware.
When OP says "We've only scratched the surface of the cloud-enabled TAM." I think he's referring specifically to below that scale, where they mostly don't have in house operators.
Of course there is room for self hosted solutions, which will continue to grow as well.
But if you think we are anywhere near peak-cloud then I would be eager to understand your logic. The global cloud market is in the midst of a ~25% CAGR run that shows few signs of slowing. Companies that need mindcrushing amounts of compute are being founded daily in sectors that remain nascent (AI, IoT, space exploration, bio-sciences). Steaming services, data processing, etc etc etc.
You're REALLY going to buy a few racks of PowerEdge servers and call it a day day? Of course not. Easily accessible, well managed cloud services are the only sane option.
And that doesn't even touch on all the barnacles on the hull of that ship (SaaS, Dev Tools...)
> But if you think we are anywhere near peak-cloud then I would be eager to understand your logic.
I have no idea what you mean by "peak cloud".
I do know for a fact that sizeable tech companies are already moving their services away from cloud providers because it makes absolutely no sense to pay small fortunes per month, sometimes costing more than the hardware it takes to run it, just to run stuff on a cabinet you are quite able to afford.
I also know that the bulk of these cloud services cater to cargo-cult customers plagued with FOMO.
Most organizations from startups to enterprises generally benefit from the cost, scalability, security, technical support, and other benefits of cloud storage and computing.
Startups, for example, can focus on development and distribution when using cloud based tech for at least some part of their stack; and needless to say enterprises can scale quickly and securely.
Within this spectrum, where and what is this “sizeable” company that benefits from doing all this themselves?
Peak-cloud - full saturation with no Market left to penetrate
Of course sizeable companies of all sectors are using hybrid. Always been the case. Both things can be and are true: owned is growing. Cloud is growing
The bit about not being satisfied with clocking in to Google... I've been fortunate enough to have been approached by some of the big tech companies, but it never appealed. Even if I survived the interviews and had an offer, I have this hunch that working there is ultimately a trap. I see older people on linkedin posting pictures of hanging up their badges after 15 years, and I can't help but think they were kept there by really comfortable circumstances. But the future isn't at these big comfortable companies.
part of the reason i think was that there was simply to much money floating around, between crypto and almost any idea being funded, you could cash in on the ex googler fact and get more money then staying at google.
Cloud worked at the time because building your own infrastructure was a pain. Then when it got cheaper to host your own, most people in IT ended up being millennials that have never seen a real server, only containers and aws.
18 comments
[ 6.0 ms ] story [ 73.7 ms ] threadWhat is this (data-fueled cloud-enabled TAM)?
I agree 100%
Using amazon or azure to host your service? Charging by subscription rather than upfront? Mining your customer's data for fun and profit?
> Using amazon or azure to host your service?
Yes, the the cloud helps businesses understand, manage costs, and scale without the headaches of haggling with Oracle or enterprise providers. Also, less in-house infrastructure labor needed. Cloud is even easier if a data service provider is managing it for you.
> Charging by subscription rather than upfront?
Yes, SaaS margins and cash flow structure are something investors love about the data services space. SaaS is not going away.
> Mining your customer's data for fun and profit?
Well, enabling your customer companies to mine their own data for their fun and profit, with less engineering labor and more operable outputs.
It's not possible to ignore the fact that "cloud" is a marketing term for renting someone else's hardware, and beyond a certain scale it's always better to run our code in our hardware.
When OP says "We've only scratched the surface of the cloud-enabled TAM." I think he's referring specifically to below that scale, where they mostly don't have in house operators.
But if you think we are anywhere near peak-cloud then I would be eager to understand your logic. The global cloud market is in the midst of a ~25% CAGR run that shows few signs of slowing. Companies that need mindcrushing amounts of compute are being founded daily in sectors that remain nascent (AI, IoT, space exploration, bio-sciences). Steaming services, data processing, etc etc etc.
You're REALLY going to buy a few racks of PowerEdge servers and call it a day day? Of course not. Easily accessible, well managed cloud services are the only sane option.
And that doesn't even touch on all the barnacles on the hull of that ship (SaaS, Dev Tools...)
I have no idea what you mean by "peak cloud".
I do know for a fact that sizeable tech companies are already moving their services away from cloud providers because it makes absolutely no sense to pay small fortunes per month, sometimes costing more than the hardware it takes to run it, just to run stuff on a cabinet you are quite able to afford.
I also know that the bulk of these cloud services cater to cargo-cult customers plagued with FOMO.
Most organizations from startups to enterprises generally benefit from the cost, scalability, security, technical support, and other benefits of cloud storage and computing.
Startups, for example, can focus on development and distribution when using cloud based tech for at least some part of their stack; and needless to say enterprises can scale quickly and securely.
Within this spectrum, where and what is this “sizeable” company that benefits from doing all this themselves?
Of course sizeable companies of all sectors are using hybrid. Always been the case. Both things can be and are true: owned is growing. Cloud is growing
How is it no market to saturate when the CEO of Amazon (and former CEO of AWS) said that only 5% of all IT spend is on any cloud provider? https://accelerationeconomy.com/cloud/amazon-shocker-ceo-jas...
yet
s&p pe ratio was as bad as both 2000 an 2008:
https://www.longtermtrends.net/sp500-price-earnings-shiller-...
buffett indicator was as bad as it has ever been by a wide margin:
https://www.longtermtrends.net/market-cap-to-gdp-the-buffett...
home price to income is worse than in 2008:
https://www.longtermtrends.net/home-price-median-annual-inco...
the last time real interest rates were this low the fed had to wreck the economy to get inflation under control:
https://www.longtermtrends.net/real-interest-rate/
so, yeah, not as bad as the precursor to 2000 or 2008, but rather as bad as both combined, but with interest rates already rock bottom...