o.k. shortish non-technical tweet thread this morning on capital Vs operational financial expenditure, because I only just learned a new way to think about it yesterday and it blew my mind. Maybe it's obvious to more, but it wasn't to me!
-
Show this thread
-
Since we first launched, one of AWS's biggest benefits has been that customers can trade many big capital expenses; building data centers, buying software licenses, and so on for operational expense. In general, AWS is pay for what you use, with monthly bill.
2 replies 0 retweets 5 likesShow this thread -
For a long time I've thought this is pretty simple. It's painful and wasteful to keep big chunks of money around and makes CFOs unhappy. Capital costs can be financed too, like buying a car, but there's an interest rate to pay and so the net-present-value cost is higher. Simple!
1 reply 0 retweets 1 likeShow this thread -
Now I knew that that model is too simple; there's also waste in capital expenses because depreciation models and amortization aren't perfect. Basically you might buy something, and it can become obsolescent and you never get all the years out of it you wanted, and can't sell it.
1 reply 0 retweets 2 likesShow this thread -
But what I didn't really "get" until yesterday is that many CFOs prefer operational expenses because they are structurally easier to lower than capital expenses. This was an "aha!" for me. Here's how it clicked ...
1 reply 1 retweet 5 likesShow this thread -
In general, when you have a big capital expense, like building a DC, it's a process that's bounded in time. A company might put a contract out for bid with suppliers, and they'll make their offers, and you pick one and you get on doing it.
1 reply 0 retweets 2 likesShow this thread -
but often things go over budget for unforeseen reasons, you can't just pause and litigate that too much, because there's time pressure too, and then the thing is delivered and you move on, and do another exercise a few years later.
1 reply 0 retweets 2 likesShow this thread -
Basically you get a limited number of high-stakes points at which you can try to save costs. If you did it often you might get good at it, but then it would be classed as an operational expense anyway, not a capital one.
1 reply 0 retweets 3 likesShow this thread -
Whereas with operational expenses: the organization can optimize costs on their own time, as and when the people or resources to do it are available. Wait for the quiet quarter, or whatever, do a "find some cost-savings" exercise, and make it stick as much as you can.
1 reply 0 retweets 6 likesShow this thread -
This is *structurally* much better way of doing business. Simple example: You can train people how to look for cost savings on an ongoing basis; not just hope you have the person who was around the last time we bid for a big contract, etc ...
1 reply 1 retweet 6 likesShow this thread
I got all of this from a conversation with a CFO. Like I said, probably simple and obvious to many, but it wasn't to me. I never "got" that level to the whole thing, I thought it was just about budgetary math. But it makes total sense. End-of-thread!
-
-
Replying to @colmmacc
Tangentially related, when we were first building our (fully serverless) IoT application on AWS, the discussions with purchasing were fascinating. Our account team had helped us estimate cost per robot based on our estimates of message volume, etc
1 reply 0 retweets 3 likes -
Replying to @ben11kehoe @colmmacc
And so with purchasing, they want to know how much the cloud is going to cost per robot per year, to include that in the build cost per robot. But they were used to manufacturing costs. X widgets per robot at
$Y per widget buying Z million at a time, contracted for W years...1 reply 0 retweets 2 likes - 3 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.