It's difficult to describe the bewildering amount of change in the technology space over the last decade in a single diagram ... however, undeterred I've tried to catch the fundamentals ... 
Same patterns are occurring. In 2008, the future giants of media were Netflix (focused on the user, cloud and reducing assets) not the Blockbusters (focused on cashflow, building stores and data centres). Ditto today. The future giants of robotics are companies like iRobot.
It's always a hoot to go through those past Blockbuster presentations from 2007 to 2008 ... the self belief, the investment in assets, the hubris, the mindless strategy that was bereft of any understanding of basic economic patterns.
I see the same occurring today with serverless ... oh well. Some will spend huge fortunes on assets that don't meet any specific user need and they will fail to deliver on whatever paltry benefits they claim it will create.
Well, my advice to CEOs in 2008 was to ask your CIO about Cloud. If they dismiss it then fire them.
M advice to CEOs in 2018 is to ask your CIO about Serverless. If they dismiss it then fire them.
You'll thank me later.
M advice to CEOs in 2018 is to ask your CIO about Serverless. If they dismiss it then fire them.
You'll thank me later.
"we're going to build our own" ...
#facepalm.
If you must, as a rule of thumb, for an estate providing functions to others then ... one ops person + $300k p.a. for every 10M active users. This is for everything, electricity, buildings insurance, software, tin, monitoring etc.
#facepalm.
If you must, as a rule of thumb, for an estate providing functions to others then ... one ops person + $300k p.a. for every 10M active users. This is for everything, electricity, buildings insurance, software, tin, monitoring etc.
"But, we spend more than that on consultants to design it!"
Then don't. Have a policy of public serverless first, then public cloud (containers etc) if you absolutely can't in a specific case. After five years, you'll have enough experience to laugh at your home grown plans.
Then don't. Have a policy of public serverless first, then public cloud (containers etc) if you absolutely can't in a specific case. After five years, you'll have enough experience to laugh at your home grown plans.
But gain that experience. Take any home grown serverless efforts and shelve them until you have a high % of your engineers with five or more years experience of public serverless. Only then will you be in a position to properly judge a home grown effort.
This is why I'm not going to serverless events focused on building your own ... it's a waste of my time and everyone else's money. I find the very idea to be utterly foolish.
