There have been a lot of rumblings over the past year or two suggesting that Amazon needs to spin off Amazon Web Services (AWS). These rumblings have waxed and waned as various pundits attempt to prognosticate on what Amazon is going to do next.
I find it increasingly difficult to parse the suggestions given in articles when trying to match it to the reality of the world at this time.
I completely understand that spin offs are designed to unlock the value in a company, I don’t think it’s nearly that simple in this case. There is far more to it than just making the numbers work as part of a business case. If you read the backstory as chronicled on TechCrunch, the idea behind Amazon launched from previous integration nightmares the company had experienced in the past. Amazingly, this was 15 years ago already, long before most developers were even thinking of integration at a scale like this.
The feasibility of a spin off is not the question here. Of course AWS could be spun off into a separate corporate entity, and it absolutely would do quite well. The reason it seems unlikely to me that they would do so is that Amazon would lose a lot of the flexibility that they currently enjoy from operating the AWS platform.
The principals at AWS most assuredly look at the needs of all of their customers; just a casual glance at the list of product announcements from AWS re:Invent 2018 should provide ample evidence of their intent. Ensuring the ease of integration as well as a convenient ability to quickly harness a large number of tools continues to help businesses of all sizes.
But, I’d be shocked to learn that Amazon proper doesn’t have the ability to put their thumb on the scale to push feature development as part of AWS development timelines. That alone indicates that the value proposition of retaining control of AWS hasn’t been fully considered, suggesting that the cost of an AWS spin off is higher than previously calculated. Considering how Jeff Bezos approaches, well, everything, it seems to be a stretch that relinquishing control of a well-run division of Amazon, when the company itself depends so much on it, would be something considered unless an unexpected hardship were to occur.
The conclusion that Amazon and AWS aren’t co-dependent seems quite short-sighted when considering the technical aspects. Sometimes the math is only part of the equation, and further investigation is required.
The idea of avoiding government regulation is an interesting one, but I doubt it’s a concern the company will need to face in the near future. It seems much more plausible that an entity like Facebook would need to worry about this. The Department of Justice took on Microsoft with little to show for it; for all of the bluster of the day, Amazon seems well positioned to avoid the scrutiny of U.S. regulators.
Of increasing concern could be the European governments with the implementation of GDPR, but AWS is well ahead of this. It’s always possible that Amazon could run afoul of the GDPR privacy rules, but a company with resources like Amazon should have that well in hand. Furthermore, while I haven’t read GDPR in its entirety, it seems more likely that Amazon would be charged with hefty fines than find itself burdened by regulation it can’t keep up with.