< Language in the post quick edited by NSA for language, clarity – so, stop bitching 🙂 >
Guest Blog Post from Vlad, Distinguished Engineer, c-cloud.
As some of you know, I am a huge fan of AWS as they had pushed the envelop in cloud computing and made it super easy and quick for developers to obtain compute and storage resources. What started as dev/test is now a serious contender for production workloads. Companies like NetFlix have already standardized on AWS. If you are currently not using AWS in some shape or form, you will be seen as a technology laggard.
AWS has also advanced the pay per view business model and has allowed IT to balance their CAPEX and OPEX expenses. They made it so that you require less commitment before using a resource. ( If you don’t like it, you can stop using it next hour ).
AWS has innovated faster than many other “me-too” players in the space. Even though no serious competitors exist yet, AWS has not let its sight on customer needs and has been solving their problems.
So, in essence AWS is awesome and you should jump all the way in, right? Wrong.
Let me explain:
1. You are not Netflx – Netflix has negotiated special contracts with AWS – they get architectural reviews, early access and dedicated support from AWS team.
AWS is hiring enterprise sales teams, and may negotiate with you if you commit to large enough deal size. But, likely you don’t qualify.
2. AWS is part of Amazon – Amazon is a consumer company – In consumer companies, the pricing models are optimized to squeeze every last penny out of a low margin business. Consumer companies experiment with pricing changes to see what yields best results. Consumer pricing models do not optimize for consumer – they optimize for profits. Often, they make changes to extract more revenue and hide it so consumers don’t notice. This is not their attempt to cheat consumer, its what the model had recommended.
Amazon has mastered this art and there is now evidence that they are applying it widely across their properties. I will give you an example:
Woot is a daily deals company ( just one type of item sold per day originally) that became quite popular ( it still is ). Amazon acquired Woot and had generally improved it. One unique thing about Woot is their shipping – its flat $5 whether you buy a 50c toothbrush or a $1500 electronic item. One of the improvements Woot made after Amazon acquired them was to offer flat $5 shipping for all items purchased within a 24 hour period. This allowed people to buy more stuff ( especially during Woot-off periods when Woot runs through several items in a day). Enough time to influence consumer behavior to buy more stuff through out the day. All good, right? Well, till Amazon pricing model gods intervened. They probably observed now that we have trained our consumer behavior in a certain way, we can go back to charging $5 shipping per each order and they wouldn’t notice, and it would increase our revenues. So, Woot made this change and communicated to customers hidden in a blog post. Not even regular customers ‘heard’ of this change.
This is a totally acceptable behavior for a consumer company – we are not talking large amounts of money and their change on surface is not that big.
However, this is a big no-no in enterprises where there has to exist a level of trust between IT and the vendor they do business with. IT often is willing to pay a premium for a trusted partnership. They value this partnership more than everything else ( yes, we all joke about Golf, expensive wine and dinners ). Amazon consumer company behavior does not fit this.
This is not to say enterprise software companies dont make pricing changes – they do. For example, VMware went through a huge fiasco couple of years ago with their pricing changes. However, the difference is that the pricing change is very public and it was widely discussed. A letter highlighting the change was sent out – account teams talked to customers. When customer said its a moronic change, they rolled back some of the changes. There is feedback loop here – there is no feedback loop with Amazon.
So, in summary – Amazon behavior as a consumer company is a huge drawback for them to sell to larger set of enterprises and will cause them to face friction.
3. In bed with US govt.
Amazon is a company that is first to turn over records and do exactly what US govt asks them to do. ( Remember Wikileaks?) . It makes good business sense for them, as they are a consumer company and being on good side of govt. is generally good for lobbying and other activities.
Enterprise customers are often international organizations, and they have customers outside of US and need to ensure that these customers are not being spied on by US govt. Running these customer workloads in AWS Ireland is not sufficient as Amazon would turnover the data if requested by US govt in a heartbeat. This requires customers to consider EMEA and APAC infrastructures that are not owned by an US based entity.
These are the 3 reasons I believe AWS will not be as widely adopted by enterprises as experts would like you to believe. But, don’t take this as being bearish on AWS – there is still plenty of SMB and departmental market to be had for AWS and it will continue to grow there. And yes, we need to continue cheering on AWS for the Innovation. However, it would take some other vendor to step forward and offer a credible public cloud option for enterprises. A consumer company isn’t it.
All of this could change, if Amazon spins off AWS into a separate company that is not subject to Amazon pricing gods.