Written by Alex Iskold and edited by Richard MacManus.
We have written before about the innovative Amazon Web Services Platform. This stack was officially announced by Amazon CEO Jeff Bezos during the recent Web 2.0 summit and is now considered part of the core business strategy for Amazon. While analysts, competitors and Wall Street are pondering what to make of this move from a business sense, in this post we look at who is utilizing Amazon Web Services - and how. This post is based on personal communication with those people, along with the set of success stories available on the Amazon Web Services site.
The fact is many small, medium and even large businesses (even Microsoft), rushed to put Amazon Web Services to use. Why did they do it? Because Amazon offers a decade of experience in running one of the largest internet enterprises - and has wrapped this expertise into a set of pre-packaged services and APIs.
To remind you, here again is the Amazon Web Services Stack:

The Amazon Web Services stack is impressive in its scale and also well thought through. Amazon is methodical about this strategy and is aiming to create an offering which can truly be called an Operating System for the new Web. Many companies have already recognized the power and ROI of the Amazon platform and are literally betting their business on it.

Webmail.us is probably the most compelling success story for Amazon Web Services because of its huge ROI. It is an established business with over 27,000 customers. It had a real and simple business need - improve the cost and reliability of its backup system. After considering many alternative solutions, the company decided to utilize Amazon S3, The Simple Queue Service and the Elastic Compute cloud to address all of its needs.
The company claims to have improved its backup process and cut costs by 75%. The Webmail.us success story on Amazon contains a paragraph that nicely summarizes the technical and business gains:
"Amazon's Web Scale Computing model shifts the focus from do-it-yourself to let-the-experts-do-it. It allows businesses to scale up or down based on requirements and demand, and provides pay-as-you-go billing models. This combination allows businesses to turn fixed costs into variable costs, while knowing that their data or services will always be available."

SmugMug is another interesting success story. It is a straightforward one, because it uses Amazon S3 exactly how it was intended to be used - for storing large media. Today SmugMug hosts over half a billion photos on Amazon. And here is the real "wow factor" in this story: one week after writing the first line of code, SmugMug was storing all of its new images in Amazon S3.
The Amazon S3 API consists of just a few simple method calls. It is equally easy to implement in Java, PHP, JavaScript, Perl, C#, Python and many other languages. As the SmugMug success story illustrates, a simple API means a very quick and painless adoption.
To pepper this with more numbers: SmugMug is now backing up all of its new images to S3, which amounts to 10 terabytes of data monthly. The SmugMug site has not gone down since adopting Amazon S3 and it estimates it will save half a million dollars on its disk storage annually. So, as the company points out, S3 makes it possible for SmugMug to compete head to head with bigger companies that have deep pockets, without having to raise massive amounts of cash for hardware. So this is game changing.
As soon as S3 came out, many companies recognized an opportunity to deliver business and personal backup solutions. The model is simple - charge a small premium on the top of the Amazon S3 storage costs.
With that approach it is essentially a user acquisition battle, where the implementation and marketing become paramount. Altexa targets small businesses, while ElephantDrive and JungleDisk target consumers - but all of them share the benefit and ease of use of S3. In their success stories, the companies emphasized incredibly quick (literally a few days) adoption, cost savings and reliability.

The success stories that we have covered so far are mostly using Amazon's S3 storage service. Scanbuy however is utilizing the Amazon eCommerce API to bring unique comparison shopping solutions to mobile phones. Their claim to fame is allowing the users to lookup prices by simply scanning the barcodes of items in a store. This is a clever approach that is made possible by a combination of technologies.
One of the key technologies here is the Amazon eCommerce API, which offers unlimited and complete access to most Amazon items. Scanbuy uses the API to fetch the latest pricing information, letting the user decide if they are really getting a good deal in the store. And as the company explains, they simply could not have done what they are doing now without the Amazon eCommerce API.
So why are analysts not sure what to make of the Amazon Web Services? In an article in BusinessWeek entitled Jeff Bezos' Risky Bet, their main concern seemed to be: will businesses use this? Well in this post we've shown that for some businesses, the AWS stack provides a set of very compelling value propositions - both technical and business. And having real business success stories with ROI and cost savings in the 50-75% range, makes it basically a no brainer.
We think that the real question is: does this work for Amazon? Is it ready to be a software and Web Services company? Will Amazon be able to scale this business indefinitely... and most importantly: are the margins high enough for it to be worthwhile? We have to believe that Jeff Bezos and the Amazon team did the math - and that the answer is absolutely yes.
TrackBack URL for this entry: http://www.readwriteweb.com/cgi-bin/mt/mt-tb.cgi/2882
Comments
Subscribe to comments for this post OR Subscribe to comments for all ReadWriteWeb posts
I did the math, too. I was thinking of using their EC2 service and maybe S3 for a project I'm working on, but it turned out to not be worth the cost. In the short-term it seemed like a good deal (low startup costs) but over the long term, it was not.
Some reasons that others might think the same:
1. Stored data has to be regenerated and refreshed so often when a user views or downloads the data in a particular format, that it's not worth storing that generated format anywhere.
2. Running EC2 for a particular process that only runs whenever something particular happens (and that happenstance is more rare than site access), I really started to see little difference in the longterm on EC2.
I think that the only reason these computations worked out the way they did for me (I'm going with dedicated hosting) is that my project has more eccentricities than most hosts are willing to deal with cheaply, but has less traffic than makes AWS cost effective.
I love the idea. I'd like to work on something, sometime, that could do use AWS well.
I think this is an amazing service for startups in the financial domain or gaming which cannot afford to spend the time or money for setting up the infrastructure.
I find the idea very good and want to work on some project which can use and leverage the APIs and infrastructure provided by Amazon.
As they say today, in the web 2.0 world, startup is easy and affordable. Amazon Web services works in the same direction.
I find that http://www.BackupReview.info to be the best website for online backup news, articles and information. This site lists more than 400 online backup companies and ranks the top 25 on a monthly basis.
Well, there's also S3 Backup http://s3bk.com/ that I develop. It already has a number of features that all other S3 based solutions lack and the app keeps moving forward. Hopefully it will become the preferred way to backup to S3.
I was really excited when I heard about it but when I looked deeper I found its not for me and I suspect a lot of others too. If you can afford the cost then its great but otherwise its not.
s3 is useless for high volume web apps unless you use s2. why? the bandwidth seems cheap but to store all these massive files they have to go via your own website or via an s2 hosting website. The winner is s2 here because there is no bandwidth cost between s2 and s3. Now if you are using shared hosting or a VDS then you have to take the bandwidth cost twice, once to upload to you and once to upload to s3. Also how many shared hosting companies are gonna be happy with you using all your available bandwidth. Once the files uploaded get into the 100' of megs area then your host really good bandwidth allowance looks quite weak.
I like the idea but they need another service to get around the problem of getting the files up there.
also the cost of s2 looks good until you realise they don't promise any uptime. so you then need to look at clustering instances and this gets real expensive and its not for the average developer.
I recently found a very interesting website:
http://alreadylinked.com/
There you can purchase ad space for your Blog etc.
This article read like an infomercial to me :(
IMHO These web services will succeed. Amazon are externalising what they already have. So, the risk is limited and there's demand there. I use both EC2 and S3.