5 ms·

Don’t look now, but AWS might be a billion-dollar biz

103 points by bond 15y ago

treeform 15y ago
Not surprising, they brought cloud computing to the masses and are doing a pretty good job at it too.
- arctangent 15y ago
  Agreed. It would be incredibly naive to suggest that AWS is anything other than a huge source of income for Amazon for a long time to come.
jontas 15y ago
I see their stock crash today as a pretty temporary thing. I bought a lot of shares today around 200 which I hope is close to the bottom of this decline. AWS growth combined with their new hardware combined with the holiday season should give them a nice boost by early next year.
- wanorris 15y ago
  I think the main thing will be how their margins play out. Amazon is a big enough company that investors expect decent margins even while they're growing rapidly. Perhaps this is an unreasonable expectation, but if so, I think Amazon should be more aggressive about giving guidance. Are there basically just no margins in their retail business? Are they intentionally depressing profits to keep taxes down? Why isn't there plenty of money there? Because even if they're investing a ton of money in expanding their distribution system to allow for growth, normally that gets treated as a capital cost and amortized over a period of time, not just expensed in the period they pay the money out. The same is true for upfront investment in their Kindle Fire business (edit: or in AWS infrastructure). I'm not trying to attack Amazon -- they're a savvy company that has proven to be hugely successful through any number of trials and crises. But it's just not clear to an outsider what's going on with their business at the moment.
  - jontas 15y ago
    I agree with you, I am just betting (hoping?) that the current crash is temporary and they will recover in the near future. Their revenues are also down significantly (I think it was something around 40%) which can't be explained by large investments effecting their earnings.
    - vannevar 15y ago
      Their revenues are also down significantly... That is incorrect; revenues were up 44% (http://phx.corporate-ir.net/phoenix.zhtml?c=176060&p=irol-newsArticle&ID=1621411&highlight= http://phx.corporate-ir.net/phoenix.zhtml?c=176060&p=iro...)
      - jontas 15y ago
        Well I fail at reading comprehension.. thanks.
  - hyperbovine 15y ago
    Personally I've never been able to figure out how they manage to make money overnighting me dental floss for $4. (That said, I still own their stock and was glad to scoop up some more of it this morning on the cheap.)
    - jaredsohn 15y ago
      I wondered the same when I recently ordered a mousepad (via one of their partners) for ~$3 with free shipping, but since they used US mail I can understand how they managed it; they can perhaps do the same for small quantities of dental floss.
      - hyperbovine 15y ago
        I mean, somebody got paid to load my 90 yds of Reach onto an airplane in the middle of the night and fly it from Tennessee to California. With virtually no added cost to me, since I have Prime which I share with 5 other people in my family. New economy or not, there's just no way that makes economic sense. I know the argument is that Prime alters one's spending patterns, and ten paying customers are subsidizing every yokel like me who's too busy/lazy to walk to the drug store. Still it feels like an unsustainable model. I often wonder if we'll look back on this period as the apotheosis of consumerism, when cheap and abundant fossil fuels made it possible behave like we all owned matter transporters.
- pdx 15y ago
  I like Amazon, but with a PE of 90, I can't make myself buy them. At some point in the future, when all their growth is behind them and they become a value investment as opposed to a speculative investment, they need to make me between 5% and 10% on my money if I buy the whole company. That's how I evaluate stocks. That means a PE of 10 to 20. In what year do we believe that Amazon will be making 5X to 9X what it's making now, given it's current maturity? If that year is near, than, sure, perhaps buy here. I just can't imagine that kind of growth for such a large company.
  - vannevar 15y ago
    Current revenue growth is around 50% annually; if that continues, it will reach the lower edge of your target (5X) in 48 months. Assuming of course it can maintain its margins.
    - mikeryan 15y ago
      Am I missing something or don't your assumptions assume that over those same 48 months Amazon's price remains static? Note that any assumption of Amazon's margins remain the same is kind of hard pressed. Their revenues this quarter was 40% greater then the same quarter last year - yet their earnings were down 73%.
      - vannevar 15y ago
        The question was 'when will Amazon be making 5X what they make now?', and the answer (at current growth rates and current margins) is 48 months. Which is not a long time, thereby explaining the high P/E. If you think the growth will slow substantially or the margins will shrink substantially (or both) in the next 48 months, then the PE is too high. Otherwise, perhaps it's not so crazy after all.
  - kkowalczyk 15y ago
    Personally I think that Amazon has essentially unlimited growth opportunity, much higher than even the most amazing, most profitable product companies like Apple or Google. Product companies have a Microsoft problem: after they dominate a given market (and Microsoft pretty much dominated a lot of the most valuable segments of software market) and that market matures into slow growth, they have a very hard time finding new opportunities and consequently only grow with the size of the market, and as we've seen with Microsoft's flat stock price, Wall Street doesn't like that. To me, Amazon's is Merchant Of Everything, both physical and digital and Everything is a very big market. Amazon grew to its current size by exploiting shift to on-line purchasing but today they're big enough to start taking advantage of their size. Walmart conquered U.S. with lower prices but Amazon can offer even better prices because they don't have to operate physical stores with staff, just warehouses. Amazon is not done until they have a warehouse for each Walmart store, Best Buy store etc. and their competitive advantage grows with every warehouse they build. And when there are more Amazon's warehouse than Starbucks, there's international market to conquer. Additionally Bezos has shown that he's very comfortable competing aggressively, running the business at the edge of profitability, which makes sense if you think that Amazon's ultimate ambition is to drive every other merchant out of business by offering lower prices and become the Merchant Of Everything.
kjw 15y ago
Can anyone think of any ways to collect data on just how big AWS really is? So far it seems like everyone is just trying to infer based on the "other" line of their revenues and anecdotes from the cloud community.
- jontas 15y ago
  Their servers send an indication in the http headers: Server: Apache/2.2.21 (Amazon) I guess you'd need to crawl the web and look for those headers. Or maybe you could look at IP addresses? It would certainly be difficult to do with any kind of accuracy but you could probably get some decent estimates if your sample size was large enough.
  - 0x12 15y ago
    Very large numbers of Amazon servers are used for something else than cranking out HTTP pages. Expect a lot of them to be crunching numbers for bio-informatics problems, physics simulations and so on. That's why there is a CUDA enabled instance. Rendering web pages is actually one of the worst use cases for Amazon from a bang-for-the-buck perspective, especially when you factor in bandwidth.
    - samstave 15y ago
      Crawl and multiply by 3 (assume that 30% are direct web servers)
      - nl 15y ago
        Is that 30% based on anything at all? When people are building 30,000 core compute clusters [1] on EC2 - presumably with zero publicly available web servers, I'd be very interested in any methodology that provides reasonable estimates of revenue based on public web servers. http://arstechnica.com/business/news/2011/09/30000-core-cluster-built-on-amazon-ec2-cloud.ars http://arstechnica.com/business/news/2011/09/30000-core-clus...
  - kjw 15y ago
InclinedPlane 15y ago
Regardless of what you think about the 37 Signals folks one bit of their advice has always struck me as being particularly insightful and powerful: Sell your byproducts. There are two very strong reasons to do this. First, it can be an excellent business. You've created something to solve some problem or remove some pain that your company is feeling, it's extremely unlikely that your company is so unique that it's pain isn't shared by other companies. And it's unlikely your internal tools would be of no interest or utility to other companies. Second, internal tools are typically of terrible quality. There are various reasons for this but it's a very common pattern due to fundamental pressures and incentives. By selling internal tools you force them to have owners and you force them to have a quality sufficient to be acceptable to the market. This generally vastly increases their quality, which provides a benefit to everyone who uses them, including you.
- kkowalczyk 15y ago
  Except it has been reported by Amazon insiders that AWS is not a byproduct of Amazon's other work but a very much stand alone product with it's own dev group with little connection to devs working on Amazon proper and that Amazon was relatively slow in adopting AWS product for implementing Amazon's internal software. Amazon always tried to give the impression that they use AWS but (at least in early years) those were half-truths (i.e. statements vague enough that if you squint at it one way you can read "Amazon's internal systems are build in AWS services" and if you squint at it differently, they might just as well say "AWS's servers are hosted in the same data center as Amazon's servers"; public perception was the former and the reality was closer to the latter). That, of course, doesn't mean that 37 Signals' advice isn't good, just that this is not an example that confirms that advice.
  - pg 15y ago
    Yes, this is true. When I first heard about AWS I assumed Amazon was selling excess capacity (because why would an online retailer set out to build something like this?), but in fact, as you say, it was a completely separate project. I've heard from people who used to work at Amazon that they don't use AWS for much there.
salem 15y ago
There was an interesting comment from the stack overflow guys that they guessed that using AWS could cost them 4x more. The evernote guys are another high profile team that went their own way. I wonder what that says about the margins for AWS, or is the AWS architecture not actually a more efficient use of hardware, power etc compared to old school setups at co-lo's.
- kkowalczyk 15y ago
  There are several reasons why AWS does (and should) cost more than a generic solution. 1. If you just look at the cost of bandwidth and hardware, there's no doubt you can find cheaper solutions in the commoditized basic hosting market. AWS solutions, however, remove a lot of manual work that goes into setup, configuration and maintenance of servers. If you're a startup, it costs you (a lot of) money to hire those people. It also cost you time wasted not working on the core of your service. Up to a certain size, the fixed costs of additional personnel heavily outweigh higher bandwidth costs. Evernote and Stackoverflow have extremely high traffic and might have crossed the threshold at which it's profitable for them to look for savings in hosting bill. A vast majority of startups is not in that position. 2. Unique services demand a premium. Most of the hosting companies provide commoditized services. AWS is much more unique. 3. Services higher up the stack demand a premium. The most basic hosting service offers you a bit of physical space in their building and only give you power for the server and network connectivity. Amazon provides a wide variety of services on top of that. Developing those services and keeping them working is much more costly than just ensuring power is on and network is up. They have to charge more (they have higher costs) and they can charge more (they provide a richer service). 4. Ultimately, the prices are set by the market. We can't really say anything about whether Amazon is more efficient at using hardware or power than, say, SoftLayer, based on their prices. First, their services cost much more to provide. Second, they are free to set their margins at whatever level they please. They could run the service at a loss or they could charge an outrageous premium. We don't know. People are clearly willing to pay for AWS services so at the very least they've set the prices in a way that, given additional benefits, is competitive with traditional server hosting.
  - salem 15y ago

103 points by bond 15y ago

treeform 15y ago

Not surprising, they brought cloud computing to the masses and are doing a pretty good job at it too.

arctangent 15y ago
Agreed. It would be incredibly naive to suggest that AWS is anything other than a huge source of income for Amazon for a long time to come.

jontas 15y ago

I see their stock crash today as a pretty temporary thing. I bought a lot of shares today around 200 which I hope is close to the bottom of this decline. AWS growth combined with their new hardware combined with the holiday season should give them a nice boost by early next year.

wanorris 15y ago
I think the main thing will be how their margins play out. Amazon is a big enough company that investors expect decent margins even while they're growing rapidly. Perhaps this is an unreasonable expectation, but if so, I think Amazon should be more aggressive about giving guidance. Are there basically just no margins in their retail business? Are they intentionally depressing profits to keep taxes down? Why isn't there plenty of money there? Because even if they're investing a ton of money in expanding their distribution system to allow for growth, normally that gets treated as a capital cost and amortized over a period of time, not just expensed in the period they pay the money out. The same is true for upfront investment in their Kindle Fire business (edit: or in AWS infrastructure). I'm not trying to attack Amazon -- they're a savvy company that has proven to be hugely successful through any number of trials and crises. But it's just not clear to an outsider what's going on with their business at the moment.
- jontas 15y ago
  I agree with you, I am just betting (hoping?) that the current crash is temporary and they will recover in the near future. Their revenues are also down significantly (I think it was something around 40%) which can't be explained by large investments effecting their earnings.
  - vannevar 15y ago
    Their revenues are also down significantly... That is incorrect; revenues were up 44% (http://phx.corporate-ir.net/phoenix.zhtml?c=176060&p=irol-newsArticle&ID=1621411&highlight= http://phx.corporate-ir.net/phoenix.zhtml?c=176060&p=iro...)
    - jontas 15y ago
      Well I fail at reading comprehension.. thanks.
- hyperbovine 15y ago
  Personally I've never been able to figure out how they manage to make money overnighting me dental floss for $4. (That said, I still own their stock and was glad to scoop up some more of it this morning on the cheap.)
  - jaredsohn 15y ago
    I wondered the same when I recently ordered a mousepad (via one of their partners) for ~$3 with free shipping, but since they used US mail I can understand how they managed it; they can perhaps do the same for small quantities of dental floss.
    - hyperbovine 15y ago
      I mean, somebody got paid to load my 90 yds of Reach onto an airplane in the middle of the night and fly it from Tennessee to California. With virtually no added cost to me, since I have Prime which I share with 5 other people in my family. New economy or not, there's just no way that makes economic sense. I know the argument is that Prime alters one's spending patterns, and ten paying customers are subsidizing every yokel like me who's too busy/lazy to walk to the drug store. Still it feels like an unsustainable model. I often wonder if we'll look back on this period as the apotheosis of consumerism, when cheap and abundant fossil fuels made it possible behave like we all owned matter transporters.
pdx 15y ago
I like Amazon, but with a PE of 90, I can't make myself buy them. At some point in the future, when all their growth is behind them and they become a value investment as opposed to a speculative investment, they need to make me between 5% and 10% on my money if I buy the whole company. That's how I evaluate stocks. That means a PE of 10 to 20. In what year do we believe that Amazon will be making 5X to 9X what it's making now, given it's current maturity? If that year is near, than, sure, perhaps buy here. I just can't imagine that kind of growth for such a large company.
- vannevar 15y ago
  Current revenue growth is around 50% annually; if that continues, it will reach the lower edge of your target (5X) in 48 months. Assuming of course it can maintain its margins.
  - mikeryan 15y ago
    Am I missing something or don't your assumptions assume that over those same 48 months Amazon's price remains static? Note that any assumption of Amazon's margins remain the same is kind of hard pressed. Their revenues this quarter was 40% greater then the same quarter last year - yet their earnings were down 73%.
    - vannevar 15y ago
      The question was 'when will Amazon be making 5X what they make now?', and the answer (at current growth rates and current margins) is 48 months. Which is not a long time, thereby explaining the high P/E. If you think the growth will slow substantially or the margins will shrink substantially (or both) in the next 48 months, then the PE is too high. Otherwise, perhaps it's not so crazy after all.
- kkowalczyk 15y ago
  Personally I think that Amazon has essentially unlimited growth opportunity, much higher than even the most amazing, most profitable product companies like Apple or Google. Product companies have a Microsoft problem: after they dominate a given market (and Microsoft pretty much dominated a lot of the most valuable segments of software market) and that market matures into slow growth, they have a very hard time finding new opportunities and consequently only grow with the size of the market, and as we've seen with Microsoft's flat stock price, Wall Street doesn't like that. To me, Amazon's is Merchant Of Everything, both physical and digital and Everything is a very big market. Amazon grew to its current size by exploiting shift to on-line purchasing but today they're big enough to start taking advantage of their size. Walmart conquered U.S. with lower prices but Amazon can offer even better prices because they don't have to operate physical stores with staff, just warehouses. Amazon is not done until they have a warehouse for each Walmart store, Best Buy store etc. and their competitive advantage grows with every warehouse they build. And when there are more Amazon's warehouse than Starbucks, there's international market to conquer. Additionally Bezos has shown that he's very comfortable competing aggressively, running the business at the edge of profitability, which makes sense if you think that Amazon's ultimate ambition is to drive every other merchant out of business by offering lower prices and become the Merchant Of Everything.

kjw 15y ago

Can anyone think of any ways to collect data on just how big AWS really is? So far it seems like everyone is just trying to infer based on the "other" line of their revenues and anecdotes from the cloud community.

jontas 15y ago
Their servers send an indication in the http headers: Server: Apache/2.2.21 (Amazon) I guess you'd need to crawl the web and look for those headers. Or maybe you could look at IP addresses? It would certainly be difficult to do with any kind of accuracy but you could probably get some decent estimates if your sample size was large enough.
- 0x12 15y ago
  Very large numbers of Amazon servers are used for something else than cranking out HTTP pages. Expect a lot of them to be crunching numbers for bio-informatics problems, physics simulations and so on. That's why there is a CUDA enabled instance. Rendering web pages is actually one of the worst use cases for Amazon from a bang-for-the-buck perspective, especially when you factor in bandwidth.
  - samstave 15y ago
    Crawl and multiply by 3 (assume that 30% are direct web servers)
    - nl 15y ago
      Is that 30% based on anything at all? When people are building 30,000 core compute clusters [1] on EC2 - presumably with zero publicly available web servers, I'd be very interested in any methodology that provides reasonable estimates of revenue based on public web servers. http://arstechnica.com/business/news/2011/09/30000-core-cluster-built-on-amazon-ec2-cloud.ars http://arstechnica.com/business/news/2011/09/30000-core-clus...
- kjw 15y ago

InclinedPlane 15y ago

Regardless of what you think about the 37 Signals folks one bit of their advice has always struck me as being particularly insightful and powerful: Sell your byproducts. There are two very strong reasons to do this. First, it can be an excellent business. You've created something to solve some problem or remove some pain that your company is feeling, it's extremely unlikely that your company is so unique that it's pain isn't shared by other companies. And it's unlikely your internal tools would be of no interest or utility to other companies. Second, internal tools are typically of terrible quality. There are various reasons for this but it's a very common pattern due to fundamental pressures and incentives. By selling internal tools you force them to have owners and you force them to have a quality sufficient to be acceptable to the market. This generally vastly increases their quality, which provides a benefit to everyone who uses them, including you.

kkowalczyk 15y ago
Except it has been reported by Amazon insiders that AWS is not a byproduct of Amazon's other work but a very much stand alone product with it's own dev group with little connection to devs working on Amazon proper and that Amazon was relatively slow in adopting AWS product for implementing Amazon's internal software. Amazon always tried to give the impression that they use AWS but (at least in early years) those were half-truths (i.e. statements vague enough that if you squint at it one way you can read "Amazon's internal systems are build in AWS services" and if you squint at it differently, they might just as well say "AWS's servers are hosted in the same data center as Amazon's servers"; public perception was the former and the reality was closer to the latter). That, of course, doesn't mean that 37 Signals' advice isn't good, just that this is not an example that confirms that advice.
- pg 15y ago
  Yes, this is true. When I first heard about AWS I assumed Amazon was selling excess capacity (because why would an online retailer set out to build something like this?), but in fact, as you say, it was a completely separate project. I've heard from people who used to work at Amazon that they don't use AWS for much there.

salem 15y ago

There was an interesting comment from the stack overflow guys that they guessed that using AWS could cost them 4x more. The evernote guys are another high profile team that went their own way. I wonder what that says about the margins for AWS, or is the AWS architecture not actually a more efficient use of hardware, power etc compared to old school setups at co-lo's.

kkowalczyk 15y ago
There are several reasons why AWS does (and should) cost more than a generic solution. 1. If you just look at the cost of bandwidth and hardware, there's no doubt you can find cheaper solutions in the commoditized basic hosting market. AWS solutions, however, remove a lot of manual work that goes into setup, configuration and maintenance of servers. If you're a startup, it costs you (a lot of) money to hire those people. It also cost you time wasted not working on the core of your service. Up to a certain size, the fixed costs of additional personnel heavily outweigh higher bandwidth costs. Evernote and Stackoverflow have extremely high traffic and might have crossed the threshold at which it's profitable for them to look for savings in hosting bill. A vast majority of startups is not in that position. 2. Unique services demand a premium. Most of the hosting companies provide commoditized services. AWS is much more unique. 3. Services higher up the stack demand a premium. The most basic hosting service offers you a bit of physical space in their building and only give you power for the server and network connectivity. Amazon provides a wide variety of services on top of that. Developing those services and keeping them working is much more costly than just ensuring power is on and network is up. They have to charge more (they have higher costs) and they can charge more (they provide a richer service). 4. Ultimately, the prices are set by the market. We can't really say anything about whether Amazon is more efficient at using hardware or power than, say, SoftLayer, based on their prices. First, their services cost much more to provide. Second, they are free to set their margins at whatever level they please. They could run the service at a loss or they could charge an outrageous premium. We don't know. People are clearly willing to pay for AWS services so at the very least they've set the prices in a way that, given additional benefits, is competitive with traditional server hosting.
- salem 15y ago