My buddy Al, who works at Amazon, posted an interesting comment on my recent next-gen web hosting post. In it he mentions this Forbes article, The Death of Hardware, that discusses the pendulum swing back to timesharing, a.k.a. utility computing.

Snippets:

The company divulges almost nothing about its costs or margins but is said to run its Web Services business on huge networks of computers costing as little as $300 each. Ten cents an hour adds up to $876 a year in revenue (assuming nonstop usage). If hardware lasts two years and if, let’s say, electricity and other overhead cost as much as the hardware, Amazon would have a gross margin of 45%, better than what it gets on books.

In the last two months of 2007 the number of items stored at Amazon Web Services grew 40%, to 14 billion units. (Units vary in size from a couple of bytes to 5 gigabytes, and Amazon keeps the totals secret.) That’s a faster growth rate than in the April-October period.

Amazon’s s3 storage service now handles 30,000 requests to its database per second.

Surely it is just a matter of time until Google and Microsoft do some flavor of this.

I can see two strategic challenges that would make it tough for Microsoft: deciding which platform to use (versus offering naked or nearly naked linux boxes like Amazon does and Google probably would) and putting together a billing infrastructure for purchasing cloud software and services.

Thanks for the link and your thoughts, Al.

Interesting WSJ article out today: Bill Gates Issues Call For Kinder Capitalism.

In a speech at the World Economic Forum in Davos, Switzerland, the software tycoon plans to call for a "creative capitalism" that uses market forces to address poor-country needs that he feels are being ignored.

Outgoing Microsoft Chairman Bill Gates talks to The Journal’s Rob Guth about his concept of creative capitalism. (Jan. 23, 2008)

"We have to find a way to make the aspects of capitalism that serve wealthier people serve poorer people as well," Mr. Gates will tell world leaders at the forum, according to a copy of the speech seen by The Wall Street Journal.

Scott Annan, CEO of Mercury Grove, followed up on my recent blog posts on tools for private online discussion forums. He wanted to find out a bit more about where Mercury Grove fell short on my requirements, and get any constructive feedback I was willing to offer. Oh, and maybe make me more of a believer in the product while he was at it.   Based on Scott’s input I made a few updates to my second post. I’ve also summarized our discussion below.

Scott explained Mercury Grove is aimed at both intra- and inter-group collaboration, especially within organizations that have lots of group activities or projects running in parallel. In my case I’m really after more of a free-form discussion forum for a relatively small number of people… I don’t have a need for the collab-related features such as task lists, projects, blog integration, and calendars.

I stated in my second post that the integration of blogs struck me as a little odd. Scott tells me this feature is actually used a number by companies with geographically dispersed groups that need an easy way to keep up to date on other groups’ news. Makes sense.

One nice ability that I didn’t notice on my quick test drive was the centralized summary of activity across all groups you belong to. Related web groups need not be information silos.

We also talked about differentiation between Mercury Grove and other online collaboration services like Basecamp. It’s a crowded space. Mercury Grove is targeting the middle to upper end of the market, from what I can see. Pricing is a particular challenge; many people expect everything to be free on the Web nowadays, at least at the entry level. Like Basecamp, Mercury Grove has a free "Community" entry-level offering, which gives you up to 50MB of disk space. I do suspect there is a gap between their Community and SMB lines where they’re losing price-sensitive tire-kickers. Perhaps a Ning-like pricing structure, where you can buy additional storage in 10GB chunks for $10, would make sense? Of course, adding billing and storage in small increments isn’t zero cost. And positioning your product as one-size-fits-all rarely works either, especially when some of your customers are large companies that want to pay a lot more than the little guys. Tough strategy challenge.

In any case, next time I am looking for project collaboration tools I will give Mercury Grove a deeper look.

Thanks, Scott, for getting in touch.