Amazon owns the cloud

Back in May I posted about Amazon's Q1 results, the first in which they broke out AWS, their cloud services, as a separate item. The bottom line was impressive:

AWS is very profitable: $265 million in profit on $1.57 billion in sales last quarter alone, for an impressive (for Amazon!) 17% net margin.

Again via Barry Ritholtz, Re/Code reports on Q2:

Amazon Web Services, ... grew its revenue by 81 percent year on year in the second quarter. It grew faster and with higher profit margins than any other aspect of Amazon’s business.

AWS, which offers leased computing services to businesses, posted revenue of $1.82 billion, up from $1 billion a year ago, as part of its second-quarter results.

By comparison, retail sales in North America grew only 26 percent to $13.8 billion from $11 billion a year ago.

The cloud computing business also posted operating income of $391 million — up an astonishing 407 percent from $77 million at this time last year — for an operating margin of 21 percent, making it Amazon’s most profitable business unit by far. The North American retail unit turned in an operating margin of only 5.1 percent.

Revenue growing at 81% year-on-year at a 21% and growing margin despite:

price competition from the likes of Google, Microsoft and IBM.

Amazon clearly dominates the market, the competition is having no effect on their business. As I wrote nearly a year ago, based on Benedict Evans' analysis:

Amazon's strategy is not to generate and distribute profits, but to re-invest their cash flow into starting and developing businesses. Starting each business absorbs cash, but as they develop they turn around and start generating cash that can be used to start the next one.

Unfortunately, S3 is part of AWS for reporting purposes, so we can't see the margins for the storage business alone. But I've been predicting for years that if we could, we would find them to be very generous.

Be Careful What You Wish For

Richard Poynder has a depressing analysis of the state of Open Access entitled HEFCE, Elsevier, the “copy request” button, and the future of open access and Bjoern Brembs has a related analysis entitled What happens to publishers that don’t maximize their profit?. They contrast vividly with the Director Zhang's vision for China's National Science Library., and Rolf Schimmer's description of the Max Planck Institute's plans. I expressed doubt that Schimmer's plan would prevent Elsevier ending up with all the money. Follow me below the fold to see how much less optimistic Brembs and Poynder are than I was.

IIPC Preservation Working Group

The Internet Archive has by far the largest archive of Web content but its preservation leaves much to be desired. The collection is mirrored between San Francisco and Richmond in the Bay Area, both uncomfortably close to the same major fault systems. There are partial copies in the Netherlands and Egypt, but they are not synchronized with the primary systems.

Now, Andrea Goethals and her co-authors from the IIPC Preservation Working Group have a paper entitled Facing the Challenge of Web Archives Preservation Collaboratively that reports on a survey of Web archives' preservation activities in the following areas; Policy, Access, Preservation Strategy, Ingest, File Formats and Integrity. They conclude:

This survey also shows that long term preservation planning and strategies are still lacking to ensure the long term preservation of web archives. Several reasons may explain this situation: on one hand, web archiving is a relatively recent field for libraries and other heritage institutions, compared for example with digitization; on the other hand, web archives preservation presents specific challenges that are hard to meet.

I discussed the problem of creating and maintaining a remote backup of the Internet Archive's collection in The Opposite of LOCKSS. The Internet Archive isn't alone in having less than ideal preservation of its collection. It's clear the major challenges are the storage and bandwidth requirements for Web archiving, and their rapid growth. Given the limited resources available, and the inadequate reliability of current storage technology, prioritizing collecting more content over preserving the content already collected is appropriate.