Tim Wu wrote a fantastic book on the history of communications and media where he pointed out that patterns tend to repeat themselves with every new medium, starting with an open ecosystem full of enterprising entrepreneurs leading eventually to a closing down opportunity set due to the market power of the new winners. In the book he covers the telephone, the radio and television and forecasts that the same will happen to the Internet. With today’s Kindle Fire announcement, things certainly seem to be headed in that direction.
First of, as many will write, the new Kindle strategy is terrific. It’s about the content and the price. Even their “top of the line” Fire at $200 has the chance to become the first personal mass market tablet device. At current course and speed, I could see the iPad becoming the family desktop of the tablet space— a shared machine that is bought once per household while the Fire takes the “one in every bag” position (of course I doubt Apple will stand still this time next year when Amazon has sold 1/2 to 2/3 as many Fires as iPads).
Back to Wu’s book: the most significant part of yesterday’s announcement to me was Silk, the browser that runs half its MIPS on Amazon’s web services. Nominally the idea is that the web’s architecture is not ideally suited to a low bandwidth, screen & CPU constrained device and that by pre-processing all of the requests and content on a powerful server, the experience can be sped up significantly. It’s not a new idea: Blackberries worked like this along with Danger’s Sidekick, Opera Mobile, and Skyfire’s browser (an investment). However, I’m just not sure that the improved user experience is the end of the story.
After all, the Fire is a Wifi only device so unless they are thinking forward to some world where they go 3G (instead of LTE), it can’t be bandwidth alone. Plus, the existing mobile ARM cores are plenty zippy when it comes to running web content (witness iOS). Even the Kindle Fire can be set (likely through some deep options menu no one will find to “client only” mode). And to boot, the cry of “mobile first” has publishers and app developers racing to optimise for the new relevant target of the tablet and the smartphone.
Additionally, why take the execution risk of server-side rendering? Not only does this add a variable cost that Amazon has to support for every user surfing their web on the Fire, but it also creates a dependency on a web service that has to be managed, debugged, and made compliant with all of the browser idiosyncrasies that currently live on the client.
I believe the real reason for squirting all of the web through Amazon’s AWS straw comes down to control, the most important bit being ownership of the clickstream data. Plenty of people have raised the privacy issue thus far and have quoted Amazon as saying that the data will remain anonymous and aggregated is different from not captured or “incapable of being used to build profiles.” At my last company I had someone from Amazon’s early days (before cheap storage, Hadoop, and the big data obsession) who told me that Amazon had kept every click on every log since inception despite the fact that it used to cost a lot of money to do so. There is no doubt this is a company that understands the value of data when it comes to discerning shopping intent, the lifeblood of a low margin high volume e-commerce retailer.
If I were Best Buy or B&N, I’d be worried. All signs today point to 7-10% of e-commerce happening on mobile devices (smartphone, tablet), with some people believing the tablet may come to take 50% of that in the next 2-3 years. With the Fire Amazon has effectively extended their storefront to this new platform in a closed and proprietary way, a completely rational strategy. But with Silk, they’ve gone one step further, choosing a page out of Microsoft’s old “embrace & extend” playbook applied to something much more relevant than the DOM or some other geeky browser standard, the currency of commerce on the web (data). Bold, and bit scary.
In a day when all the rest of the Web 1.0 companies seem like dump trucks crashing into each other in the night, who’d have thought that the humble little bookseller which just 12 years ago was jokingly known as Amazon.org (no profits) would come out on top of the next major phase of the Internet?