I just left Joe Kraus a comment that describes why I think his startup, JotSpot, is doomed – or at least will have a very short Tail. J The thing is, in order to be the Head of a healthy, ever growing Long Tail, many, many, many (did I mention “many” enough times?) people and businesses must be able to benefit (i.e. make real money) from your Body (e.g. API, tools, etc.). Which is why I believe SalesForce.com has a viable Long Tail because it's supported by a great body that consists of the Sforce Platform and an On-Demand Architecture.
In Google’s case, as most people already know, its search engine is the Head while AdSense is its Body, and all the bloggers and Web-based businesses who display Google ads via AdSense are part of its Long Tail. To that end, Google is attempting to make its Long Tail longer (and fatter) with Froogle, Google Maps, Google Scholar, Google Print, and on and on. Essentially, Google isn’t just trying “to organize the world's information and make it universally accessible and useful” – at the same time, they’re also trying to monetize the world's information and make it universally reusable for advertising. Of course, they wouldn’t say the latter statement anywhere because, oops, it just doesn’t jive with their well publicized “don’t be evil” corporate code of conduct.
To wit, remember the whole brouhaha over the Google Toolbar’s AutoLink feature a few weeks ago? Well, simply put, if you (even just attempt to) monetize someone else’s information, intellectual property, or hard work without that someone making a fair share of money, you are being evil. I wonder how folks like Winer, Scoble, Rubel, Mossberg, et al would have reacted to AutoLink if they were able to easily profit from the technology without doing any extra work other than to continue writing blog posts and articles? In AutoLink’s case, Google launched the Head without a viable Body to support the Tail. J If it's so evil, why will Google still release it? Because it provides a very significant shortcut to Google's attempt to monetize the world's information by potentially monetizing every Web page that a Google Toolbar user sees.
Interesting perspective. I'll think about when I want to start up a business or offer a service.
I must qualify my "so evil" phrase above by adding that it would only apply to AutoLink as soon as Google generates any direct or indirect revenue from the Web pages that AutoLink links to. From what I can tell, AutoLink referenced Web pages do not have any ads -- not yet, that is. Hmmm, if Google would guarantee that AutoLink references Web pages will never have any ads, then I would actually lean towards embracing AutoLink as a "not evil, and very useful" feature.
One more thought on this: Google would have to do 2 things in order for AutoLink to be "not evil" - at least from my perspective:
1) as stated in my previous comment above, guarantee (or at least mandate) that AutoLink referenced Web pages will never have any ads
2) enable Web content providers to use a meta tag (similar to what Microsoft enabled for its IE Smart Tags feature) to disable the AutoLink feature on each Web page
I appreciate your perspective on JotSpot. I honestly don't know if we'll be successful in our approach, but I do know that it certainly seems worth trying. Let me see if I can address your points in reverse order.
Richness of experience:
I agree that fat clients have it way better than the web, but I think there is a serious, long-term shift toward rich internet interfaces. Certainly Oddpost, Google and others are/were leading the way and we're investing pretty heavily in this area. But, I concede the point entirely, the richness of web based interfaces will remain inferior to think clients for some time to come.
Lack of APIs:
This one isn't true. In fact, JotSpot's philosophy while not open source, is certainly open data. We haven't published the documentation around our APIs, so they might as well not yes exist, but we intend to have a large variety of interfaces in and out.
I agree with you on this issue. Or more to the point, that it is an issue. Like rich interfaces, I think time is on the side of hosted services but the question is 'how soon'. We believe that JotSpot can be successful as a hosted service, but you should see some news from us soon on this front as well.
Thanks again for the comment.
For completeness, as I stated in my comment to Scoble's post (http://radio.weblogs.com/0001011/2005/03/15.html#a9639), here are two more features (though these are user side as opposed to the server side requirements I mentioned above) that AutoLink must have in order to be "not evil":
1) no default selection for content providers; the user gets to set this upon initial use/click of an AutoLink (no "auto opt in"), and
2) let the user easily change the content provider selection (simple "opt out") at any time.
We urge you to invest some more time checking out white tail deer anatomy pages.