Portals are the New Newspapers

My last post was about newspapers and how they have been deconstructed.   I made the analogy of newspapers as information mainframes – legacy feature bundles based on pre-aggregation of information services.  To paraphrase the point, the Internet crushes attempts to build proprietary aggregations.

Now examine the current feature-war strategies of GYM.  They are building the same generic stack. But the web doesn’t stack. It topples.  Or should I say it tuples?

The core premise of a value-added stack is the friction associated with disaggregation.  I could buy my french fries at McDonalds and go to Wendy’s for the cheeseburgers, but the incremental value of Mac over Wendy french fries isn’t great enough for the effort.  The cheeseburger is the bigger part of the value chain. So it drags the inferior good along with it by aggregation, unless and until Wendy’s fries really suck.

The classic strategy for a David fighting Goliath is to go deep. Be best of breed on one thing.  It doesn’t always work, but it works best when the friction against switching is lowest.  Where else is it lower than on the Internet?

If GYM are building Web mainframes, how do they get deconstructed?  Web services deconstruct them.  The first salvo was fired today by Alexa.  As I said a few months ago, business models are going to be the bottleneck to web services adoption, not technology.   The one advantage that GYM have is they can make the end user experience free with advertising.  Of course what they are really doing is allocating the value received (the cost per click) across an internal stack of services.   The David fighting a Goliath must live or die on his service.  His value can’t be buried in the stack. 

Alexa has created a pricing algorithm for their API.  This may be the first explicit price-per-use model for a public internet web service API.  I cannot think of any others, though there may well be some.   Alexa now has put anyone and everyone in the portal business. I could host Alexa and use Google’s Adsense and, Bingo! I have my own monetizing portal on a $10/month web site.   Add APIs for calendars, instant messaging, and other web services, and the GYM stack is entirely blown apart.   Revenge of the long tail.  Let a million portlets flourish... Vertical value stacks are replaced by horizontal services.

The only thing standing in the way is the evolution of market-based prices for other APIs.  What’s a calendar worth? $1.00/1000 uses? What’s web email worth?  $1.00/1000 messages? What’s Yahoo’s VOIP worth – oh, we, know, its $1.00/minute.  And so forth….

This will be the new service economy – the Web Services Economy.

I may be the only one, but I see an elegant irony.  Newspapers are being decimated by unbundlers.  These are the first generation Internet services (Yahoo, Ebay, Google) who unbundled content from advertising, and unbundled them both from physical delivery.  Search has been the killer application – the cheeseburger.   

However, the Web’s pressure for unbundling is relentless. Witness skyrocketing tiny companies like Goowy that do nothing more that allow you to use best-of-breed web portal services without being tied to any one portal. If I can now re-package search by using an API from the open market, portal-based search is no longer the cheeseburger.  It is just dead meat. 

Portal strategies are antithetical to a Web Services Economy.  The pressure to deconstruct is relentless. The center cannot hold. Just as newpapers are the new mainframes, Web portals are the new newspapers.

Comments

[jim Forbes]

I think your analysis of Alexa is masterful. i also think many newspaper will not be intellectually up to the challenge of shifting to a new model. There could be some exceptions to this however, and I think mostly they will be in the small community newspaper market where the paper is an intrinisic part of local life. The same could be said for local radio stations whose websites are underutilized as gateways and portals.
--jim forbes

[Brian Schneeberg]

Along the lines of your reasoning, I think MS, Yahoo, and AOL are all now realizing that they cannot force folks to only see their bundled portal content, but must provide the functionality to allow them to add their own favorites (usually ala RSS). Hence, you have MS Live and Yahoo Widgets coming to the fore. But, I'm with you, longer term I think it is all about web services, which is what my company is betting on (along with embedded analytics and data mining). In fact, we are developing a next generation portal that will eventually allow users to pick and choose from any web service they'd like to interact with (not just monitor) through our platform. In terms of business models, I think micropayments (i.e., pay per sip for various web services) and advertising will be key.

I'd love to get your feedback on what we've done so far with our Mentations offering; you can download the beta from www.mentations.com (right-click on things for options).

[Norm Waite]

Here is a big point, bigger than Google so far:

By a wide margin, the next big step forward in ads yielding revenue (ad ROI) is to be sure, whenever sending some bytes for an ad, send the bytes considering carefully the characteristics of the person being sent to.

Three stages of ad ROI: (A) With old world ads, want to sell something, then pay for an ad in the mass media. Nearly none of the people seeing the ad are likely to respond at all. (2) With Google, search on "MIDI music format", and can see a lot of ads for rock music, country music, Broadway show music, light classical music, MIDI keyboards, etc. Better than (1). (3) Send only ads on the music the person is really interested in, say, opera by Mozart, Puccini, Wagner, and Verdi. For such a person, all the ads for rock music would be wasted or even irritating. With such better 'targeting', ad ROI has to be several times higher than with the best available so far.

Privacy concerns? Sure.

So, the company best able to have and use the characteristics on the people will have a big advantage. Holding these characteristics is nearly a natural monopoly. Tough to believe that Google won't be the best at this. Closely related, improve the accuracy of Google searches by a factor of several by, again, considering not just the search terms but the characteristics of the person searching.

For how to 'process' such 'characteristics', some mathematics anyone? Ah, what a shame, that 'information technology' should have to encounter some mathematics! I mean, what does mathematics have to do with 'information'!

[Zac Pessin]

Rip,
May I recommend the book:
The Support Economy.

Your discussion of dis-integrating original bundles of service to allow better intermediation is taken one step further as a form of concierge layer between end-users in an Economy and "manufacturers" of services and goods. No longer are even the "manufacturers" of services in complex economies the best at delivering those services to users in the high-touch manner that will consistently command (and win) higher premiums from consuming audiences.

The logic is not new, but it's nice to see you highlighting its adoption and acceleration.