Umair Haque / Bubblegeneration

Thursday, February 09, 2006

What Yahoo Should Do Next

What amazes me about Yahoo is that it's actually a company with great ideas. Better ideas than most of their competitors - whether Google or media. I mean Semel's four pillars are the segments of my new media value chain. Obviously, Yahoo gets it.

But somewhere between the rubber and the road, these ideas invariably get diluted, degraded, and robbed of all value creation potential. It's like the brains at Yahoo design Miuras - and somehow, what ends up being delivered to the market are K-Cars.

Case in point: Yahoo to trial revenue sharing to consumers for using search.

Now, you might think it's just price competition - but in fact, this is almost a great idea. You can almost imagine the context - some Y exec saying "we can still grab a 20% margin if we share a bit with consumers". True - but doomed to fail, for reasons of massive adverse selection which we won't go into here.

Instead, let me offer some humble advice to the Ys that <3 Bubblegen. Here's how this idea could be great. Remember Amazon's Mechanical Turk? Cross this with that.

Don't pay people to use search - pay people to help improve Yahoo search. Give anyone a tiny micropayment for a tiny contribution to Y search. Leverage the massively distributed specialization of the edge to improve/filter/rank results.

There are many ways to approach the problem - tags, rankings, ratings, etc - for which revenues are redistributed to peers. The exact mechanism isn't so important - the intent itself is.

Think about the huge marginal benefit you realize from this strategy versus the simple price competition of incentive for use - in this case, you realize leverage. Think about what kind of game-changing return you could realize for a piffling few million bucks invested in this kind of strategy.

The result would be the most radical strategic innovation - it would absolutely redefine the search value proposition if done right, and let you begin to reshape the deep economics of search on its own terms.

Think about it, run the numbers, email me if you wanna chat about it.

-- umair // 11:45 AM // 6 comments

Comments:

But people are tagging things for free already right now in various Yahoo! properties and it's not being input into their algorithm. You have to fix this first before you pay people to tag more. I suspect it's on it's way. A good idea you have.

# //

Thomas Hawk // 12:22 AM

Yep - I think rev share is *the* model for social software going forward. However, where money becomes the major motivation to contribute, I think we'll see big problems.

# //

PeteCashmore // 6:17 PM

One good way to improve search info databases content rating and classification is to abandon yahoo dir, dmoz, and other static directories.

Big search engines should stimulate appearance of focused and good designed web pages link exchnage directories. People behind them will work for money.. And thus will provide good results.

The question is - which directories are good, and which are crap?

# //

dog names // 2:56 PM

Good design!
[url=http://erdzamfv.com/pash/ykin.html]My homepage[/url] | [url=http://ugqyijls.com/qqil/pudw.html]Cool site[/url]

# //

Anonymous // 2:50 AM

Well done!
My homepage | Please visit

# //

Anonymous // 2:50 AM

Nice site!
http://erdzamfv.com/pash/ykin.html | http://qlisblcl.com/zdfj/pcrm.html

# //

Anonymous // 2:51 AM