[Sigia-l] "Best Bets" the Yahoo way
Listera
listera at rcn.com
Wed Mar 3 04:41:31 EST 2004
"Christina Wodtke" wrote:
> Yahoo has always had people to pay to be included in the directory...
What Yahoo and Inktomi do is their business, critiquing what they do is
mine.:-)
> Both of them are just ways to guarantee you get included in a timely manner.
Or another way to admit that Yahoo is either incapable or unwilling to index
what's out there in a timely and equitable manner.
> there are two core concepts in web search: pay for placement and pay for
> inclusion.
You mean as framed by Yahoo? There's no law that says you must ask for money
for inclusion. Google doesn't.
> pay for placement are those adsense and the overture ads you seen on the top
> and right of your "real" results. On a given query then can be more useful
> than the results, on others they are merely in the way.
They are not commingled visually or otherwise with the real results.
> Yahoo does not let paid inclusion affect the algorithm,
When $100 million or more annually is at stake, as projected, how can the
user be sure that that is the case?
> though there has been conversations that it might actually improve relevancy,
> since a paying company *might* be a more authoritative source.
Not much point to discussing that further once you intellectually buy into
the notion that one's authoritativeness is indexed to one's
ability/willingness to pay.
> Liek all elements of their algorithm, Yahoo relies on strict blind testing to
> determine relevancy.
So the people who pay Yahoo to essentially go up the relevancy tree are
fools?
This is so much like saying that hundreds of millions of dollars given to
congressional reps have no influence on how they vote because they are "not
for sale." Are there still people who believe this nonsense?
> This is a war folks, between major search players, and until someone wins
> decisively it is way to dangerous from a business point of view to do
> something as foolish and game-ending as offer irrelevant results.
So you're saying Google results are irrelevant because they don't take
money?
> ""Any time you accept money to influence the results, even if it is just for
> inclusion, it is probably a bad thing," Mr. Page said." Yes, it is a bad
> thing when you rival has found a way to get a larger fresher index that you
> have publicly shunned.
That's like arguing that political junkets for congressional reps with
corporate reps on balmy islands is a good thing because sipping margaritas
by the beach lets them get to know each other more efficiently.
I'd imagine it'd be trivially easy for Google, for example, to extract money
to get listed faster and better. But they haven't done it. I'm not sure how
this is bad for the user. Unless perhaps the results I get from another
search engine that takes money for inclusion are spectacularly better. My
initial impression is otherwise.
> BTW, this is a disingenuous title-- a best bet is when someone has done a
> query analysis and has crafted the first few results to be the "best" ones.
My long-standing opposition to "best bets" is that it artificially alters
the results, no matter what the criteria is, no matter how noble the reason
to do it might be. It's an intervention, usually opaque to the user.
If some of the items in the results are there *only because* they paid
(otherwise they might not be included, which is the reason companies pay
cash to get specially indexed), the results are already gamed.
Yahoo cannot ask for money to increase someone's chances of getting into the
results set and then turn around and claim the payment has no affect on
relevancy. Yahoo would then be in the untenable position of deceiving either
their paying customers or the users, or both.
----
Ziya
People don't want to buy a quarter-inch drill.
They want a quarter-inch hole.
More information about the Sigia-l
mailing list