Home Messages Index
[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index

Re: Google Page Depletion Continues

__/ [ www.1-script.com ] on Wednesday 17 May 2006 05:18 \__

> Big Bill wrote:
>> So, you're now better off than you used to be, right? Maybe this is a
>> genuine attempt by Google to encourage us away from the concept of
>> gaming their algo to get to the top of the serps and instead to sit
>> back and observe best practice and let them by use of some obscure and
>> as yet understood method bring us more visitors than we formerly had
>> when our sites were listed as indexed and we had high serps.
> How's that possible, I wonder? You've got knocked down to the third page
> that only 5% of users ever visit yet you still get same traffic? If you
> are not on top of the SERPs, you are not getting any traffic, period. So
> the name of the game is still the same.

Agreed. It is around 5% for page 3 and provided that sites lose their status
(as perceived by the algorithms), they will inevitably fail to top the
SERP's. The number of indexed pages has become rather ridiculous. While I
still get decent traffic from Google (Honest! I am yet to investigate the
reason**), 99% of my pages appear not to reside in the index (site:URL), but
when I search for phrases, the pages are there. It seems like some anomaly
or inconsistency in the algorithms (essentially a mistake). This isn't the
case with every site that I check, so others were lucky. Some sites that I
look at have hundreds of thousands of pages. All still there...

**Maybe it's the "I suffer, but so does my enemy" point of equilibrium.
Before I started 'losing' pages, I had a fortnight of sudden # of referrals

>> I'm serious (again!), this could be their way of dealing with spam,
>> taking away the incentive to be at the top of the search results. Why
>> did we all want to be there in the first place? Because that way
>> supposedly we got the most visitors. Now maybe they are presenting us
>> with an algo that mysteriously gets us more visitors than if we were
>> at the top so there's no point spamming to get there any more.
> You are seriously kidding, right? Randomizing results as a way to combat
> spam? SERP #1 page is supposed to be the most relevant. If 5 seconds later
> it gets "randomized" down to SERP #21, did it become less relevant?


> You are probably giving them too much credit. If something is happening,
> it does not necessarily mean that it's under their control.

Especially when you talk about tera- and perabyte datacentres, which work
like horses and chew bandwidth 24/7.

> I personally think that there could couple possible reasons for the
> current problems:
> #1 they managed to loose a huge chunk of their database on a large enough
> number of datacenters couple months ago. Now they are struggling to sync
> the datacenters having to make do with a smaller subset of the database in
> the meantime.

That'd be an 'ouch'. Could they have pulled out some old stuff from backup?

> #2 "Big Daddy"-style indexer's performance is considerably less than that
> of the old one's. Since the start of the Big Daddy in Nov '05()? crawlers
> bring more data than indexers can actually parse/analyze/store due to more
> sophisticated analysis every page has to go through. Web grows
> exponentially yet I believe they can only buy new servers at a particular
> rate because AdWords income should have flatten out by now. So, given no
> technological breakthroughs lately, they are not going to catch up. Or at
> the very least it is not going to be easy.

Maybe the Register was right after all. Maybe they are choking on all that
spam -- that which used to be excluded. I have always estimated that only
about 10% of the Web was indexed. This could explain the crawling lull that
Borek, yourself and I are witnessing.

Best wishes,


Roy S. Schestowitz      | Windows all-in-one: Word, IE (for E-mail) & iTunes
http://Schestowitz.com  |  GNU is Not UNIX  ¦     PGP-Key: 0x74572E8E
  7:35am  up 19 days 14:32,  8 users,  load average: 0.73, 0.49, 0.43
      http://iuron.com - proposing a non-profit search engine

[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index