Forum Moderators: open
Inktomi claim to have "examined" 1.7 billion [I assume on their site that 1.7 million is a typo] URL's. This is not a searchable dB but in theory forms the basis for a large chuck of their algo and is referred to as the "Inktomi Web Map". They claim to have examined these URL's to create, amongst other things, a map of the link structure of the web. In effect, and to be fair they do not claim otherwise, this means that a large proportion of their spidering is for the purpose of building the Web Map, they have no intention of listing most of the URL's they visit. Many here comment that they have many, many visits from Slurp [the Inktomi spider] and never appear in the dB and/or their content is not updated, this is normal as they are building the Web Map not expanding their dB.
Best Of The Web
Inktomi use the information contained in the Web Map to create a dB of 110 million [not a mistype, million] URL's which they call Best of the Web [BOW]. This forms the core of their dB and is the primary index which is searched at their partners, in fact some of the partners use only this dB [and the paid inclusion of course].
GigaDoc
Gigadoc contains approx 500 million URL's and is a combination of BOW and 390 million other URL's which have not made the cut to BOW. Some of the partners search this dB, but URL's contained in BOW are shown first as a rule.
Paid inclusion etc etc
All the Inktomi paid inclusion URL's are contained in a separate dB which is "injected" into the search results, this stands for other paid ads i.e. from their partnership with Looksmart's PPC listings . In my mind I see paid inclusion as equivalent to being included in BOW.
Removal of Dupes
No URL can be in more than one dB, it is one or the other or none. Take for example an existing URL which is contained in BOW, upon paid inclusion it will be removed from BOW and is then only in the paid dB. If you stop paying then it will be removed from the paid dB and from Inktomi itself. In theory, after an as yet undetermined time, it should reappear in the Gigadoc db and then make it's way into BOW [on the assumption it was "good" enough for BOW before paid inclusion].
More stuff here:
The Inktomi Difference [inktomi.com]
I foolishly paid for many pages from my site at the commencement of paid inclusion and changed these to unindexed URLs later. In no case was a previously indexed page (in BOW) out of the index for more than 4 weeks, even though they were never resubmitted.
Another point which may be of interest is the fact that it is very difficult to change the ranking of a BOW page through paid inclusion on a major keyword in a very competitive area (I still have my home page on PI) - the only time the change really shows up is at the same time as the normal BOW update. This can lead to some misunderstandings. In my case I did a tweak and 3 days later moved up some places (not noticing this was a BOW update). Did some more tweaks - nothing happened, left it and noticed at the end of the next month I dropped from #1 to #4 :(
Now I've changed it back but will have to wait until the next update to see the changes (if any).
Moral of the tale - don't pay if you are in BOW.
I'd go a little further. Decide which dB you want to be in and go for that one.
I'm wondering something.
The "inktomi web map". Does the information existing in this section contribute to link pop? I'm a bit confused about this, because it seems that if you search for your links on, say, hotbot, it's only going to report links existing in "best of the web" and "gigadoc". So does the "web map" consist of a whole heap of hidden link information?
For example, 10 of my links may exist in "web map" - those sites may have been spidered by slurp. But they might not be in "best of the web" or "gigadoc" - do these links still count?
Any thoughts?
A search engine goes out, indexes the web, and makes it searchable. The whole goal of the Inktomi spider is to get only those docs they have decided are the "best of the web" and then leave to the side all the rest. Some of which might be commercial, and hence, they have some small justification for PFI.
However, there is a ton of content which is not commercial, and readily available at any decent search engine, such as Wisenut, Alltheweb, and Google. My favorite place comes immediately to mind, Webmasterworld. I'd bet my shorts that Inktomi has not indexed even one tenth of the incredibly valuable, non commercial content which makes up these amazing fora.
That post, and the comments, were illuminating for me. I've not yet paid Inktomi for anything, and now that I read this, I know my decision, and that of so many other SEO professionals and users, to not support Inktomi makes perfect sense.
BTW, I assume that 1.7 billion is a typo, as well. I think they just want to sound better than ahem, google or wisenut. And google only bumped their number up when Wisenut officially launched, lol. What kind of math have they been using???
Welcome to WMW and the ink forum. Yoiu are either in the paid or the non-paid db. One has the potential for ranking well and one is a waste of resources...
-s-
>>>but a marketing tool.
You are right on the money...
-s-
This thread has been most informative! I have often wondered why I have sites which can rank #1 on google and wisenut etc for a range of search phrases - but are absolutely off the radar on Inktomi engines on the same phrases???... I am referring to sites with no paid Inktomi listings (i.e. they are non commercial sites) - and yet they are listed in the Inktomi database. But even though they were much more relavent (as evidenced by Google's ranking, Wisenut etc) - they are pushed so far down - the value of their listing is zero. I think the conclusion is - Commercial sites I SEO I'll pay for Inktomi (and continue to pay)- but non commercial sites - forget em. Why go to all the trouble to pad out a paid advertising service to try to make it actually look like a search engine????
Welcome to WebmasterWorld and the INK forum. I wish INK could find a balance between the cash they need and the search results someone using their product needs.
Thanks for your thoughts...
continued [webmasterworld.com]
