Forum Moderators: open
At first glance, INK-optimized sites look like they'll fly fairly well on Fast. Still, a few "survival" questions come to mind.
What does the algo target as spam?
How deep is the spidering?
Does Fast like third-level domains? (something.mydime.com)
Submittal frequency? X per day, or shall I just unload DumpTruck on it?
Ink optimizeds don't do real well from what I can see - Alta is more appropriate. And I know you have been around long enough to remember; it *feels* like the old Infoseek of 97-98.
There really isn't much you have to do other than simple optimization. Kw's laced through headers and page.
Yes they index third level domains.
Submittal freq is a relative question since they only accept the base level domain. So have those links in order for the spider to follow off the home page. They go atleast one level deep, and I think it depends on how many pages total it finds whether it decides to go two levels deep. Here is one place where cloaking is still appropriate - how long it takes Fast to wise on that is anyones guess.
The difficult thing about Fast is detecting how many pages of yours it has in the index. Lycos is returning a number on 'links', but darned if I know where it is getting it (it is not from fast that I can detect).
Links: there does appear to be support for the "links:www.domain.com" syntax, but it doesn't appear to be accurate. Link numbers as tracked since last week tuesday are fluctuating quite a bit. It also looks like any domain with a "-" dash in it is not accurate.
The cool thing about the sheer size of Fast's database (300million pages) is that they find things others don't. One of the big things they run into, are LOG FILES. Search on some standard terms found in log files or apache index files (left as an exercise for the reader). You can find some interesting stuff. (Like cloaked pages from sew).
Which reminds me: link back to yourself on every page. If a page theif is going to steal your page and cloak it, at least get credit for the link to the root domain.
I just went and ran some spot-checks. I'm not doing too well with Fast. Not terrible, but not anywhere near as high as I've been doing recently in AV and INK --but my rank there is elevated because of themes. *sigh*
Currently for one of my sites - I have 3800 pages out of a total of 4500 in Fast's database.
To check I typed in the usual url:mysite.com. Am I missing something?
Pete, how did you submit? One root url, or more?
At the time, I was big on hallway pages. The site has loads and loads of accommodation listings. So I submitted hallway pages with links pointing to these pages as well as the main site pages themselves. I had submitted approximately 66 pages (4 a day), and then bang - fast started crawling and deep crawling. (Total sessions = 142 and hits = 7490) towards the end of Jan, beginning of Feb. Pages were indexed and all over the place for the keywords (site took up 6/7 out of the top 10)for a number of searches which was pretty worrying.
But, its now June and all the pages are still there and holding there positions which is great!
Regarding their relationship with Lycos, it would obviously be great for me if it became permanent. But, the relevancy of their results can be seriously dodgy. A client wants his site optimized for the term "weight loss". Check out the results below:
http://www.ussc.alltheweb.com/cgi-bin/search?type=all&query=weight+loss
If htey carry on producing this, they will have to kiss their partnership with Lycos goodbye!
Maybe the limitation is something new? Or I'm just seeing a 'depth limitation'?
Now that fast is allowing full urls, has anyone noticed Fast actually going straight for that url? I can't. It has been going back to the root domain and then walking down into the site.
- "url : domain doesn't actually work. It includes all references to your site including links. It might also contain pages with just "url" on them."
Quite right. But, I stand by the 3800 out of 4500 total. This new site has poor (in fact - shocking) link popularity. The pages listed are from the site.
and
- "Now that fast is allowing full urls, has anyone noticed Fast actually going straight for that url? I can't. It has been going back to the root domain and then walking down into the site."
Checked out the log files - Fast's webcrawler was hitting the full URL direct and then walking.
