| 7:30 pm on Aug 14, 2000 (gmt 0)|
I had a shock the other day (saturday) submitted to Altavista on Wendnesday and was listed on Saturday. That is a first for me.. Sometimes feels like a hit and miss with them..
| 9:01 am on Aug 15, 2000 (gmt 0)|
I got one in today that was submitted on the 7th of this month, so from the UK it's 7-8 days.
| 1:43 pm on Aug 15, 2000 (gmt 0)|
new site, submitted aug 3rd--showed up this am.
| 4:04 pm on Aug 15, 2000 (gmt 0)|
Yup, there seemed to be an update today, moved up to #4!!!!
| 4:30 pm on Aug 15, 2000 (gmt 0)|
However, I have a BIG feeling that link popularity counts, but in a weird way. One of my keywords is "car rentals" and the top site is Budget. On altavista, budget has 1105 links pointing to it. However, they don't use themes and only a few of their pages mention car rentals. So I am starting to believe that link popularity is a major factor, but maybe just for the top site? If you look at results, the #1 slot tends to have the most links pointing towards it, and from then on it seems to be all themes.
Just some thoughts.
| 5:16 pm on Aug 15, 2000 (gmt 0)|
mildly competitive hyphenated keyword targeted domain (13,399 other sites) in top 5 with 6 content quality but low link quality links + strong themes = theme algo 99.9%, link algo .1%
| 5:40 pm on Aug 15, 2000 (gmt 0)|
Greg's last posted has prompted a question. Do the engines ever use more than one algo for their search results ?
In the case mentioned by Greg the top site could be returned by an algo in which link pop scores well, with the remaining sites given by a themes algorithm.
Has anyone seen any evidence of this sort of thing happening ?
A novel way of testing a new algo could be to use the new algo to return the even numbered search results, with the old algo returning the odd numbered results. Click throughs could then be monitored to determine which is the most relevant.
| 3:16 am on Aug 16, 2000 (gmt 0)|
I don't think what we're seeing here is two different algos. Although Budget could better theme their site for the word "rental" they have used the word "car" at least once in the body text of every single page. They have also used the whole phrase "car rental" once in the title, keywords, and description of every page. Although they might not be perfectly themed, AV's use of relevant link popularity is strong enough to move them past sites with slightly better themes.
| 5:38 am on Aug 19, 2000 (gmt 0)|
maybe they are using cloaking technology? How would a person know?
| 8:06 am on Aug 19, 2000 (gmt 0)|
I also wonder if Alta Vista boosts companies that own a Real Name. That could be a factor in Budget's top spot.
| 3:50 pm on Aug 19, 2000 (gmt 0)|
In regards to realnames, I don't think they help to much, or at all, because I have real names on a lot of domains, and even have one that's #4 now, but its #4 because of themes I believe. I've had realnames for at least 9 months, and for 8.5 of those, my ranking was way down there. But I'm never sure.
| 5:18 pm on Aug 19, 2000 (gmt 0)|
>link popularity counts, but in a weird way
I keep getting drawn back to the related pages feature at AV.
I have a domain targeted at a competitive [ish] two word phrase which in the last week has seen good improvement in ranking. The improvement has been seen across a wide range of less competitive stems as well as the targeted phrase.
What has changed:
A couple of months ago the text hyperlinks, used to navigate the site, were changed to include more keywords. The site moved from the 50's to the 20's.
Overall link pop has slightly improved over the last few weeks but nothing really striking and stands at 23 links.
The pages returned by like:www.thedomain.com have shown the most dramatic change.
To cut a long story short the previous pages returned as "like" were almost exclusively from two identical pages containing links to local companies, which included the domain in question. AV grabbed the sites slightly above and slightly below the domain's listing and showed these as related pages. Interestingly AV also showed sites as "like" from the same page, which although much further down the page, are related to the targeted keywords.
The change now is that AV has included more pages within the "like" results returned which are related to the sites targeted key words. It seems that AV has had a "sudden" realisation of what the site is about and is now showing this through the "like" results returned. If this is true I expect more results to be added to the "like" results and that these will be on theme.
A small experiment, no cheating ;)
The links below will run a "like" search at AV, see if you can guess what the site ranks well for. [Try not to look at the search box or address bar, pretend it's a banner ad ;)]
One [altavista.com] Two [altavista.com]
| 2:50 am on Aug 20, 2000 (gmt 0)|
Can someone tell me how a person can find out if a site is using cloaking technology? Um, please.
just trying to figure out what I am up against.
| 1:47 am on Aug 21, 2000 (gmt 0)|
If they're using a good IP based cloaking system the only way you can tell is by comparing what info the search engine returns to what page you actually see (title, description, etc.)
If they're using a agent based cloaking sytem (which most people don't use), You can use a program like web bug that will allow you to fake different user agents.
| 3:09 am on Aug 21, 2000 (gmt 0)|
when you say information search engine returns, what are you referring to. The site description listed at search engine?
What about pages I try to view source and there is little data for keywords, description, etc... yet they seem to rank high. I am working in the insurance sector. I have managed relatively high rankings, yet I am strategizing to obtain better. I have not used cloaking and basically make my pages search engine freindly and attempt to keep it aesthetically appealing. Alta Vista I am told likes short doorway pages, yet I have not really run across those for the higher placing pages. So I am thinking they may be using cloaking.
| 4:42 pm on Aug 23, 2000 (gmt 0)|
"what are you referring to. The site description listed at search engine?"
Yep, compare this title and description to what you see in the source code, if they don't match don't emulate that page. (remember that some search engines don't use meta descriptions)
"What about pages I try to view source and there is little data for keywords, description, etc... yet they seem to rank high"
It's still possible to rank high without even using meta keywords or descriptions. If they're not using a description check to see if the search engine is pulling body text to use as a description.
"Alta Vista I am told likes short doorway pages, yet I have not really run across those for the higher placing pages."
I wouldn't pigeon hole AV as just liking short pages. Most of my high ranking pages have a good amount of body text.