Forum Moderators: bakedjake
I'm 3rd for my keywords on msn.
But in straight ink I'm no. 1 and in L$ I'm no. 1 (after the promos)
The only other factor I can think of affecting things here is the difference between L$ and M$Ns algorithms. I can only imagine M$Ns algo also includes either link counts - which I have less of than the M$N no 1 and/or DirectHit influence. The no1 guy is number 1 in DirectHit and has thousands of incoming links. I have neither yet (well 10's of incoming links).
Hopefully some clues there ;)
Marcia -- the risky bit is that you have very little control of the description - though I was quite pleased with the one I received. Better than the 5 measly words Yahoo could be bothered with.
In terms of where I would spend $299 - Yahoo every time. Unless my target market is very M$ biased.
Gethan
When you're talking about your MSN listings, are you referring to LS or Ink? If it's Ink paid pages, it's not links, it's what's on the pages, from what I'm seeing. Some new pages I had added last week are ranking just as they'd be expected to be based on looking at the other sites for the keywords, and there's no link factor involved with those pages at all. Oh, they're linked between them with text links, and in fact the index page was pushed up a bit, but that's on-page, the added text and increased keyword density. For the new pages, I'll do some minor modifications on a couple and watch the changes.
algorithm involving the URL, title and description for many of the listings returned However it also appears that some of the top listings do not fit the algo. Perhaps there is some payment for placement at work here, or some off page criteria
Mark, in some cases the URL, title and description appear to be a relevant factor, but when you say some of the top listings do not fit the algo, those are exactly the ones causing confusion.
Payment for placement? Who would have been paid? There's nothing on the LS or MSN sites, publicly available anyway, about being able to pay for preferential placement. The question is whether MSN is applying an algorithm to the directory sites, which would make the results more uniformly predictable than they are, or whether they're coming from LookSmart already assigned some sort of point system. In other words, is the ordering mechanically determined, or a result of human intervention of some kind based on site quality (or whatever)? The lack of uniformity is what makes it so questionable.
Nothing is ever random, ever.
>no way to get onto the first or second page
Sometimes you can only achieve what is achievable, look for the best ROI.
Tip 1- Work with the source when viewing the SERP's.
Tip 2- Use the MSN directory to identify the Cat. If you are looking to rank at MSN don't even look at LS, it will only confuse you.
Please explane that some more
/Ove
Possibly a combination of the two, perhaps the LS reviewers add a site rating and additional "hidden keywords".
MSN - Web Directory ranked 3 of 31
and in the webpages no 1 (or 32nd).
Looksmart - Ranked 1
Inktomi - Ranked 1
DH - not ranked at all.
I can only conclude that MSN uses a different algorithm to search the Looksmart data. So anyone got any more ideas?
There seems to be too many terms and too wide a variety of sites that turn up as those confusing top results. So, I would discount the idea of a secret unnanounced pay msn scheme. The fact that we haven't heard of it here and it would seem to be widespread by the serps if it existed makes that unlikely.
I also see those weird cases where analyzing only the publicly visible category data (url, title, discrip) fail to explain the positioning.
I don't see any strong correlation to inktomi results.
It does seem to be related to link pop, but that may not be a direct relationship(ie: webmasters who know how to rank well in msn also know how to rank well in google).
The sites in my limited study that are screwing up the results seem to be consumer comparison or news type approaches.
Try "golden" 1-6 are on topic (sort of) but there is barely a mention of the keyword. Results from 7 onwards are obviously using the Title, URL and description.
Others to try are "wood" & "timber".
Some notes about MSN.nl which has a similar system:
[webmasterworld.com ]
The new Microsoft Search logic knows how to iterate through a query, break the query string apart, and figure out which words go together. Behind the scenes, it constructs a Boolean search that essentially says, "Look for content that contains all of these terms, or any known acronyms for these terms, and get me the Best Bets and Preferred Terms suggestions for each term." For instance, a query for "IE5 Win2K" would expand to something like this: "("IE 5.0" or "Internet Explorer 5.0") and ("Win 2K" or "Windows 2000")." This translation happens behind-the-scenes and is transparent to the user.
Does that section hold the key then ?
Seems that it may do for these two searches:
timber [search.msn.com] & lumber [search.msn.com]
The results are nearly identical.
advanced synonym matching
