Welcome to WebmasterWorld Guest from 184.108.40.206
Forum Moderators: mack
Get ready for MSN Google! These results appear to shout "RELEVANCY"!
And according to MSN during the Chicago SES they would be launching by Mid-January! Stay TUNED! :>~
Bring the heat MSN and Yahoo! Google deserves the competition!
joined:Oct 10, 2004
First time I've really paid attention--looks much better than that Google mess.
Because it is a violation of the TOS to link my sites or the sites of anything to which I have any interest, I will post the following keywords, so you MSN beta boosters can see for yourself how brain dead is this MSN beta.
In the MSN beta and in all the other popular SEs, such as Yahoo, MSN (not beta) and others, search for:
Korbel has been around for about a hundred years. It has a domain of www.korbel.com. While www.korbel.com is on the page 1 SERPs of most SEs, it ranks on page 2 of the MSN beta. Why? Because "wine", a second keyword, was added to the search. This resulted in matches deeper in the Korbel web site, which apparently this junk MSN beta is loathe to "contemplate".
This is not nitpicking. It is a major flaw. Sites that provide content on multiple subjects are penalized by this simpleton SE that is the MSN beta. A site that provides content on multiple subjects will naturally organize the site into sub folders and sub-sub folders as the content becomes more specific.
There are literally hundred of sites that appear on page one in all the other SEs that do not appear anywhere in the MSN beta because the content page is not in the root folder
Most SE engines are not so stupid as to believe that "drill down" is some kind of mortal sin. That is a specious argument anyway, because, after all, this in the Internet, and hundreds of other sites that may link to a sub-sub folder do so because that is where the content exists.
As I said before, those that think this MSN beta is the Holy Grail should stop and consider whether anyone will actually use it. This is a massive technical failure, a perfect example of what happens when poorly enumerated requirements are released to a programming team.
Someone higher up in Microsoft needs to be aware of the problems this junk will bring to MSN, and put a stop to it before all the glad handers who think this garbage is "ready for prime time" foist this on the general public.
However, I have a number of sites that feature particular brand name products, and beat the manufacturer in the SERPS, sometimes two or three times. Not all of them are as well-known as Korbel, but some are even more well known.
When I look at the manufacturer's sites that are ranking poorly, it's readily apparent that nobody at the company understands or wants to engage in SEO.
So far, MSN search is a mixed bag. I'm seeing some good quality results in some categories, and not so good in others.
But, for me, MSN is doing a much better job than Google. For the time being....
Even though our site has been continuously the #1 listing for over 2 years on all the other major search engines for a 2 keyword search, Yahoo (MSN), Teoma (Ask Jeeves), Mama, and Google, our domain name does not contain either keyword, so I guess that is why MSN beta does not list any pages from the site, anywhere, using any search terms other than site:<domain name>. (shrug: That faulty SE requirement is sure to produce false and misleading SERPs)
However, www.korbel.com is the only site in the world that is www.korbel.com and it is the manufacturer. The fact that MSN beta does not list the site on page 1 unless a 1 word keyowrd search is used will turn users off. These are the other users that were not already been turned off by not finding sites that have relevant content that are not listed by MSN beta.
After making the last post, I contacted MSFT corporate, requesting someone look into these MSN beta issues before it is released and causes problems for MSN.
None of the other SEs require Korbel to SEO for "Korbel wine" to grant korbel.com a spot on page 1.
Our site (not korbel) is the the #1 2 keyword site on all other search engines for almost 2 years running, yet not a single page of the site is in any MSN beta SERP.
It did not get to be #1 by accident. While I have been given additional optimization tips lately, the site was obviously considered the most relevant by all the other major search engines because it has never slipped from the #1 spot once it achieved that spot almost 2 years ago.
If the MSN beta algo is not defective, then perhaps the site is not listed anywhere in MSN beta because I publicly criticize the MSN beta for its shortcomings.
There has been no logical explanation for the site to be completely excluded, so all that is left is illogical petty reasons for the exclusion of this and hundreds of other relevant expert authority sites.
Csnet seems to be absolutely right with his assumption that they like to find documents in the servers DocumentRoot to rate it high. This must be some kind of M$ logic ...
joined:Apr 13, 2002
Strictly speaking, msn is correct. However it could also be said that MSN search might need to do some semantic guessing, and assume the searcher doesn't know the difference between wine and champagne.
Incidentally, Korbel ranks number one for Korbel Champagne, which is how it should be ranking (if you consider how Korbel sees itself). I just searched for "Korbel" on MSN Beta and it came up number one. So it is evident that MSN knows what Korbel is. It's more of an issue with how Korbel sees itself and how it has decided to market itself. Read on...
Now here is a different search. A search for Roederer Wine has the Roederer Estate in Anderson Valley as the number one result. But Roederer is known more for their champagne, and they don't rank for that. But it's not really champagne, even though that's how you and I may refer to it. It's sparkling wine.
Roederer's poor ranking for Roederer Champagne is not MSN's fault: Roederer Estate is located in California, not in France, so the correct reference to their product is Sparkling Wine, and that's how they are marketing their product (correctly, I might add).
Now here's where it is going to blow your mind: Do a Google search for "roederer wine" and you'll see that Google fails to rank the Roederer estate. MSN wins.
Korbel optimized their pages for champagne, specifically California Champagne. Strictly speaking that may be incorrect. Champagne is from France and sparkling wine comes from everywhere else. So it's Korbel's choice to target the word champagne and not wine, and MSN picked up on that.
The proof of my statement is if you do an allintitle search for wine on the Korbel website: zero results. But if you do an allintitle search for champagne you get five results.
The reverse is true for Roederer. Allintitle for champagne is zero. Allintitle for "wine" produces many results. What's confusing Google is that Roederer modifies most instances of the word Wine with the word sparkling (sparkling wine). MSN doesn't make that disctinction and is able to produce the result.
Who is right and who is wrong? Both are right and both are wrong. It's an opinion of what exists on the web. The results should be different. MSN results should not be measured against Google's results, because then you are presupposing that Google is right. Google is not. It's just their opinion.
Korbel has been around for about a hundred years. It has a domain of www.korbel.com.
csnet, you seem to imply that 'brand age' = internet savvy.
I'll give you a tip. You couldn't be further from the truth. Go and view the Korbel site with a text browser like Lynx. Search engines don't care how long the 'brand' has been around. Tell Korbel to read a few web design guidelines at w3c.org. Ask them what accessibilty means to them.
Let's face it MSNbeta. SUCKS and I was glad to see it was announced, but man, what a mess! The best ranked pages on my site came and went, came and went, and now my whole site is gone from BETA, (does beta mean crap?).
I just read that to mean what it sounds like. But if it makes you feel any better, I hate Yahoo because they kick out sites that have a lot of links. So I know the feeling, but I think one has to look at a bigger picture before judging a search engine.
A technical site (unique content, decent and not spammy, relevant backlinks including DMOZ, PR5 in Google) was in beta.msn on position 15...30 out of 180.000 for a technical one-word keyword search during the past weeks, but gone now. Just gone.
The same search on Google shows that site on posn #1 out of 400.000 since many months.
Number #1 in Google, #2 in yahoo, #7 in the old msn, nowhere in beta.msn.
Must be interesting algos to work out such different results.
Looking at the SERPs of MSN and Google, it seems that both SEs are describing 2 completely different sets of different Internets.
Hopefully the MSN won't roll this "beta" into production until their relevancy algo gets adjusted.
Let's face it MSNbeta. SUCKS and I was glad to see it was announced, but man, what a mess!
Simply search for something like "News" on beta.search.msn.com!
You will find something like "News-Record.com - Piedmont Triad" or "DallasNews.com ¦ News for Dallas, Texas" on the front page between CNN and BBC.
...the key word is RELEVANCY
Korbel's website is horrible, it's almost all images, the word wine doesn't appear anywhere that I could find, yet you want this result to be invented? If they had a half way competent web person, they would be number one for korbel wine, a simple addition to their home page title would do the trick.
Flash splash screen, etc ad nauseum. Pick a better example, one that actually demonstrates failure. What I see is success in this case, the site does not tell msn beta it's about wine, so msn beta doesn't return it. Blame the web people doing their site,not the search engine.
MSNbeta has lots of bugs, too many for it to be close to ready, but this isn't one of them.
My point was that the domain is korbel.com and even if it is a completely brain dead site it should have been included in the page 1 SERP, with or without a "wine" keyword, just based on the domain name alone.
I am sending you a private message with a link to what I believe are much better examples.
Why should a search engine read your mind, fix your mistaken impression for you that the korbel site has something to do with wine? At least based on the korbel site this is an error, so the search engine should not return this site, since you asked for a combination of keywords that don't exist on that site.
Clearly you are looking for korbel and wine. So you should get back pages that deal with korbel and wine. Not Korbel or wine.
Plus that site uses almost all images for its text, no title tags of any significance, really almost no text content at all for a spider to collect.
I'm sure returning the proper pop culture reference is a huge challenge, so it's not surprising that component of the algo still needs a lot of work, it's probably one of the technically more challenging things you can do since there will be so many wide and varied references to any particular pop culture event or product.
I've seen similar weaknesses, in similarly difficult to implement parts of the search algo, this is why it's in beta and not live gold.
By the exact argument you use above about Korbel wine, MSN beta fails miserably. There are literally dozens of sites that use the exact search phrase repeatedly in their headers and in text throughout the sites that are not shown in the MSN beta SERPs and are shown in all 6 of the other SERPs.
This clueless MSN beta lists sites on page 1 that have only 2 of the 3 or 4 keywords somewhere on the page. It is a ridiculous situation, and it is not some difficult process to solve. Even Google gets these right.
Without quotes, all the other SEs give priority to exact phrase matches.
However, if you watch where it doesn't work very very carefully, I think you can get a very good sense of how it's written. As you see component after component begin to function more or less correctly, this will get harder to do. I'd recommend watching it work, and not work, and see how these things can teach you what you need to know to make it work for you long term. this is kind of a unique opportunity.
Just this morning I found a much more severe failure, not subtle, good enough to sticky to msndude. Single word, nothing subtle about that. That one suggested a deeper level error in their real algo.