| 12:03 am on Sep 7, 2005 (gmt 0)|
msn results are pretty good, just a matter of time before the rest of the world catches on. Reminds me of how G took out A
| 12:35 am on Sep 7, 2005 (gmt 0)|
msndude - I'd like to suggest that you guys offer a little more guidance on how websites can be more "acceptable" to MSN. I know you want to have your own identity but compared to the other big boys I find the results are significantly different. In my case the site is well respected and #2 in the other guy's engines for a generic business keyword. But I'm nowhere to be found in MSN for same word. I'm guessing there is something that MSN does not like but I have no clue what that may be.
| 12:44 am on Sep 7, 2005 (gmt 0)|
MSN has been very kind to our site. It indexes us extremely well and gives us a fair amount of traffic.
I watch my keyword sector constantly and feel the top 100 results much more pertinent than google's serps.
We are a brick and mortar representation site.
I do feel that spam reports are ignored. I made a couple about a month ago and seen no result from it. I know hand to hand spam fighting is hard in such a vast info world but, really, it would separate you from the rest of the pack. I see some of the most common OLD blackhat practices resurfacing in all the major engines.
I do believe that msn has quality results and thankfully it doesn't pin everything on how old the site is and if it is in the odp. I feel you have fresher results...at least in the categories I watch.
Msn just needs a better public relations firm.
| 2:18 am on Sep 7, 2005 (gmt 0)|
I'd love to see every search engine take this approach:
HEAVY HANDED PENALTIES in general that are manually reviewed at the request of webmasters. Spam is a huge problem and it needs to be eliminated once and for all.
Then get some quality control personel in to manually review sites paid by $$$ from webmasters.... maybe $50 for a review or something.
This does 3 things:
1) It gives you less spam and a mechanism for webmasters to dispute being penalized.
2) While reviewing penalized sites you can refine your algo by observing why all the good sites are getting penalized.
3) You can introduce a human element and put some sort of ranking in there based on humans so that the final ranking is human+algo based.
I don't think any algo will ever be good enough, and you can't hand-review every site, but I think penalizing first and reviewing later while tweaking your algo is the best approach.
| 3:02 am on Sep 7, 2005 (gmt 0)|
Too be honest go and have a look at the Australian results. Not good. But I understand you are trying hard. But surely you can do better than what is currently being produced. We have more UK results than anything else.
| 3:47 am on Sep 7, 2005 (gmt 0)|
1. Slow indexing of large sites.
2. Too much algo weight on the page name [ i.e. widget.html ]
3. Too much algo weight on anchor text off the homepage, to internal pages.
4. Favors small, high keyword density pages too much.
5. Very easy to game with old seo tactics.
6. Needs a algo to weed out sites that have no html name after the / . [ directory ]
msndude feel free to sticky if you want more info.
| 4:02 am on Sep 7, 2005 (gmt 0)|
For me start.com is to hard to read with that stupid side bar and colors / also colors on msn need to be fixed make it CLEAN and easy to read and you will win
Have grade one kids or kids that can just start to read > get there views on the search engines: see what they like , make it simple and clean
Eg: the orange font is so rough looking and crappy, to me it looks like bad fonts < smooth it all out and your golden
Keep up the good work
| 5:16 am on Sep 7, 2005 (gmt 0)|
10pm, west coast time, 65 seconds to load 'next' on a search, maybe 45 seconds for the first search page to load, over very fast broadband connection. That's not going to cut it. Still trying to run msn search on windows maybe? Send the server team my condolenscences.
Same time, WebmasterWorld, espn, etc, load in about 1 second, so it's not a local network issue.
Still can't put my finger on why I can't use msn search, usually it's fast, but this type of slowdown can simply NOT happen. This is low traffic time too.
I do like the page html/css code of the search.msn.com page, and the start.com page, that's good, very nice work on that. Opera has some new css bugs, so if there's problems, make sure that's not it.
| 6:01 am on Sep 7, 2005 (gmt 0)|
MSN Search has been down for a the last few minutes for me and is giving out an error message something like:
"An error occurred while processing your request" and it provides a reference number of some sort.
| 6:13 am on Sep 7, 2005 (gmt 0)|
Since when did solicitations for spam reports become headline worthy or even acceptable at WebmasterWorld? VERY uncool.
More than spam reports from WebmasterWorld members, MSN needs users. Once they increase the number of users of their engine, natural reports will start flowing.
For a competive phrase "widget reviews", I get 50-100 visitors a day at number 5 on google and 5-10 a day at number 1 on MSN. The number triples or quadruples during xmas.
Let me repeat .... the problem is not spam, the problem is increasing the user base and getting the "buyer with the credit card on the keyboard" to use MSN.
| 7:13 am on Sep 7, 2005 (gmt 0)|
Nice to know MSN is listening.
My site is getting about 24600 referrals from Google alone and compare to that msn is sending only 750 a month. My site is about technical tutorials etc and fully indexed in MSN.
This site is up since 5 years now and i haven’t changed or optimized it since last 2 years and it is never optimized professionally and see the difference between you and Google. pm me i will give you my url if you want to compare.
| 7:33 am on Sep 7, 2005 (gmt 0)|
Site descriptions using parts of the meta keywords tags need to be taken out of the SERPs. It does not look good having a list of keywords in there.
| 8:00 am on Sep 7, 2005 (gmt 0)|
I think some of you miss What MSNDude ask us!
MSNDude want to know where they are short in racing with other SE.
Some of you like them just because your site ranking with them well and vice-versa.
But what I like to told to MSNDude is CONVERSION RATIO.
MSNDude I found that searches from G converted 3 out of 4 but with you I found it 1/10 or even low.
I also ranking with you well but still prefer G on above mentioned ground.
You need to give Traffic which really intrested to buy or pay for what they look with you.
I am also aware that this is not my SEO shortfalls that can not convert traffic to sale from you because at same time I converted WELL with G for same site/keywords.
Hope you got me,
Have a nice time,
Thank You for Listning.
| 9:26 am on Sep 7, 2005 (gmt 0)|
|MSNDude I found that searches from G converted 3 out of 4 but with you I found it 1/10 or even low(er) |
That's the key thought! Good observation Mann, if it is reflected across the industry. It is a better yardstick than click through rates as it measures whether the user is not yet getting a better relevency, not whether the SEOs are achieving their perception of relevency.
How about this - providing webmasters with a cool tool to track "conversions" from MSN natural search (like the G does for Adwords and many stats packages have any way) and MSN's kick back is knowing whether algo changes help surfers go THROUGH the sites, rather than just TOO the sites.
Not every webmaster would use it (including me) as they have their own tools and for me if gives away too much info too quick, but for most, the benefits would outweigh the disadvantages and most would see the end results as better search results for the user. I don't use Urchin, but my guess is that Google are ahead of the curve on you in this area.
| 10:15 am on Sep 7, 2005 (gmt 0)|
thx for giving us the opportunity to speak from my side as well. I think the pros here pointed out the most annoying things regarding the algo already, so let me say something about the interface (i know taste is not discussable):
The results page of search.msn.com IMHO uses too much fancy stuff to look so simple.
1.) a tiny grey dotted line between the results would really clean up the page = line only between different domains, not before an indented paragraph
2.) the HUGE link with the green/white arrow "Show more results from long-domain-names-make-it-look-strange.com" in an extra line steals you space for one more result on a 1024x768 full screen ... make it "More from this site" and drop the pic.
3.) that "corporate blue" for the title of the site as a link is totally boring! use the full blue scale #0000FF and for visited links: make them corporate blue ... go down 1 px in the headline size - the headlines are HUUUUUGE
4.) the top navigation / search box is totally against all interface rules, packed with stuff some people even never heard of: stretch it with the 8 px you saved from making the headlines smaller.
For a mass product: settings, help, espanol and the search builder link should be under (or around) the logo to seperate pure FUNCTIONALITY from CONFIGURATION and SUPPORT.
With these changes, I think I would like the layout, but I also think this interface is not mass friendly and I am no hero in design, but you asked for opinions ;-)
| 10:54 am on Sep 7, 2005 (gmt 0)|
Your major problem is a lack of temporal depth of data. Your map of the interent over time jsut isn't up to the job of powering a major search engine yet. To fix that, you can wait, or buy
The rest of your problems are related to the fixes you have resorted to to compensate for this basic problem. You are over reliant on on-page factors, and I don't think you've got your link data handling balanced correctly yet
Oh, and your hub / authority model is skewed, but I don't have the depth of search data myself to accurately describe the problem, but I HAVE seen some shocking examples
| 11:01 am on Sep 7, 2005 (gmt 0)|
I will support the comments about spam.
On Google, I rarely see blatent spam - reguarly pages that are spam but take a bit of investigating to verify that spaming is occuring.
MSN - the spam jumps up in your face - blatent hidden text (not even cleverly hidden) and missleading redirects to other sites.
I will PM you some examples as you suggested in your opening message.
As to search relevency etc - that is always going to be subjective to the person doing the search and we SEO's will always complain when we are not top :)
However, I do think the spam problem is quite serious at the moment - and the feedback option is frankly pointless as has already been pointed out.
Thanks for your time.
| 11:20 am on Sep 7, 2005 (gmt 0)|
To gain creditability your SE needs to deliver traffic and at the moment that is not happening ... for my sites at least and I seem to recall others have made similar comments.
I'm talking about top 3 ranking for a very competitive travel and tourism niche, so we are not talking about some obscure topic. The MSN traffic is so low that it really has to be considered as irrelevant.
You can perfect the algo technology until it is undeniably the best of them all but without traffic, the webmaster/SEO fraternity will see it as a failure.
I'm glad you are listening, so..."Please sir, can we have some traffic?"
| 12:35 pm on Sep 7, 2005 (gmt 0)|
The traffic conversion is the main problem - I am doing very well on MSN, especially with my new sites being on the first page near the top, but I'm getting very little traffic compared to Google where the sites typically feature on pages two or three.
Surely MSN gets oodles of traffic, so what can MSN do to improve this?
| 12:42 pm on Sep 7, 2005 (gmt 0)|
Open up your search algo code.
1. You'll see increase in search traffic and involvement of people (initially mostly webmasters)
2. you'll get more people to learn MS technology
3. Algo will improve
4. Other Open source advantages
| 1:10 pm on Sep 7, 2005 (gmt 0)|
Help webmasters earn traffic. Develop a system where webmasters can understand why their page is ranking where it is. If webmasters know the reasons behind ranking they will improve.
Imagine a system that tells webmasters things like, this page/site looks fine, but ...
... site is registered for just one year?
... competitor has 2 extra internal/external links that this page
... not sure why this related keyword is not on this page?
... this para looks spammy
Webmaster will know what to do ... and we know you can have a big Disclaimer page.
A system like this will open up glitches in MSN search algo to the world. But will the problems cease to exist by hiding them?
Help webmasters earn. They will promote MSN search.
| 1:23 pm on Sep 7, 2005 (gmt 0)|
Uber_SEO conclusions also reflects my experience. I'll send you a sticky with what I think is a good example.
| 1:41 pm on Sep 7, 2005 (gmt 0)|
--To be honest, it looks like it favours poorly optimised, poorly linked in sites, over big sites that you'd expect to see at the top of the SERPs.--
MSN Dude, Don't heed this. Your objective is to deliver the best and most relevant pages to your customers. I don't believe you are in business to please the whims of professional optimizers and their customers.
There is no reason why a poorly linked site without optimization should rank below another one. The value of the site is more important that what container and tricks were used to publish it.
As for constructive suggestions, I would say, talk back to people and try to answer their queries when they contact you. I have sent sticky notes to you and the team at MSN Search, and you never replied. If you start responding to Webmasters, then you'll gain their trust and they will do the promotion for you.
It's another way of tackling the problem of making MSN better, by using real PR - i.e., public relations and customer service, as opposed as using an algorithm and technology. If you can improve on this and do it better than Google, you will gain something over which one cannot put a dollar sign easily - good will.
| 3:27 pm on Sep 7, 2005 (gmt 0)|
The spam and duplicate content is definitely an issue. In one of my client's fields, I notice that one of their competitors has come up with three or four sites that have the exact same content (except for different colored pages and different domain names) and they take up the majority of the first 20 listings for their keyword - they even have the company name in the page title, so it's pretty blatant. I've reported spam over and over the past three months or so, but the same people are there. And why did you take the number of results on the page OFF the page? (it used to say 1-30 of XXXXX results, and then the next page would say 31-60, and so on)
| 3:40 pm on Sep 7, 2005 (gmt 0)|
Wow, great conversation and I appreciate that it is fairly balanced.
Our goal is to provide the most relevant search engine for end users. Period. At the same time we want to provide webmasters with the best possible experience when our search engine interacts with your sites. This includes everything from politeness to thorough index coverage. Over the past 6 months there have been small things we have done for this and we will continue to do more to hit both goals.
One other topic that has been raised is the amount of traffic that MSN Search drives to your site. A number of folks have said that Google drives 3 times as much traffic as MSN. It is worth noting that this may be inline with other larger trends. Depending on what market research / competitive intelligence firm you believe Google has about three times as much search traffic as MSN Search in the United States. Not something we are proud of, however, I want to make sure people understand the context. Additionally, different search engines target slightly different demographics. While your site may appeal to users of search engine X it does not mean that it will definitely appeal to users of search engine Y.
Hope this helps.
| 3:46 pm on Sep 7, 2005 (gmt 0)|
One thing I really appreciate about MSN's search engine is that you are willing to include new sites in your SERPs much more quickly than Google (or Yahoo).
I realize this increases the risk of displaying spammy sites, but it really helps new sites.
| 5:08 pm on Sep 7, 2005 (gmt 0)|
One thing that I've noticed in a big way, (in the gambling related industries) is the sheer lack of spam pages that redirect to the actual gambling sites, with affiliate id's attached. There must be hundreds, take a look at 'poker' for example... I've submitted spam reports on tons of sites in MSN, I don't think many have been dealt with.
| 5:19 pm on Sep 7, 2005 (gmt 0)|
I dont follow MSN search to closely but some people seem to be very happy with the freshness of the crawl and looking at my results the pages that are in the index are fresh.
Obviously this is good but the depth of the indexing is not so good - so perhaps more concentration should go on deep indexing with less freshness for these pages - and keeping the fresh of the fresh for the cream at the top.
Just a thought.
| 6:42 pm on Sep 7, 2005 (gmt 0)|
Thanks for engaging more than in the past - I think your team will find the WebmasterWorld crowd very informed and helpful as you improve and advance the MSN tools.
Also, I hope you'll consider sending more reps to the Las Vegas Conference - I had a long talk with one of your team members in New Orleans and was very impressed with his sincerity regarding making MSN a great search tool.
Google has been superb in making their search engineers available for extended question session and you should consider this approach.
| 9:36 pm on Sep 7, 2005 (gmt 0)|
"Our goal is to provide the most relevant search engine for end users. Period."
False, and you know it. By geolocating searches you deliberately as a matter of corporate policy choose to not serve the most relevant results.
And THAT is the "period".
And then of course, having relevance as the "goal" is so 1997 that it is mind boggling. Look for quality MSN, look for quality. Any search engine looking to relevance as a goal these days will be, well, it'll be as bad quality a search engine as MSN is.
| 11:47 pm on Sep 7, 2005 (gmt 0)|
|A number of folks have said that Google drives 3 times as much traffic as MSN |
Have you done a proper survey across a reasonable number of sites? You might be hearing what you want to hear.
For my sites, with a variety of audiences, MSN referrers vary from 0.2% and 1.2% of the Google referrers (on two sites with millions of pages), to 6% on several medium size sites, to a maximum of 13% on a site with few pages.
Just to clarify that first figure, it's 4,007 referrers from Google yesterday, 9 from MSN. Way less than a third...
It does seem to be inversely proportional to the percentage of the site indexed, so if you could keep most of the pages you crawl in your index I'd expect your share to rise.
I do realise driving traffic to sites is not the objective, but that is a visible side effect of being successful at finding what the searcher is looking for.
| This 128 message thread spans 5 pages: < < 128 ( 1  3 4 5 ) > > |