Forum Moderators: open
"Any clue as to the possible role greater reliance on semantics is playing in your never ending quest for more relevant results?"
I'd say that's inevitable over time. The goal of a good search engine should be both to understand what a document is really about, and to understand (from a very short query) what a user really wants. And then match those things as well as possible. :) Better semantic understanding helps with both those prerequisites and makes the matching easier.
So a good example is stemming. Stemming is basically SEO-neutral, because spammers can create doorway pages with word variants almost as easily as they can to optimize for a single phrase (maybe it's a bit harder to fake realistic doorways now, come to think of it). But webmasters who never think about search engines don't bother to include word variants--they just write whatever natural text they would normally write. Stemming allows us to pull in more good documents that are near-matches. The example I like is [cert advisory]. We can give more weight to www.cert.org/advisories/ because the page has both "advisory" and "advisories" on the page, and "advisories" in the url. Standard stemming isn't necessarily a win for quality, so we took a while and found a way to do it better.
So yes, I think semantics and document/query understanding will be more important in the future. pavlin, I hope that partly answers the second of the two questions that you posted way up near the start of this thread. If not, please ask it again in case I didn't understand it correctly the first time. :)
As i have noticed, I believe personally that the aim of these updates is to get rid of those crappy spam sites, while trying to keep the high quality sites remaining at the top of the serps
[edited by: penfold25 at 6:18 am (utc) on Feb. 16, 2004]
While I am not exacly sure what you mean, I assume that you mean to say that, since mid-November, non-commercial sites were appearing more than they "should" have in the Google serps for search terms that are used by commercial sites, and because commercial sites choose to use those terms that they should automatically be conferred with some sort of priority in the serps.
Comment: My observations about what had been happening with search terms in Google's money search terms SERPs runs completely the opposite to your observations. I posted about that issue several days ago, and I pointed out that information sites about money were "filtered", leaving just the commercial ones, for the most popular money search terms. In any case, isn't it a bit over the top to argue for a preference for the commercial use over non-commercial use of certain terms in serps that no-one pays for anyway. If you want that sort of preference you can go to other payperclick, etc engines, or use adwords.
I can't comment on the broader balance in the serps due to the Brandy changes, however I can do so in my area of activity: the money search terms serps in 64 appear to be much better balanced between commercial and info sites than before Brandy. However the spammers are still there, as I have advised GoogleGuy in a "BrandyUpdate" email. (Keeping fingers crossed on that one!)
Trumble- If you have lost 70% of your Google referrals already, that may not be a problem since the "change" hasn't propagated to enough datacentres to have that sort of effect yet. Have you tried your queries here- http://64.233.161.98
GoogleGuy- Are we able to use the "What are some other things to look out for?" section from that page on our websites?
I noticed that on 64 that major city areas were showing pre-florida style rankings; however, smaller city or town areas were not.
I've noticed this too. It's helpful to learn it's not just me. Let's hope they fix it soon. Nice to see what's on 64 as far as it goes.
Do your results coincide with leading cities being hit the hardest? Put another way, are you fine with 80% of your cities, but wiped out in the majors?
This is the way the pattern started for me, but I think allan may be right about the sequence in which they're fixing it.
I've felt for years, though, that Google has weighted CityName in inbound link text much more than it should, so CityName Widget searches have often been skewed.
On CityName Widget searches, high PR sites of entities that contain CityName as a part of their name and also contain the word "widget" somewhere on their home page will often rank ahead of solid widget sites in local areas.
It could be that that the multiplying factors/filters in the recent updates were also sensitive to CityName anchor text in a way that compounded the problem on big city searches for less than major sites.
It may (also) be that widget inbounds might tend to contain fewer CityName keywords than widget keywords.
I'm also seeing from (just a few) test searches in areas I monitor that Google 64 is particularly sensitive to word order, so searching for cityname widgets gives surprising different results than widgets cityname. Haven't compared this to Florida and Austin, though.
One of my longtime test searches to gauge where Google is at in CityName search is san francisco public relations. (Mods... I hope it's OK to cite this specific search. It's one we've talked about for years). Take a look at the results on 64, and compare them with AllTheWeb (both with and without the phrase quote rewriting), and you'll see how much better ATW handles this particular search than Google does. (At least the results are a lot prettier and more apparently what the search is looking for). At the same time, Google is notably better than ATW in some other searches... for the same reasons, I think, that it suffers on this one.
GoogleGuy wrote ...
<If I want to buy a diamond for someone, I might go on the web and just search for a place to buy a diamond. But a typical user is also going to want to know about their purchase. Things like color, carats, clarity, and so on that people want to find out about. I probably would want to know about the different organizations that certify diamonds, along with some believable opinions about the organizations themselves and their value. >
and ...
<This is all just my personal take of course, but I'd recommend building the sort of resource site that people can use to read and research, the sort of site that people bookmark and return to.>
----------------------------------------------
Does this mean that Google's vision of the future Internet is as a massive online encyclopaedia with little commercial content?
I mean if I just want to buy something and I already know about it I may not want all this information. For example if I want to buy a car I probably don't want to read about stuff like the inner workings of an internal combustion engine. If I did I would search for "internal combustion engine", (which I did as a test and got some great results.)
Conversely I also tried searching for "car" and 6 of the top ten results were car rental sites, which contradicts Google Guy's statement. I mean I did not even add the words hire or rental? This must be Latent Semantics in their broadest sense :-)
My own personal problem as (a one man consultancy) is that I have an authority, original content site that is just about 100% informational but it has been completely dropped for reasons unknown to me. All my communications to G about this have been ignored so far. I was effectively providing the kind of content that Google were seeking before all this started and they pulled the plug on me? A bit of work to be done yet methinks?
To conclude, I have no problem either way, shopping or encyclopaedia. I would be very happy for Google to go down the encyclopaedia route. Just make a statement, let us know clearly that this is your intention and that Google is no longer for people who are shopping and the SEO's and shoppers can move away somewhere else.
I personally like FreshBot and put "News" and "Information" on my pages.....
I have 1 site, using 2 domains, this is because 1 domain is for international use and the language is english, the other domain is the same site in dutch. I don't know if this is considered as spam? The languages are different. And some products are only for dutch use only.
I also have another question. On the site mentioned before, i sell for example widgets. In this thread i read something that Google is becomming more and more like an encyclopedia. If you are searching for the artistA widgets you are looking for a commercial product. If the results are non commercial, these results are illegal, because you have to pay the buma stemra, if you offer those widgets.
I also saw that my competitor is using hidden words and so on and he uses 8-10 domain names all containing almost the same site. He interlinks between his sites (a linkfarm) He is for every search term almost on the #1 position. I don't think this is fair. I'm trying to make a site that is offering products and i don't use this techniques. My site is very clear and almost on the bottom of the page, but the spammy site is in the #1 position. I thought that Google recognizes this kind of spam, well it is very clear that he uses it, but he is still in #1 position. I reported it already as spam a while ago, but still nothing has changed. I know that G will receive a lot of spamreports. Are those spamreports handled manually or not? If so, maybe the report isn't looked for yet. If not, the bot still doesn't recognizes it.
Does this mean that Google's vision of the future Internet is as a massive online encyclopaedia with little commercial content?
I don't see why information and commercial content should be incompatible - just so long as users can find the one they're looking for. I still believe that most users start by looking for content and then go on to compare prices.
It's the old marketing adage: People don't buy products, they buy solutions. Say my problem is to keep my girlfriend happy on Valentine's day. One solution would be an "inexpensive" artificial diamond that looks as good as the more expensive ones and is guaranteed not to fall apart for three months ;-) So give me content to convince me she'll love it then a button to buy.
As far as an online encyclopaedia goes: If I was a diamond retailer I would be extremely happy for my advert to printed alongside the "Diamond" entry in the Encyclopaedia Britannica.
I mean if I just want to buy something and I already know about it I may not want all this information. For example if I want to buy a car I probably don't want to read about stuff like the inner workings of an internal combustion engine. If I did I would search for "internal combustion engine", (which I did as a test and got some great results.)
Yes but its all about what is relevent to your search.
Google will return your site so you can buy a car but its going to give you the best site that has the most information regarding the car or buying a car.
It does take sometime to get your head around the thinking of Google but at the end of the day Google is trying to provide you with search results that don't just contain the keywords you typed but pages that are relevent to the keywords entered.
i.e
property is relevent to apartments
Automobile is relevent to Cars
Sindy is relevent to Barbie
ooops sorry about the last one got carried away :)
Any way you catch my drift, Google also returns plural and none plural words. They know which words are the most popular and cross examin the keywords entered to other words that are also popular and related.
The sites that are high will have a few good related links outbound and inbound, plenty of unique content and above all no spammy techniques.
p.s still waiting for 64.xx.xx to come through here in Europe
You may be extremely happy with this but do you really think it would do you much good? I would think that the Encyclopaedia Brittanica gets about as much traffic as Magellan.
But then again, Encyclopaedia Brittanica (I spelt that again without a dictionary) KNOWS that it is an Encyclopaedia and as such its job is not to sell stuff.
Or maybe a search engine where the article on "what to look for when buying a widget" comes up before 30 identical pages that list nothing but widget product specs...
can the toolbar tell if your site is bookmarked? or if youve returned off a bookmark? Might give weight to Bretts ideas on toolbar usage and serps.
http://www.google.com/webmasters/
I meant guidelines, not rules.
