The Google domination factor (GDF)

[Mikkel deMib Svendsen]

I know I opened up a more general thread on local issues on local stats, but then I thought it may make sense to kick start with this easy task: Drawing a regional map of what i decided to call: The Google Domination Factor (GDF).

In most of Europe (that I know of), and probably other regions too, Google is by far the most dominating index. In fact, so dominating that many companies chose to completely ignore the focus on anything else. At markets where it seems Google deliver +90% of the visitors I understand why comapnies chose to just do Google.

Anyway, I often run into the question of just how dominating Google is at the various European markets. So, I can start out with Denmark - where I live ...


GDF for Denmark: +90%
Websites featuring Google search used by Danes:
Google.com, Google.dk and local partners (major portals): TDCOnline, Eniro and TV2.

[andrewgoodman]

Hi Mikkel,

GDF, or the M!M!M! factor, as Danny would call it :D -- I like it.

Now where are you getting your stats in this case... basically on personal experience with server logs, etc.?

Depending on how one defines it, GDF can't be much lower than 90%, can it? At least in terms of the subset of referrals that are "free search index referrals." Indexes like FAST have never taken off, and even if Yahoo were to pick up steam, it's paid inclusion so results would vary wildly from client to client.

[seobook]

Now where are you getting your stats in this case... basically on personal experience with server logs, etc.?

My sites would likely be biased the way the Alexa system is (toward webmasters and marketers). My numbers are mostly Google but a poor representation of the web as a whole.

While it is fairly non commercial (and thus somewhat biased toward results which are less manipulated) the Wikipedia is also fairly comprehensive. They published referer logs.

http://wikimedia.org/stats/en.wikipedia.org/ref_200405.html

[Dodger]

Indexes like FAST have never taken off, and even if Yahoo were to pick up steam, it's paid inclusion so results would vary wildly from client to client.

FAST is an animal of a different breed. It has supplied other search engines with results for many years, and as far as I can still see...it is supplementing Overture results still. You will never know if you are looking at a FAST result in most cases.

And with all due fairness to Yahoo Search, that paid inclusion (SiteMatch) only represents about 1% of their index and does not give them any preference over organic results (according to Yahoo anyway).

I think with time, Yahoo might give Google a little run for their money and tap into some of these European Markets. In some way they are starting right now, it was not too long ago that they became CNN's search on their main site. Albeit, not my choice to do an Internet search from ... but they are part of the AOL/Time Warner Network. Just some food for thought.

[bwelford]

I rely only on organic SEO and up till recently most search traffic came largely from a variety of Google engines. Yahoo! was rarely there and MSN even rarer.

I am seeing a phenomenon in the last few weeks with Yahoo! referred traffic growing significantly week by week. I think it must be linked to the greater activity of the Yahoo! Slurp spider and faster indexing of web pages. I find the Yahoo! and Google systems are picking up new web pages at much more comparable rates.

Of course the question of dominance still depends on how many searchers are using either SE. I see nothing that will undercut the 3 to 1 or so advantage that G has versus Y! on this.

[AussieWebmaster]

The Asian market is one area where Google does not dominate but they did just buy into Biadu. My take on that is if Sohu or Sina actually act on their statements of wanting to enter the American market, Google simply amps up things at Baidu and Google.cn and they have to scurry home to protect their local turf.

[Mikkel deMib Svendsen]

> Now where are you getting your stats in this case... basically on personal experience with server logs, etc.?

Good question!
In my example, for Denmark, I used the stats I get from a great number of local sites I've worked, and work, with, across many different vertical markets. So, it's a mix of facts and stats and my own feel of it. We probably won't come any closer to the "real" number.

Lets get some more local numbers on the table! :)
Just give your best shot on what you feel is the right GDF for your region - others are welcome to comment the numbers. I will collect all the numbers as we get them in and put them in a table: The Official GDF table :cool:

[Dodger]

I am seeing a phenomenon in the last few weeks with Yahoo! referred traffic growing significantly week by week. I think it must be linked to the greater activity of the Yahoo! Slurp spider and faster indexing of web pages. I find the Yahoo! and Google systems are picking up new web pages at much more comparable rates.

Of course the question of dominance still depends on how many searchers are using either SE. I see nothing that will undercut the 3 to 1 or so advantage that G has versus Y! on this.

Two sites that came online in March are showing differences in indexing at Google and Yahoo. Both are aware of the sites, and have decent amounts of backlinks....but in one case Google is outindexing 3 to 1 over Yahoo, and the other Google is indexing 15 to 1 over Yahoo. On another existing site, which was a past Inktomi inclusion...Yahoo has NOT picked up on two new product lines (outside of the links on the Home Page) and Google has indexed all of the pages. My gut feeling on Yahoo's index ... dirt slow and barely existant.

I have noticed a rise in Yahoo referrals over Google though on the commercial sites. But Google seems to be the popular referral for informational sites. My opinion is that Yahoo is being used for Shopping more than Google, while Google is the go to search for information or research. But I have to admit that I only have a few sites to gauge that by (on the information side) and Google is indexing them quicker. If Yahoo does index these sites more aggressively, that could change.

[Gaia]

About Italy,
Definitely here the GDF factor is real.
From a research we have made last April about Italians use of the search engines:
79,3% of internet users declare Google is their favourite search engine.
And the 53,6% say they always use Google (while the 25,7% very often)!

If you consider that the 2nd search engine is Virgilio (8,8%), and Virgilio results are part from its directory and part from Google, you understand why our clients measure our job by their position on Google.

[NFFC]

I know two things for certain. Firstly Mikkel has forgot more about this stuff than I will ever know, secondly SEO's are lazy.

I don't think that projecting search engine market share on the basis of SEO's logs is a good idea. Google is the easiest SE to use in driving traffic to a site, I believe there is an inbuilt bias towards Google if we take these stats from SEO's.

We were the fortuinate recipiant of a nine month Google ban a year or two ago. I'll tell you this, when you have to find a different source of SE traffic there is a lot of choice out there.

It will vary from sector to sector but we see G at a level somewhat lower that 40%, B2C ecommerce. I think that 40% is about right, still huge but not domination. Personally I can only see that % falling.

Statmarket may have there own bias in stats but my feeling is they are much nearer the truth than a bunch of SEO's ever will be.

http://www.statmarket.com/cgi-bin/sm.cgi?sm&press&week_stat

[Mikkel deMib Svendsen]

I am not sure what your point is, NFFC, but for many regional areas in Europe there are absolutely not a lot of search alternatives. For some markets there are, like the US, but not for all. This just give Google a lower GDF in the US - thats all :)

I would be very happy if anyone could tell me where I can get just 10 or 20% of the traffic in Denmark that we see coming though from the Google index (including all local distribution). I do not think it's there.

I know the same is true for other regions. This thread was simply about trying to establish how dominant Google actually is. If they are less dominant in one region, great - I do not personally like to see them as dominant as they are here in Denmark. I do not think thats good for anyone - maybe not even Google.

[NFFC]

>I am not sure what your point is

That your logs as an SEO do not reflect reality, they reflect the fact that you are an SEO. Currently Google dominates SEO's logs, that doesn't mean they dominate the "real" www.

By "you" I don't mean you personally, just SEO's in general.

Or in one line:

GDF is a myth, it only applies to those too lazy to seek out other traffic sources.

[Dodger]

GDF is a myth, it only applies to those too lazy to seek out other traffic sources.

What country are we talking about there partner? And what if the searcher is too lazy to seek out these alternate sources that you speak of, are you going to teach them how to do that? Or is it just all of a sudden going to appear on there browser screens?

[NFFC]

>And what if the searcher is too lazy

Its not about the searcher, its about *you*.

What I'm saying is that if your Google % is in excess of 40% then you are a lazy SEO.

Let me qualify "lazy" a little. If I hired you as an SEO I would expect great Google rankings, I would judge you on that, after all they are the top dog. You would look to meet my expectations.

I'm not going to hire you, I will do it myself. My ROI differs from yours and so does my Google %.

[Dodger]

Since this thread was about the GDF in Europe (more than the USA) what good will it do to be found in other SE's right now.

But I am not an SEO, so what do I know. I do know what my logs tell me and I posted that here inPost number 8 (http://forums.searchenginewatch.com/forum/showpost.php?p=3105&postcount=8).

I think that it depends on what sector people are talking about when they start to discuss SE referrals. Some are showing an upswing in Yahoo without stating what type of site it is that is showing the upswing. I believe that may have some bearing on things -- and that bearing has to do with the type of searcher using these two search engines and not the website or the SEO.

[Mikkel deMib Svendsen]

> What I'm saying is that if your Google % is in excess of 40% then you are a lazy SEO.

Please, there is no reason to call people lazy just because they do not share your oppinion. Keep to the topic :)

This is not at all the way I have seen things. I have analyzed hundreds of sites over the years - both SEO'ed and not SEO'ed sites and for many regions, like Denmark and Germany I do see a true Google dominance - like it or not. This is not just numbers I am making up but can be confirmed by many local webmasters.

It's not a lot of help to suggest that SEOs look for other stats if there are no valid regional ones. And thats often the case. Using Denmark again as an example we do have public stats on the use of the largest websites, but not how many searches they have and more important: How many click throughs.

When I am talking about the Google Domination Factor I mean how many actual visitors they bring. I don 't care that some portals can show millions of users, or even searches, if only a very small amount click through.

Also, just to be clear, I am not just talking about Google.com but about all the websites regional users use to search that feature the Google index. As many of the biggest local portals (e.g. Kvasir/Eniro in Norway) use Google search Google effectively becomes very dominant indeed.

For the US market you are right. I agree that you should be able to do more than 50% search traffic from non-google engines if you do your SEO well targeting the US. But for many other regions it's not possible, in my experience. I simply can not find those highly searched at websites you are talking about in my region :)

[addvisors]

I optimize sites in many european countries, and find the GDF to be various depending on topic. Eg. when running sites that are primarily non-technical in their keyword mass, the GDF is more or less 90%+, but when the searches become technical and "nerdish" :) even though the searches are in the local language, I find that the GDF gets lower, as Yahoo and MSN gets a bigger share.

Anybody seeing the same pattern?

[Mikkel deMib Svendsen]

Very interesting numbers, Kim. I am sure other regions can report similar variations across vertical markets and target groups. So far we've been focused on the general GDF in this thread but it's even more interesting if we can dig out some more details.

[andrewgoodman]

Slightly off topic as they are referring to US numbers, but I do find the WebSideStory numbers from the March 30 2004 press release an important bellwether because they do refer to real server log stats from thousands of websites:

search referrals:

41% - Google
27.4% - Yahoo
19.6% - MSN

What is left unstated here, though, is the proportion of Yahoo and MSN traffic that is coming from paid inclusion. In MSN's case we have seen a mishmash of inclusion methods with the rising and falling fortunes of Ink & LOOK, and now we wait for what seems like forever for MSN to release their own search product. If the number of referrals can go way up depending on one's paid inclusion spend (remember they are not only paid inclusion now, but paid inclusion with a PPC chaser), then we aren't comparing apples with apples.

Until MSN and Yahoo renounce paid inclusion we are going to be faced with some difficult challenges in ascertaining what is a fair "GDF" number, as some marketers will spend heavily on paid inclusion, which Google does not offer.

[Dodger]

I optimize sites in many european countries, and find the GDF to be various depending on topic. Eg. when running sites that are primarily non-technical in their keyword mass, the GDF is more or less 90%+, but when the searches become technical and "nerdish" :) even though the searches are in the local language, I find that the GDF gets lower, as Yahoo and MSN gets a bigger share.

Anybody seeing the same pattern?

I am seeing the opposite on US based searches. Hmm...that is interesting, I would of thought that the high incedent of Google referrals would have been the same in Europe. It seems that there are also a large amount of techie type programmers in the OpenSource area coming out of Denmark and Germany in particular ... I kind of assumed that is all that you guys used the Internet for over there. ;)

[AussieWebmaster]

Slightly off topic as they are referring to US numbers, but I do find the WebSideStory numbers from the March 30 2004 press release an important bellwether because they do refer to real server log stats from thousands of websites:

search referrals:

41% - Google
27.4% - Yahoo
19.6% - MSN

What is left unstated here, though, is the proportion of Yahoo and MSN traffic that is coming from paid inclusion. In MSN's case we have seen a mishmash of inclusion methods with the rising and falling fortunes of Ink & LOOK, and now we wait for what seems like forever for MSN to release their own search product. If the number of referrals can go way up depending on one's paid inclusion spend (remember they are not only paid inclusion now, but paid inclusion with a PPC chaser), then we aren't comparing apples with apples.

Until MSN and Yahoo renounce paid inclusion we are going to be faced with some difficult challenges in ascertaining what is a fair "GDF" number, as some marketers will spend heavily on paid inclusion, which Google does not offer.


I agree with you Andrew.. the impact of paid inclusion is not factored into search many times... though attempting to isolate by that method would have to isolate the numerous submissions to Yahoo directory to get Google attention

[addvisors]

My guess is that techies use more than one engine to find the best results.

I especially see it on a few sites offering games etc.

[bwelford]

It's funny that metacrawlers such as Vivisimo or Metacrawler have never seemed to attract as much interest as you would expect, given that a Google or a Yahoo! can have brief periods of apparent confusion or just plain dumbness. :confused:

[seobook]

I agree with you Andrew.. the impact of paid inclusion is not factored into search many times... though attempting to isolate by that method would have to isolate the numerous submissions to Yahoo directory to get Google attention

I might be crazy, but I do not think paid inclusion usually has much of a real effect on most sites I do SEO for. I really don't care if the other guy is using paid inclusion if I get 3 times as many links and outrank him even while I am sleeping. Granted in some industries where you need to list well for different buzz words you may need paid inclusion, but as far as general SEO goes I totally ignore the effect of paid inclusion on most sites I work on.

[seobook]

It's funny that metacrawlers such as Vivisimo or Metacrawler have never seemed to attract as much interest as you would expect, given that a Google or a Yahoo! can have brief periods of apparent confusion or just plain dumbness. :confused:

the problem is that there are only a few search engines for those meta search engines to weigh in, so they to acquire a good bit of that dumbness.

[wiseMouse]

the problem is that there are only a few search engines for those meta search engines to weigh in, so they to acquire a good bit of that dumbness.

Not only this, but by covering more of the web, meta search is not more accurate. In fact, without knowing how the individual indexes create their ranking you have quite some trouble to make a more relevant list of results.

More to the topic, in Switzerland, I read somewhere, Google feels pretty hard since the surfers there use one or two local search engines, established long before Google existed...

Generally, I think if we consider a Google Factor, we should go with the smaller number (41%) since they're based on larger numbers of sites (i.e. a wider variety of surfers).

[Andy AtkinsKruger]

Looking at tracking stats on many sites, Google dominates the UK with around 66% of visitors.

MSN and Yahoo fluctuate between 10-15% each depending on the organisation.

Ask Jeeves is about 6% - but occasionally has bursts of activity on some sites and has beaten MSN and Yahoo on occasions.


Mikkel's initiative to get a guide to the level of activity in local search is a good one and I don't agree with NFFC that SEOs are lazy and can't make judgements on where the visitors are coming from. Firstly, we can see the visitor levels change on sites as new sites are indexed by other search engines - and we can compare rank positions. Added to that, we can use Overture and Wordtracker to compare the relative importance of phrases and therefore spot which ones are missing from the search engines.

User research is expensive and slow to get a degree of accuracy only slightly better - as it doesn't show people who switch from search engine to search engine.

Andy

[Mikkel deMib Svendsen]

> More to the topic, in Switzerland, I read somewhere, Google feels pretty hard since the surfers there use one or two local search engines, established long before Google existed...

I am in no way an expert in Switzerland, but as I understood it there is not really any proprietary search technology just for Switzerland - the portals do the same as they do in the rest of the world: They license it from either Google og Yahoo!Search.

The way I calculate GDF I include all Google distribution - basically, all sites Switz users use to search Google. For most regions this usually include Google.com, Google.local + one or more distribution partners

[kenpomachine]

In Spain, and acording to Netratings, Google get 90% of the searches. In my experience, the percentage is a bit higher for a big commercial site (with around 400,000 visitors a month).
Besides, in the last year or so, MSN has lost some of its autosearch market, as more and more people go directly to Google now. This trend comes from my website's stats, but I think they apply to the general market as well, as people are becoming more internet savvy.

[AussieWebmaster]

It is hard to deny Google dominates... I see the same thing too... in Spanish, German, Chinese and Japanese even... the next one we are rolling out is Arabic so let's see if they are the ones there also.

[Patrick Berry]

To add my little grain of sand to the mountain:

a client site runs multiple ppc´s and dynamic banner campaigns in 6 european countries. Also tv campaigns, press campaigns wct . This is an official tourist informacion site, and has a huge number of resources.

The GDF is high. between 65% and 85% of search engine generated traffic.

Oddly SAPO dominates the traffic generated by SE´s in Portugal but it gets its results form Google. I(yes it does have a directory but little traffic)

[AussieWebmaster]

To add my little grain of sand to the mountain:

a client site runs multiple ppc�s and dynamic banner campaigns in 6 european countries. Also tv campaigns, press campaigns wct . This is an official tourist informacion site, and has a huge number of resources.

The GDF is high. between 65% and 85% of search engine generated traffic.

Oddly SAPO dominates the traffic generated by SE�s in Portugal but it gets its results form Google. I(yes it does have a directory but little traffic)
And the results are the English based SERPs... interesting.

[andrewgoodman]

It's funny that metacrawlers such as Vivisimo or Metacrawler have never seemed to attract as much interest as you would expect, given that a Google or a Yahoo! can have brief periods of apparent confusion or just plain dumbness. :confused:
When there really were 6-7 different indexes with different crawlers, different indexes, and markedly different algo's, PLUS at least a couple or three major unpaid directories, metasearch was cool. The underlying reality of metasearch now is it's a lot like a Google search with more paid listings. In Vivisimo's case, a lot of the underlying results are from companies like LookSmart, Lycos, and MSN. If the user trusts Google as a search brand more than they know or trust Vivisimo or any of those underlying players, they won't exactly be falling over themselves to use it, I imagine. In the old days, the much smaller users base for search really looked at these tools with some reverence, but the marketplace has become much more sophisticated and suspicious. They know about index size and paid listings and so forth. Google has survived intense scrutiny, whereas companies like LookSmart have often buckled under pressure. Users seeing a page of metasearch results that feature LookSmart results would probably be sceptical.

Of course, lots of emerging technologies, such as Google Scholar and various shopping engines, are really kinds of metasearch, aren't they? Actually I think there is considerable overlap between the idea of metasearch and any kind of info aggregation. If Froogle imports user reviews from Bizrate, or if Yahoo Local aggregates four or five or twenty types of info on a page, that's helpful to the user, just as the idea of Dogpile concatenating search results from different engines once seemed cool to people (but it's getting tired as a concept).

In other news, Mamma just bought Copernic (the enterprise search part of the co. was recently spun off and renamed) and hardly anyone noticed. Metasearch definitely is not capturing users' attention these days.

[seobook]

>They know about index size and paid listings and so forth. Google has survived intense scrutiny, whereas companies like LookSmart have often buckled under pressure. Users seeing a page of metasearch results that feature LookSmart results would probably be sceptical.

I think most people are rather unaware of much of the underlying business models behind search and concepts like paid inclusion.

I think people just trust googe because:
-they know of it
-they have been told that its the best
-it is fast and easy to use