|
Percentage of U.S. Searches Among Leading Search Engine Providers |
|||
|
Oct.-07 |
Sept.-07 |
Oct.-06 |
|
|
www.google.com |
64.49% |
63.55% |
60.94% |
|
search.yahoo.com |
21.65% |
22.55% |
22.34% |
|
search.msn.com |
7.42%* |
7.83%* |
10.72%* |
|
www.ask.com |
4.76% |
4.32% |
4.34% |
|
Note: Data is based on four week rolling periods (ending Oct. 27, 2007, Sept. 29, 2007, Oct. 28, 2006) from the Hitwise sample of 10 million US Internet users. * - includes executed searches on Live.com and MSN Search. |
|||
|
Source: Hitwise |
|||
don't come close to their own experiences with search traffic. Since I get into a lot of conversations around Google's dominant market share and see plenty of raised eyebrows from savvy technologists, bloggers, and marketers, I figured I'd ping my good friend, Matt Tatham at Hitwise, to see if he could help solve the mystery, which of course, he did (thanks Matt!).
Take a look at the chart below - it shows the percentage of upstream traffic (that's all traffic - search engines, direct type-in, referral links, email, etc.) to technology and tech media websites. Hitwise places these in two separate categories, "IT and Internet" and "IT Media." A site like Sourceforge would probably fit in the former, while SEOmoz would be in the latter.
The graph tells us that Google sends 15.73% of all traffic to all websites. That alone is an interesting figure, as it tells us something about how we surf the web - yes, search is certainly a huge driver of traffic, but it may not be the juggernaut we sometimes think it is. 75%+ of web traffic comes from other sources: following links, typing in domains, email, bookmarks, blogs, and social media.
More relevant to our discussion, however, is the data showing that 31.95% of all traffic to "IT Media" sites comes from Google and 5.78% of all traffic to "IT and Internet" sites has Google as the referrer. Let's compare this data against Yahoo! and MSN/Live to see whether and how much Google is over-represented in the online tech world.
Yahoo! sends 6.92% of all traffic to "IT Media" sites, 1.90% to "IT & Internet" sites, and 4.43% of all website referral traffic.
Matt's taken the time to break out Live.com vs. MSN.com, which gives us a peek into the distribution between the two (Live's still much smaller). But I've combined the data for easy comparison, and between the two, Microsoft web search products send 1.61% of traffic to all websites, 2.1% of traffic to "IT Media," and 0.65% to Internet & Technology sites. Taken together, this data gives a chart that looks something like:
Takeaways:
- If you're in IT Media, there's a good reason you're seeing 75-85% of your search traffic come from Google. Don't panic - you're not being "under-represented" in Yahoo! & MSN, that's just the way it is.
- If you look at the Internet & Technology column, you might be surprised to see that Google's only at around 70%, but my guess is that Yahoo! is actually listed in Hitwise as an "Internet & Technology" website, and they do a very good job of sending a lot of traffic back into the Yahoo! network. Still, 70% is bigger than the 64% market share number, and my guess is it's even larger than that.
- Google sends 72% of traffic to websites compared to Yahoo!'s 20% and MS' 7.4% - these numbers don't quite match up to the search shares, but that's to be expected. This data tells me that Google is slightly more successful in sending actual traffic to their search results than Yahoo! and Microsoft - perhaps there are more "frustrated" searchers on the other two engines, or perhaps they are both "retaining" more of their searchers inside their own content networks.
I get 85% of my traffic from Google - what am I doing wrong in Yahoo! and MSN/Live?It could likely be that you're doing nothing wrong at all, and Google simply dominates the upstream search traffic in your niche. For those who are curious, here's SEOmoz's breakdown of search referral traffic:
Hitwise doesn't specifically monitor the webmaster/Internet marketing niche, but I suspect the numbers are even more Google-centric, as webmasters tend to be extremely biased in Google's favor (at least, from a usage perspective).
BTW - I'm not 100% sold on the accuracy of Hitwise's data. I think that relatively speaking, it's usually solid, but I wouldn't go out and swear in court that the numbers are precise. I usually take them with a +/- 25% swing, but it's certainly better than nothing (and considerably better than Quantcast, Alexa, or Compete).
Rand - I think that your point about Hitwise's data is spot on; useful but not gospel. I like the guys there a lot (hi Heather!) but have heard some worrying things about data they don't take into account.
Anyway, in the spirit of openness, here's out referral stats since we launched the new site:
And of that:
Make of that what you will!
To be fair though, we do see a much larger market dominance from Google in the UK than they do across the pond.
Manley's search referrer stereotyping rule of thumb:
Whilst professional men* represent the largest market share for e-retailers, cost per conversion on the other engines can often be considerably less than it is at Google.Even for those of us with themes or geographical locations which mean we get almost all of our traffic through Google, there is a knock on ROI from ranking in other engines, particularly Yahoo!, which is inbound links from their users boosting ranking as in the Google SERPs.
*I am not suggesting that there is an amateur class of man.
I think that, in Latin America, academic audience uses Google.
Haha. I am definitely in the amateur class of man.
No wait, I'm supposed to be a robot. Damn.
Give your "stereotyping rule of thumb" (like it!), I'd love to break down the search queries people have used when coming from each engine and compare the quality of the search terms. Is there a noticeable difference in quality or style and are the differences noticeable across a large number of different sites?
Did I just create a new project for myself?
I've seen this change based on niche. My English Grammar site gets excellent term vs content matches in MSN/Live, ok on Google and not so much on Yahoo.
My SEO site gets very good matches from Google and usually Yahoo, though sometimes Yahoo seems to be streching things... I can see why 14thC would rank for some terms even though they clearly don't fit the theme (right words, wrong box kinda thing). MSN/Live sends so little traffic it's hard to tell.
My business/leadership site seems to be doing fine in Google and the others are catching up so no judgement there yet.
When it comes to finance I suspect this an area Google dominates. Over 90% of the search traffic is from Google. The matches are good to very good. The matching quality against Yahoo is excellent, just not enough of it - is there ever enough?!
My friend's metal fabrication site matches the overall averages Rand listed above and the keywords are excellent matches for the content in Google and Yahoo. MSN sends so little traffic it's hard to tell yet.
I suspect the index size makes a big difference here, especially if the index is scewwed toward certain industries. That can throw off the balance of what words mean in other industries.
A small niche site of mine used to get some very strange traffic from Google. Yahoo and Live were way more accurate. Google seemed to pull words and phrases from all over the site, stir them up, misunderstand how they were used and send in tons of users. Who all promptly clicked "Back." Things are a lot better now; this was about six months to a year ago.
Y! and Live sent far less traffic but people usually stayed on the site for longer.
I have been perusing logs and it does seem that there is a fairly even spread of search query idiocy.
I'm seeing a lot of typos from Yahoo! users and a lot of old school Ask Jeeves style 'questions' from MSN and Live, but the grepping I have been doing is on too small a selection of sites and over too short a time frame for me to draw any real conclusions (it's well past midnight in the UK now and apathy is kicking in).
I believe that a search engine, like any tool, is something which one becomes more proficient at using with practise and it is likely that, given enough data, I will find a fairly even spread across the board.
Even a hardened professional like myself has made appalling blunders when searching before now.
I used to see a lot more of that too. Especially in the waterwell industry. That's where I realized how important a holistic approach to web design and SEO is. I'd say about a year ago I saw it clearing up and see it much less on Google and more on Yahoo!
Go figure!
Rand's post reminded me that I needed to do another full discoluse post on the stats which is here.
Are you planning another annual round-up Rand?
Hey Rand,
Can you give us any color on the difference betweek "IT Media" and "IT and Internet"? Why would Sourceforge fit into the former and SEOmoz the latter? Having the definition may be key to understanding this.
On a common sense level, these data make sense for IT professionals. They are more likely to install software (alternative browsers, toolbars) and less likely to rely on general interest portals. Both of these factors point to higher Google usage.
My recollection is that the data you provided is very representative of the referrer mix we experienced at internet.com.
Ezer
The most interesting point to me (and the reason I am so into gadgets) is where you said...
"The graph tells us that Google sends 15.73% of all traffic to all websites. That alone is an interesting figure, as it tells us something about how we surf the web - yes, search is certainly a huge driver of traffic, but it may not be the juggernaut we sometimes think it is."
The role of search engines has been going down for awhile as far as being a source of traffic for the overall web. It is because of that trend that search engines spend so much time and resources on alternate products.
Which is also why Search Marketers should be thinking of themselves as Internet Marketers - driving traffic (and conversions) to our clients from all sources - not just search engines.
Man, I'd probably get really depressed if I realized search from social sites are driving more traffic than traditional SEs. And the relevance is pretty good too...
See this post from the (where else) Hitwise blog on how Topshop gets more traffic from networks than search.
No I'm saying I do - I'd just be depressed if I actually realized it!
I just started doing work for an ag related site this year. This thread made me curious so I had to check. Overall numbers aren't spectacular compared to ther niches, but my search engine distribution is a bit different so I thought I'd share.
Search engine traffic breaks down to...
So a bit less google, a bit more Yahoo/MSN/Live than others might be seeing. I think it breaks down to the level of web-savviness of the folks in our niche. Many farmers are either a bit older, a bit less computer savvy, or both compared to other niches. Many search with whatever their homepage happens to default to. That said, I think those numbers will change as computers start to play a larger role with farms and ag-business.
Plus www.farmerneedsawife.com is going to have an effect evenutally :)
I used to work on the website of Farmers Weekly (UK ag magazine) and the users were scarily savvy. It had the busiest forum of any of my old employer's sites, which is impressive when you consider that included a number of tech titles.
Ah - farmers DO like to talk to other farmers (even if it's via forum!). Seriously though, they do grasp things pretty quickly once they realize it is a tool that can make them more productive.
Farmers are not as dumb as TV makes them out to be. There is a lot of science and machinery involved in farming.
Absolutely - bizarrely the thing that made the forum was crises like the foot & mouth epidemic when they couldn't leave their farms..
In Latin America, Google has more than 80% of share in search. This information was provided on the SMX seminar on October in Buenos Aires.
From my sites, google share is more than 90% as follow:
Yoga site: 96%
Real Estate: 97%
IT consulting services: 98%
Tourism: 97%
Art: 97%
These results include image searches, are based on unique visitors and most of them come from google.com.ar.
I am conducting a survey on my blog of Marketing (Marketing Club of a university), and so far (with 53 votes, which is not too much) google has 98% of share. There could be a bias since is hosted on blogspot, but I suspect that most of the votes came from members of the club instead of visitors that arrived by searching.
Ono thing I would like to see is the trends of search engine, referral and direct traffic. I'm going to have a deep SEO work within a major ISP/Portal here in Brazil and I'm going nuts to define a % to achieve.
The stats for Google, Yahoo, MSN from Hitwise are fine. I get 70-80% of all traffic from Google and for my clients as well (IT and non IT). More of the sophisticated users use Google.
Yahoo SEO is different from Google. Yahoo requires more keyword stuffing for relevance and over time things start to rank in Yahoo vs. Google rankings that perhaps are more straightforward and logical.
If I optimize my site for Yahoo rankings, then Google starts to count this as SPAM. I would love to hear examples of where one can optimize for Google and Yahoo succesfully, as this remains a challenge.
I don't think Google share of the market alone is the influence here. For evidence, just look at what keywords are sending you traffic from Google and where those keywords rank in Yahoo. Chances are they just dont rank in Yahoo - IT or non IT fields.
People tend to have a bias for certain search engines - MSN, Yahoo or Google. The keywords people search across them are relatively the same - Laws of statistics.
Would love to hear answers on how others have been able to achieve both Yahoo and Google rankings for keywords.
If we take a look on a first trend we will see thet ASK loosing searchers. It is because they stopped their TV advertising om major TV channels.
Yahoo picked up in searchers simply because during Holidays more people read their news, e-mail each other, look for movies, use modile web etc...
The same we can say about MSN, they reached 10% only because Christmas time.
Thanks for this Rand.
I've always been concerned about such a difference in my referral percentages between all the SE's. I guess now I can put it on the back burner and worry about one of the many other SEO related problems I have!
I dont think it has to be all about visitors /
- on one domain I get 2x as many visits from google, but 2x as many conversions from yahoo
- on the other domain I qualify for #1's question,
my site on pallet jacks, is leaning toward 65% from MSN and 20% Google, 15% Y! ....
numbers could be skewed as I get less than 10k uniques/mo
Those data relate closely to one of my non-tech sites:
67.92% Google
17.21% Yahoo!
9.5% MSN/Live
5.37% Others
I'd be very interested to know if there are any niches in any countries which go against this trend. Can anyone here share any stats that do not match the hitwise findings?
Is it possible that only www.google-haters.com, www.yahoo-is-the-only-one-for-me.com and www.whaddayameannooneusesaltavistaanymore.com might have a different balance?
In Portugal Google rules supreme, with over 90% market share. And I think that is the global European picture as well.
Check out the Global Search Report from Multi Lingual search for more details.
Thanks for that excellent link carfeu. By my reckoning that report shows that in 11 out of the 16 countries google has the largest share. Less than I was expecting. Perhaps my perspective was skewed because I rarely think of search beyond the UK and US. Shame on me.
@rand/ciaran:
why do you guys say that you don't necessarily trust these numbers? And where do you come up with that +/-25%?
I'd like to see more information about these kinds of stats:
* where do they come from (how are they measured and how does the measurement mechanism show/hide trends which might be relevant to my site)
* how does my site fit into the different segments?
* how do I interpret the differences between search share and traffic share (or do I even need to worry about it)?
Basically:
* how is this information actionable?
Also:
* do Yahoo and MSN keep traffic in their own content properties more than Google?
* to what degree is this true? (if it is)
* if so, what does this say about my relationships with these entities?
* should I be spending time building content or getting links from their properties (not necessarily for link juice, but for referral traffic)?
I can answer a bit of that. Consider who would use stats - tech savy webmasters, IT pros and SEOs. Most of my clients never look at their stats and almost none had a stats program before I came along and implemented one.
Then of the sites that use stats, how many use Hitwise? Which categories do they lean toward and what user behaviours would be effected by that?
Hitwise also uses ISP data and the regions those ISPs are in can effect the data a great deal based on what is trendy or taboo in that town.
See https://www.hitwise.com/products-services/how-we-do-it.php for how Hitwise collects data.
Thanks for the link, I'll check it out.
Should have read this before posting the link myself!
Basically because of this https://hitwise.co.uk/products-services/how-we-do-it.php
It's not that I don't trust them, just that basing concrete assumptions on a sample size of 25 million globally is quite a big leap.
As Rand says though, it still seems a lot better than other competing tools.
This my new blog
https://www.igorthetroll.com/blog/
It is not about SEO.
First, I like to thank Rand Fishkin for his inspiration to do this. If it was not for him I would have never started this project. I also like to thank all the Mozzers who have tolerated my Rants during this year.
I kindly and humbly invite you to come and take a look what it is about.
Thank you,
Igor Berger