Hitwise Intelligence - Bill Tancer - North America

I wonder how this correlates with 'real' life . Do we spend 4% of our time searching say for socks or other items.

Posted by Geoff | May 19, 2006 02:33 AM

Hi Bill,

First time I've seen something like this, and it shows just how incredibly dominant search is to the rest of their products and services.

Your excellent post got me thinking:

Does Google suffer from Nimoy Syndrome?

What do you think?
~G~

Posted by George Nimeh | May 19, 2006 08:16 AM

Is image search gaining in share? I don't remember it ever being that high before...

Posted by J-Weezy | May 19, 2006 09:15 AM

J,

Image search has grown in overall market share, increasing from .32% of all Internet visits in December 04 to .48% for the week ending last Saturday. - Bill

Posted by Bill Tancer | May 19, 2006 09:26 AM

Hi Bill,

How do you define an "internet visit?" If I do a Google search once per hour over a 10 hour business day, is that 10 visits? How about it I do a search every 6 minutes for an hour? Is that 10 visits, too? If I keep a Gmail tab open on Firefox all day long without logging out, but and check it once every 30 minutes or so for new email, how many visits is that? Just trying to better understand what we're measuring here.

Thanks for the chart!

TB

Posted by T-Bone | May 19, 2006 10:01 AM

Two other questions, Bill. First, are these data for the the entire Internet, or just for the USA Internet? (I suppose USA -- otherwise I suspect google.co.uk and google.de would be very high on the list.)

Second, have you published similar data for MSN and Yahoo and AOL? Now THAT would be an interesting comparison.

Best regards,

TB

Posted by T-Bone | May 19, 2006 10:15 AM

That chart is interesting, but not that surprising. I'd be much more interested in seeing a chart comparing each of those categories to the equivalent Yahoo category. That would be more revealing of the success of a particular service. Of course image search is going to be a fraction of search since it's inherently a specialized search.

I'm curious how you get these numbers. Some of these services are used differently than others so I would expect that to skew the results. I would imagine a typical google search would result in maybe 3-5 page views. A map on the other hand-- you could look at for hours without ever reloading the page. Do your numbers account for the one time the map page is requested, or does it count each map tile request? If the latter I would expect to see a higher percentage.

Posted by Tim | May 19, 2006 10:16 AM

Thanks for all the comments, i'll try and get to everyone's questions. TB, Hitwise uses the IAB definition of a visit (considered the standard by most) which is defined as: "a series of page requests with a gap of no more than 30 minutes between each one." All data on my blog (unless otherwise specified) is coming from our U.S. sample of 10 Million internet users (all residing within the U.S., we also have samples for the UK, Australia, New Zealand, Hong Kong and Singapore.

I've received several requests to post Yahoo and MSN data which i will do later today.

Regards,

Bill

Posted by Bill Tancer | May 19, 2006 10:37 AM

Tim,

To your comment, see my response above. Using the IAB definition of a visit, a series of 3-5 pages viewed would still be counted as one visit (as long as there wasn't 30 minutes of inactivity between pages).

-Bill

Posted by Bill Tancer | May 19, 2006 10:55 AM

If I'm reading this correctly, Google Maps has 0.82% of the Online Mapping Market (against people like Map Quest) and Google Video Search only has 0.45% of the Online Video Market (against people like YouTube, etc)

Those seem awfuly low...

Posted by Rob Davy | May 19, 2006 11:17 AM

Rob,

Actually those percentages are of all visits to google domains (so the maps % is visits to maps.google.com as the numerator and total visits to all 20 google domains is the denominator). If I look at the Hitwise Travel - Maps category (which has 164 sites) Google is #3 w/ 7.52% of visits to the category. Hope that clarifies.

-Bill

Posted by Bill Tancer | May 19, 2006 11:32 AM

That's what I thought to start, but it was this bit that made me think otherwise:

"(Google Finance currently ranks 39th in the Business & Finance - Business Information category, with 0.29% market share of visits to the category, Yahoo! Finance in contrast, ranks #1 in the same category with 35.6% of market share)"

Does that mean 0.29% of Google visitors are at Google for Google Finance and 35.6% of Yahoo visitors are at Yahoo for Yahoo Finance?

Posted by Rob | May 19, 2006 11:36 AM

The popularity also seems to be dependent on how prominently these sub-domains are displayed.

I just noticed that the six apps mentioned on Google home page (web, images, groups, news, froogle, maps) figure in the top 8 in list shown above.

Posted by AnotherOne | May 19, 2006 11:37 AM

Rob,

Sorry for the confusion, the stat on Yahoo! Finance and Google Finance in the text of the posting are the % marketshare stats for the two properties within our Business and Finance - Business Information category.

-Bill

Posted by Bill Tancer | May 19, 2006 12:00 PM

So the stats in the text are actual marketshare (eg, how much of the online Finance information market Google and Yahoo each have)?

And it's just a coincedence that Google has 0.29% of that market and represents 0.3% of all traffic to Google properties?

Describing it as Market Share in the table is a little mis-leading, when it's more "Google Traffic Destination Share"

Posted by Rob Davy | May 19, 2006 12:05 PM

Rob,

Thats correct, Google represents .29% to category and .03% of visits to top 20 Google domains . I'll be more specific in my chart labels going forward.

-Bill

Posted by Bill Tancer | May 19, 2006 12:15 PM

Bill,
I found your analysis very thought provoking. In fact, I'd been thinking about doing something similar for MapQuest (also mentioned above) to see how they have fared against the onslought of the Search Engines. Since Accoona is a client of yours here is what I've written, based on Hitwise data. If you have any emlightening comments they would be most welcome. As I condlude, the results seems rather counter-intuitive:

I have been wondering about MapQuest these days. It seemed to me that they must be suffering from the introduction of Yahoo Maps, Google Maps, and MSN�s Virtual Earth. I�ve never been a fan of MapQuest maps, they�ve always seemed clunky and hard to manipulate to me, and the unstructured, natural language map queries that the likes of Google permit seem much superior to MapQuest�s old fashioned form-based queries. In addition, Google seems to have brought to earth high flyers that at one time dominated their markets. So it was a big surprise to me to find out that MapQuest has not only held its own from the onslaught of the Search Engines but has even thrived. Here is a Hitwise graph of the monthly all sites market share in the category of �Travel � Maps�:

(Unable to display Hitwise graph here)

As you can see, after falling about 10 points below Yahoo Maps for about 8 months, MapQuest suddenly crossed over in January 2006 to completely dominate the segment with a 56% market share, while Yahoo Maps fell to a 20% market share, a one month drop of 15 percentage points for Yahoo and a 26% one month gain for MapQuest! Google was never even a contender for the top spot and dropped 3 percentage points the same month that MapQuest swung to dominance. I have no idea what occurred in that month to create such a sea change. It almost makes you think there was something wrong with the data, except the positions have held for three months now.

Some interesting conclusions to draw: A better user interface both for the display of maps and input of geographic queries does not seem to have made a big difference in causing users to switch their allegiance. Of course, it may be that unstructured geographic queries are unsettling to a lot of people, but I happen to prefer them! It also seems that the option to overlay satellite imagery and even 3-d visualizations don�t matter a hoot to most people. While cute and fancy looking, they seem to appeal mostly to the technorati, who probably don�t use them much anyway, and not to the general user. Perhaps momentum is not as easy to change as thought. If most people are used to going to MapQuest, they�ll keep going there as long as they keep doing a decent, if not the best, job. Of course, this doesn�t seem to explain why MapQuest suddenly gained 26 percentage points in a single month. Those people weren�t using MapQuest the previous month.

I confess this all seems rather counter-intuitive to me.

/Ira

Posted by Ira Machefsky | May 19, 2006 03:43 PM

Interesting, but I would suggest not really "the top 20 Google domains." Those would include blogger.com, orkut.com, blogspot.com, and perhaps some others. This is only those that are *.google.com.

Would be interesting to get the others in there.

Posted by Evan Williams | May 19, 2006 03:53 PM

Ira,

Thanks for your post. I'll try and chart/blog this over the weekend, look for my analysis and thanks for a thought-provoking comment.

-Bill

Posted by Bill Tancer | May 19, 2006 04:01 PM

Evan,

You are correct, I only included Google branded sites within this analysis. FYI, if I include the three domains you mentioned, Blogger would appear as #6 in the category @ .55% share, Orkut would show at #10 with .31%, Blogspot resolves to Blogger so would be counted in that marketshare.

-Bill

Posted by Bill Tancer | May 19, 2006 04:14 PM

Thanks!

However, just to clarify, is the blogger/blogspot traffic actually just == blogger.com or is it combined? The reason is, blogspot.com resolves to blogger.com. However, there are lots of *.blogspot.com's. I'm guessing those you're not counting?

Posted by Evan Williams | May 19, 2006 04:29 PM

Evan,

We aggregate visits to all *.blogspot.com sites to blogger.com.

-Bill

Posted by Bill Tancer | May 19, 2006 04:36 PM

There is also a chance that numbers could be skewed due to double-visits due to book-marking reasons. i.e., users may visit their Google Personalized Homepage for the sheer purpose of checking their Gmail (using the module.) If they have new mail, they would click on the link. This gives Google's home page a hit AND Gmail a hit, even though the hit was primarily for the purpose of Gmail and not Google.

Furthermore, individuals could very well have Google's home page set as their default home page in their browser settings. It is a very fast-loading page, which is perfect for this reason. I may wish to visit Vonage.com but would start by opening the browser (which gives a "visit" for Google,) then enter Vonage's web address. I could even perform this step if I were really wanting to visit Google Calendar. All of this activity causes skewed statistics, that would artificially inflate the value for Google's home page. This, in turn, would increase the percentage for Google's home page and would, likewise, decrease the percentage for all of Google's other properties.

Also, the only reason anyone would visit Google Talk's site is to see if there have been updates, or to read about or download Google Talk for the first time. So, if 1 out of every 1000 people who are performing searched on Google are visiting Google Talk's download page, that actually says a lot of positive things about Google Talk, though the chart makes Google Talk much less popular.

Lastly, this comparison chart would tell a whole lot more if it were compared to similar charts for other jack-of-all-trades companies like Yahoo or Microsoft. In other words, if visits to Yahoo Maps happens to constitute 0.82% of all visits to Yahoo services, then this value only respresents the need for mapping services compared to the need of general search with the public as a whole, not necessarily the success or failure of a given service.

Put another way, if 1% of a grocery store's sales is Tampons and 20% of a grocery store's sales is bread, this does not say that the grocery store is doing a terrible job at marketing, promoting, or transacting tampons when compared to bread. It just shows the differences in trends between those two types of products in general.

That's just my 17 cents.

Posted by Bob Bigellow | May 19, 2006 05:29 PM

It is sad to see Google News so Low on the list...

This is a classic example of NOT exploiting the full potential of an asset....

If the news was somehow more intelligently intergrated with the SERPs, it could lead to massive traffic....

Posted by Search Engines WEB | May 19, 2006 06:00 PM

An important flaw (irrespective of raw data accuracy) is that some URL's are "product" sites (visit for each use), some are "download" sites (desktop search, toolbar and earth) where one visit leads to actual product use unseen by Hitwise, and others are in-between AJAX products (Mail, Maps), where one access equals one broduct invocation, but the amount of email reading, mapping and searching done (panning, new locations, routes, etc.) is hidden behind the one base URL accesses.

Contrast this to other products (as shown in recent Hitwise vendor comparisons) where some companies have AJAX or Flash-based apps in the same market space as an older, HTML site. For mapping, this means that the first group shows one access per user session while the second group shows an access on every attempt to move east, closer, or search. This is at least one reason for the results showing as they do.

These URL-based competitive analysis are becoming ever weaker just as those who report them (The Guardian, for example) become ever less-aware of the meaning of the data. How unfortunate.

Posted by Michael Jones | May 20, 2006 08:56 AM

Would be nice to see Google News ranked #2.

Posted by Google Big Fan | May 20, 2006 10:35 AM

Hi Bill,

I'd be interested in seeing how the fairly new Google Trends will rank in the next analysis you do.

http://www.google.com/press/pressrel/new_tech.html

Thanks,
S

Posted by Seena | May 21, 2006 03:04 PM

Seena,

Will do, I should have data on Google Trends within the next few days.

-Bill

Posted by Bill Tancer | May 21, 2006 04:20 PM

I thought answers and google talk should get more traffic..

- mike

Posted by Find in Forums | June 6, 2006 02:13 PM

This is incorrect use of terminology. This is Product portfolio share and NOT market share. The former means the share of the various product WITHIN ones own product mix. While the later means each product's share in their respective category.

Please correct it, its completely misleading.

Two questions. What do the %'s mean and how was this arrived at? And second what's the source?
[By number of visits? Or number of page views? duration? by domains/URLs?]

Posted by Ajit Narayan | June 8, 2006 11:53 PM

Google serves XMPP at talk.google.com. Why does this show up at all?

Posted by Arnold | June 15, 2006 09:22 AM

Thanks!

However, just to clarify, is the blogger/blogspot traffic actually just == blogger.com or is it combined? The reason is, blogspot.com resolves to blogger.com. However, there are lots of

Posted by sohbet | December 19, 2006 08:27 PM

The reason is, blogspot.com resolves to blogger.com. However, there are lots of

Posted by Sohbet | December 22, 2006 12:15 PM

It is sad to see Google News so Low on the list...

This is a classic example of NOT exploiting the full potential of an asset....

If the news was somehow more intelligently intergrated with the SERPs, it could lead to massive traffic....

Posted by chat | January 8, 2007 06:34 PM

Interesting, but I would suggest not really "the top 20 Google domains." Those would include blogger.com, orkut.com, blogspot.com, and perhaps some others. This is only those that are *.google.com.

Would be interesting to get the others in there.

Posted by oyun | March 8, 2007 06:28 PM

Thanks!

However, just to clarify, is the blogger/blogspot traffic actually just == blogger.com or is it combined? The reason is, blogspot.com resolves to blogger.com. However, there are lots of

Posted by oyun | March 10, 2007 02:18 PM

Thanks!

However, just to clarify, is the blogger/blogspot traffic actually just == blogger.com or is it combined? The reason is, blogspot.com resolves to blogger.com. However, there are lots of

Posted by indir | June 11, 2007 12:20 PM

Sorry for the confusion, the stat on Yahoo! Finance and Google Finance in the text of the posting are the % marketshare stats for the two properties within our Business and Finance - Business Information category.

Posted by sohbet | August 25, 2007 11:04 AM

Two other questions, Bill. First, are these data for the the entire Internet, or just for the USA Internet? (I suppose USA -- otherwise I suspect google.co.uk and google.de would be very high on the list

Second, have you published similar data for MSN and Yahoo and AOL? Now THAT would be an interesting comparison.

Posted by Free download | November 6, 2007 04:04 AM

Two other questions, Bill. First, are these data for the the entire Internet, or just for the USA Internet? (I suppose USA -- otherwise I suspect google.co.uk and google.de would be very high on the list

Posted by sohbet | February 1, 2008 08:27 PM