Slashdot Mirror
Google Launches Trends
An anonymous reader writes "Google started to offer a new Trends service that allows viewing search term request statistics split up by geographical locations and languages. In short one can use Google trends to figure out what's hot and what's not and perhaps even find cyclic patterns to pick best time to advertise. From my poking around Google trends I have noticed that there appears to be a general declining bias for most search terms that either has to do with the declining popularity of Google (i.e. less folks were using Google for the past two years) or with the declining amount of searches in general (which is highly unlikely)."
I have noticed that there appears to be a general declining bias for most search terms that either has to do with the declining popularity of Google....
Dumb. Maybe he's correct, that google popularity is declining, but the examples he gives do not back up his hypothesis.
There's less searches for Ultramax, Trance music, Madonna and Britney Spears than two years ago. Thats because those items are less popular than two years ago.
Oh, and doing a quick search for the author of this 'blog' (which is starting to look suspiciously like a pagerank pusher), I see he has an ecommerce site called ultramax music, that features trance).
So - two of his search items are related to what he's selling! An interesting way to get your name, music & company linked from a high profile website.
There are shills on slashdot. Apparently, I'm one of them.
Google launched trends in May, so this is old news and a dupe. Nonetheless, here's an interesting trend.
I've used google trends before, is this a new version or just a formal release?
I ran my own tests, and I think that the "search volume" axis is relative to the total number of searches on each day, rather than an absolute number. Most of my tests didn't show a declining bias, although I saw that in a few cases. I think the "declining bias" you saw might be caused by an increase in the diversity of search terms. Old search terms never go away, but new search terms are constantly being created as new names and catch phrases enter popular culture.
Doug Moen
I have written a truly remarkable program which this sig is too small to contain.
seems like only 3 years ago I was saying links of /. should be norel.
then maybe submitter would be mor pleasant with his choice of hyperlink text.
Here is a nice link Google Trends
have a nice day.
Oh oh oh. almost forgot, this is about 9 ice ages old. oh ohoh
do a search for
lemon, lime
it is by far the best search you can do on trends! awesome!!
#hostfile 0.0.0.0 primidi.com 0.0.0.0 www.primidi.com 0.0.0.0 radio.weblogs.com
... living in the perpetual darkness of Finlands arctic winter for a few months a year sure can make you lonely, but that lonely?
Don't click here: http://www.google.com/trends?q=ubuntu%2Cdebian
Sorry for me spell bad, not a native but I'll do my best
sorry. I am tired.
#hostfile 0.0.0.0 primidi.com 0.0.0.0 www.primidi.com 0.0.0.0 radio.weblogs.com
Hmm, I can see an upward trend here. However the data they have here is very suprising!
I tried out "Firefox, Opera, IE" and was hoping google would be cool enough that I would get a comparison of trafic from each of the browsers. No such luck. Come on all of you Google employies. One of you needs a 20% time project I'm sure. Put in a fun set of 'easter eggs' that catch browser traffic comparisons or platform traffic comparisons, or other specific comparisons for an appropriate search. I'm sure there are more interesting trends than just 'how many times did you search for x'.
I hear you man.
fucking douchebag. lets get some mints and shove them in his eyes!
man, dane cook kills me.
#hostfile 0.0.0.0 primidi.com 0.0.0.0 www.primidi.com 0.0.0.0 radio.weblogs.com
http://www.google.com/trends?q=bomb+making&ctab=1& geo=all&date=all
Indeed, this isn't a new feature, and it certainly looks like someone has an agenda here. Where are people going, if not to google? And the expected trend of any specific query is to decrease over time as popularity wanes.
There is no escaping this one...http://www.google.com/trends?q=porn&ctab=0&g eo=all&date=all
http://www.google.com/trends?q=rms%2C+esr%2C+bill+ gates&ctab=0&geo=all&date=all
Good. Cheap. Fast. Pick Two.
ice cream, snowboarding
o arding&ctab=0&geo=all&date=all
http://www.google.com/trends?q=ice+cream%2C+snowb
According to this Press Release, Google released Trends on May 10, 2006 along with Google Co-op, Google Desktop 4, and Google Notebook. It appears that the guy who wrote the article just happened to come across Google Trends, put his own website words into it, and IMO he doesn't understand how it works.
Funny createSig(Witty remark, Odd reference)
{
return (Funny)remark + (Funny)reference;
}
http://www.google.com/trends?q=Japan%2C+Europe%2C+ United+States%2C+China%2C+Canada&ctab=1&geo=all&da te=all
Whoever knew Canada was so popular?
Modern culture's hierarchy has changed!
New Maslow
/idle speculation
It would have to mean that less people are searching, because what else would these mysterious people be using to search with if not google? The browserbar that some spammer hijacked onto their copy of IE? Google is search; so it's reasonable to assume that if less people are using it, it means that less people are searching.
There exist many better trendspotting tools, e.g. try BlogScope. For example, compare YouTube and Google Video.
NetBSD is dying.
Someone please help me analyze this one.
The AACS key is NOT 0xF606EEFD628B1CA427BEA93A9CA9773F
Try this search to see that Slashdot still maintains its high quality standards :-)
Ouch. :( http://www.google.com/trends?q=slashdot%2C+digg&ct ab=0&geo=all&date=all
/. much more than Digg*
*likes
http://www.google.com/trends?q=france%2C+england%2 C+germany%2C+Canada&ctab=1&geo=all&date=all
It's truly scary when you realize why the parent trend looks that way.
Now this is interesting:
0 &geo=all&date=all
0 &geo=all&date=all
Man v. Woman
http://www.google.com/trends?q=man%2C+woman&ctab=
Men v. Women
http://www.google.com/trends?q=men%2C+women&ctab=
Ramen
Take, for example, DS vs PSP. Any look at sales charts will tell you that the DS is ahead, but PSP is ahead in searches. Unless there's a whole load of Paint Shop Pro fans skewing the results.
I'm scared of numbers that can't be written as a fraction. It's an irrational fear.
Here you go, this is much more intresting and easier to compare.
My hypothesis has finally been proven! While geeks remain unpopular, pizza and lesbians are steadily gaining in popularity.
Exactly. Looking at holiday related terms is very telling Wow... that last one is odd. There's a very telling lack of a peak there... Since july 4th was almost 2 months ago, one would assume that there should be a peak. Either people did not celebrate Independance Day in the US this year, they celebrated in a different way, or google trend's data is at least two months out of date, and therefore useless for trendspotting.
I'll never make that mistake again, reading the experts' opinions. - Feynman
While this is technically old news - the "beta" has been available to play with for some time now. I remember looking at it to prove a point about "porn searches" to my friend - that US was NOT #1 on the categories WE could think of (only about 4 so I KNOW we missed some)... and being surprised that we were further down than 10th... Arab nations almost empirically held the top places... But the point is, that this has been around now. But, it IS news that something Google has put out IS listed as 'complete' and not just "Beta" for once. That is a rare thing. It seems that all of Google stuff is "beta" forever. :o)
-- "You must be the change you desire to see in the world." Mahatma Gandhi --
Of course it isn't updated with data from yesterday. A cursory glance at the scale of the graph shows that the minor tick marks represent quarters, and there's only 1 tick mark in 2006 showing. Thus, we've got reliable 2006 data through a bit after the first quarter (likely through May and maybe into June). You can clearly see that it's starting to rise at the end, and if we had the data, we'd certainly see the July peak.
And just because it doesn't have the latest data, doesn't invalidate it's use for trendspotting. It merely won't tell you what the latest fad is.
How can any x-y graph without units on the y axis be useful? Without units this entire thing is just mental masturbation. You can play around with it if that sort of thing makes you happy, but I can't see any actual value coming out of it because without numbers it's just meaningless pictures.
Hmm, okay, so can i prove something with this?
% 2C+internet%2C+internets%2C+email&ctab=1&geo=all&d ate=all
http://www.google.com/trends?q=sex%2C+politicians
i.e, everyone wants to know where sex can be found, a heck of a lot of people want to know where their email is, a elite minority are looking for that internet that was sent them but never arrived, "internets" clearly remain off the practical radar for now, and no-one cares at all where their politicians are?
I'm beginning to think this is a marketing / pr / scaremonger's wet dream
If you swap adjectives for nouns ( use 'canadian' and 'japanese' etc, instead of 'canada' and 'japan' etc ) then the trend is not there. There is not a trace of it. I think that the trend for the word 'Canada' is a bizarre artifact. If we knew why it was there then we might know a lot more about how google trends relly works - or fails.
a n%2C+American%2C+Chinese%2C+Canadian&ctab=1&geo=al l&date=all
http://www.google.com/trends?q=Japanese%2C+Europe
http://www.google.com/trends?q=porn
Man is a slave because freedom is difficult, whereas slavery is easy.
"google evil" vs. "satan evil" very interesting!
a n+evil&ctab=0&geo=all&date=all
http://www.google.com/trends?q=google+evil%2C+sat
http://www.google.com/trends?q=woman&ctab=1&geo=al l&date=alll l&date=all
http://www.google.com/trends?q=women&ctab=1&geo=a
Pretty crazy that Iran and india are the top searchers for those words.
I'm baffled, if they want porn, they should be searching that.
The google results for women/woman yield wikipedia and a bunch of informative sites about womens's health!
Very interesting, also looking at cross country skiing shows Canada has large interest in the sport, and ottawa is the capital of it!
Canada also wins the beer category!
Vietnam searches for sex the most? Notice how "backwards" or theological countries have more of these searches??? Crazy
Apparently Turkey, New Zealand and Denmark have the highest number of searches for 'Google Trends'.
2 &ctab=1&geo=all&date=all/
So what's that all about?
http://www.google.com/trends?q=%22google+trends%2
Isn't Google Trends geolocation inherently innacurate? As I understand, it is using your IP address to determine your location, but this only gives you the location of your ISP router, and the two can be different by hundreds of miles. Since ISP routers tend to be located in urban areas, they tend to get a disproportionate amount of traffic. There was an interesting article about the problems with Google Trends, pointing out how for example the suburbs of Milton Keynes in England are constantly highly ranked in Google Trends, because BT has a large routing center in the area.
The trends are interesting, but I'd be careful about how they are used.
I think that volume search doesn't necessarily represent traffic volume or popularity. In the end, slashdot has it's crowd, which has established long ago with a very high volume, and digg is much newer. Also, means that /. ppl can remember the URL more easily.
I like this comparison.
Well, this is because women travel in packs, thus the singular "woman" doesn't really exist...
In short one can use Google trends to figure out what's hot and what's not amongst those Internet users that search on Google and perhaps even find cyclic patterns to pick best time to advertise on Google.
Maybe that sounds picky, but you have to consider the sample before you start making conclusions and generalizations.
David: Too much, there's too much fucking perspective now.
i\hbar\dot{\psi}=\hat{H}\psi
Those New Zealanders are horny bastards... http://www.google.com/trends?q=porn&ctab=1&sa=N
http://google.com/trends?q=ubuntu%2Cmacintosh http://google.com/trends?q=mac%2Clinux http://google.com/trends?q=osx%2Cubuntu (note that you get the top cities/regions for the first term)
http://www.google.com/trends?q=jobs&ctab=0&geo=all &date=all
So, the data is 2 months out of date. Therefore of very limited utility in predicting actual trends in "internet-time." Granted, it could still be useful for cyclical trends: seasonal, holiday based, back to school/etc in determining periodic demand, but not as useful as "Wow, this thing is about to get really big, and we'd better jump on it NOW." Then again, maybe that's actually a good thing for society... the act of "coolhunting" (I personally hate the word, but that's what I've read it as... trendspotting is a much better term) has really driven the fad cycle faster and faster to the point where a fad used to last a year or a couple years, now they last on the order of months. This cycle contributes greatly to the disposable culture that we are sinking deeper and deeper into with all the negative consequences. Then again, maybe fad cycles will accelerate to the point where large factories can't be tooled up in time to supply the demand, and people will start learning to DIY or be square.
Ugg... I can't believe I just posted that on slashdot. Or rather, am about to post it...
I'll never make that mistake again, reading the experts' opinions. - Feynman
Do a search for "kerry", and you will discover and interesting pattern up to election day and afterwards? Well, let's just say that I can't find a better analogy for flat-lining.
Isn't this very very old news? I think Google Trends was launched at least a couple of months back and possibly before that.
wheat + chaff :)
To reign is to serve.
I like that you can see Penny Arcade's Strawberry Shortcake strip make a big enough splash to notably affect searches on "strawberry":r awberry&ctab=1&geo=all&date=2004
http://www.google.com/trends?q=penny+arcade%2C+st
This sig is not the Zahir. Lucky for you.
Pepsi beats Coke
t ab=0&geo=all&date=all
California beats Florida (and Hawaii)
Right beats left
MIT > all
And apparently, the number of people who say tomato almost exactly matches the number who say tomato, although it varies slightly by city as seen in the chart at the bottom. http://www.google.com/trends?q=tomato%2C+tomato&c
https://www.eff.org/https-everywhere