Harry Seldon's blog

Fractals, Chaos, and Control Systems on Rails

Read

Let us talk about sex, Techcrunch and Google 2/2

Posted by Harry Seldon on December 11, 2008

After this first post about sex and marketing, let’s see my second point.
Erick Schonfeld from Techcrunch tells us about the 2008 list of the most popular search terms for the year on Google. Something is bothering me. If you notice, the first thing in the article, as it is explained, is not the “most popular” list but the “fastest rising” list. As I wanted to see the “most popular list”, I went directly to google’s zeitgeist site. Weird thing, you cannot find the list for the US but you can for all the other countries that are presented. For all the lists, I am very surprised: where is the word “sex” ? I have always been told that the forever most popular search on the web is sex. So what happened ? Did Google censor the results ? Did they cheat ? Did they manipulate the results. Did they simply make some statistical tricks to hide it ? I think that they used that last option.

If you have a look at yahoo’s results on TC’s article you will see that 3 of the top ten results are “Britney Spears”, “Jessica Alba”, and “Angelina Jolie”. I deeply apologize for saying that I do not think that guys searching for “Britney Spears” only search for song results ! So indeed people are looking for sex on the net, period. So where is this fact in Google’s results ?
The UK results are just too depressing : 1. facebook 2. bbc 3. youtube 4. ebay 5. games 6. news 7. hotmail 8. bebo 9. yahoo 10. jobs. Who can believe that Sun’s readers are not looking for something else ? About French results, who can believe that French lovers are only looking for weather and youtube ? Well, OK, Google does not tell us what French guys are watching on youtube…

Somehow the results are very biased. SEO (Search Engine Optimization) expert probably have an idea of what is going on. My feeling is that Google is not telling us exactly what they mean by query. Do they mean single query or complex query ? Let us say I look for Britney Spears sex (obviously this is for mathematical demonstration purpose only, I would never do that my dear sweetie if you read me), does this count for 1 query at “Britney Spears sex” or for 3 queries at Britney, Spears and sex. Results would be very different and using the first one will hide the word sex. So Google is most probably doing that. Web analytics experts are used to both kind of results. Indeed when you have a website you typically watch the queries leading to your site and you look both the single word queries and the sentence queries. Both give you different insights of what is going on. Knowing that Google is quite good at analytics, it is their business, I am deeply disappointed by the poor quality of the public statistical analysis of the queries they get.


The database made by the set of all the queries is just unbelievably interesting. That is the actual Google’s goldmine. First because they make their ads business out of it but also because they gather amazing intelligence information. This set is just the dream of any spy agency in the world. It is not only the most popular terms that are interesting. Using IP address, they can also check what every company is looking for on the net. They have access to the most private searches from people and company. Let’s say some guy somewhere makes a search on “a new mean to revolutionize the web search by using chaotical semantics”, well, Google knows it !


What I am saying here is that I expect a more thorough analysis of what people search because this data is after all people’ data. Anyway, if statistical data in general were more public, I guess I would not need to develop myself a public polling website.


Bonus: Writing this post I have a fun idea: making a simple website which is only a search engine, for instance using Google’s engine through the api. The only special thing it makes is to openly give a public access to all the queries. That could be fun.

Posted in | no comments | Tags , , , , , | atom

Let us talk about sex, Techcrunch and Google 1/2

Posted by Harry Seldon on December 11, 2008

Let us talk about sex: it is good for business !
I have 2 reasons for talking about sex today. Surprisingly, both come from Techcrunch posts: one about the sexy video of the moment and one about the Google’s small study about the top requests on their engine. For the first one, why this post is about sex is pretty obvious. For the second one, it is less obvious and that is my problem: where is the word sex in Google’s top requests ?

So let’s begin with the first post. Today Michael Arrington said that the sexy commercial from Fleggaard was the best commercial ever made. OK this video is very sexy. We see very pretty girls. We see them top less and even naked. And we see a lot of them in action, the action being skydiving. Ok it looks like 100 James bond girls naked. OK every man is drooling over this video. But come on “best commercial ever made” how can this be ? It is just one more commercial with naked girls. It is not really something new. So if this is the best commercial, I guess I just need to make a post with the words sex, naked, and top less to triple my traffic ? I hope I will never have to do such a thing ;-)
Is talking about sex so good for business ? Is marketing so simple ? Before these 2 posts this blog traffic is very modest but fastly increasing: in September I had 300 unique visitors, in October it was 500, in November 800. The google requests leading to this blog are very logically “rails open flash chart”, polls, “hardy to intrepid”, “harry seldon blog”. So, if in December the traffic is significantly above 1100 unique visitors (obtained from linear regression) and if the requests are more sex, naked and topless I will conclude that indeed sex is just the best marketing tool*. I will keep you posted if I see a traffic surge after these 2 posts.
For part 2, just go to this post.
* As there are many things going on around the Thinkosphere and this blog, the design change for instance, I am already seeing some traffic increase. So I will need to correct the traffic from this increase to analyze the effect of these 2 posts. Not simple but that is the beauty of stats !

Posted in | no comments | Tags , , , , , | atom

Ranking rankings

Posted by Harry Seldon on September 01, 2008

Web 2 is a lot about grading and ranking. So I guess it is logical that bloggers themselves get ranked. However there are dozens of ranking systems. Thus bloggers get their revenge: they can rank the ranking systems !
Ouriel Ohayon made a great post (in French) here and he asks the following :
Which ranking is, according to you, the most meaningful ?
You can answer below:

As I am coding the poll website thinkosphere.com you probably already know that I am a huge fan of internet polls, even if they currently lack too many features. Thanks to this post I discovered one more lacking feature : publishing a poll in several languages. I do not know about a poll engine offering this capability. I guess I will have to do it myself !

H

Posted in | no comments | Tags , , , | atom

Search

Navigate


Recent Comments

Recent Posts

Tags

actuators aircraft atc blog chaos chaos_theory charts control controllers controls crisis economy finance flight fractals git gnc gs guidance linux mandelbrot marketing navigation ns ofc on pilot rails ruby sas scs sensors statistics systems techcrunch thinkosphere tutorial typo ubuntu wifi

Categories

Archives

Syndicate


Sponsor