Author Archives: Nadir Garouche

Mozilla Finally Goes Mobile

It was announced on Mike Schroepfer’s blog – Mozilla Corporation’s vice-president of engineering – today that they have finally decided to develop a mobile version of the most popular open-source web browser.

Schroepfer explains that many reasons motivated Mozilla to create a mobile browser: more and more people access the web from a handheld device; users want a a full browsing experience; most phones will have enough memory to be able to run a first-rate browser; etc.

I look forward to using Mozilla on my mobile, if they provide something similar to what they do for desktops, this can revolutionarize the way we browse websites on the go.

Microsoft’s Testimony Before Congress Regarding Google-Doubleclick Merger: WTF?

So Microsoft General Counsel Brad Smith is about to testify before the U.S. Senate Subcommittee on Antitrust regarding Google’s merger with Doubleclick.

I’m quite puzzled by one part of his testimony:

This country doesn’t permit a phone company to listen to what you say and use that information to target ads. The computer industry doesn’t permit a software company to record what you type and use that information to target ads. Yet with this merger, Google seeks to record almost everything you see and do on the Internet and use that information to target ads.

I really don’t get the sentence about recording personal information to influence ad targeting. Isn’t Microsoft doing the exact same thing with their AdCenter program? I mean, Adcenter’s specificity is that it uses demographic data to fine-tune targeting, such as sex, age, or income. If they’re not recording information about their users, where are they getting their demographic data from?

Sorry, Brad, but I don’t get it?

Amazon Caught Cloaking

I love playing with the User Agent Switcher plugin for Firefox. I just discovered that Amazon was cloaking its affiliate links thanks to that tool.

Cloaking is when a site displays one page to a search engine bot, while a regular user will see something different. Cloaking is used to manipulate search bots’ behaviour, and therefore obtain better search engine rankings.

I was browsing Uncrate the other day and after clicking on a link to Kanye West’s Graduation page on Amazon, I realized something fishy. Uncrate used the following affiliate link on their post:

http://www.amazon.com/exec/obidos/ASIN/B000RG1FMO/ref=nosim/uncrate-20

But after I clicked, the link redirected to:

http://www.amazon.com/Graduation-Kanye-West/dp/B000RG1FMO

Wow, something is wrong here… I checked the status of the User Agent Switcher plugin and noticed that it was set to GoogleBot…

So then I decided to check the HTTP Headers of this redirection with Live HTTP Headers,another excellent Firefox plugin. And here’s what I found:

It’s a HTTP 301 Redirect!!! Click on the image (or here) to see the text version of these HTTP Headers. I’ve checked with other Amazon affiliate links, and it always redirected to the product page with a 301. Amazon is cloaking its affiliate links to get more links to their product pages and therefore rank higher in search engine results.

Clearly, this is unfair for other websites’ owners competiting with Amazon.

If Amazon wasn’t using 301 redirects on their affiliate links, they wouldn’t have that many links pointing to their product pages, and then it will be more difficult for them to get high rankings.

Mobile Web Community Still Angry At Vodafone

Luca Passani, the man behind WURFL, came up with a long rant (entitled “Vodafone UK is abusing its position) about Vodafone regarding their transcoding solution used in the UK(powered by Novarra), which reformats ALL sites visited by their users, and therefore masks the User-Agent. This is causing issues for mobile sites that rely on user agent to provide the right content to the device that requests it.

For example, here’s a regular HTTP request sent by a Nokia 6288:

HTTP_USER_AGENT => Nokia6288/2.0 (05.92) Profile/MIDP-2.0 Configuration/CLDC-1.1
HTTP_X_WAP_PROFILE => “http://nds1.nds.nokia.com/uaprof/N6288r100.xml”
HTTP_ACCEPT => application/vnd.wap.wmlscriptc, text/vnd.wap.wml, application/vnd.wap.xhtml+xml, application/xhtml+xml, text/html, multipart/mixed, */* HTTP_ACCEPT_CHARSET => ISO-8859-1, US-ASCII, UTF-8; Q=0.8, ISO-10646-UCS-2; Q=0.6
HTTP_ACCEPT_LANGUAGE => en
HTTP_ACCEPT_ENCODING => gzip, deflate
HTTP_CONNECTION => Keep-Alive

This is a standard HTTP request format containing the standard HTTP_USER_AGENT string.
Now here’s what the HTTP request looks like with a Nokia 6288 used through Vodafone UK Network:

HTTP_USER_AGENT => = Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.0.7) Gecko/20060909 Firefox/1.5.0.7 MG (Novarra-Vision/6.1)

HTTP_ACCEPT => text/html;q=1.0, text/css; q=1.0, application/x-javascript; q=1.0, text/plain;q=0.8, application/xhtml+xml;q=0.6, application/x-httpd-php;q=0.1, */*;q=0, image/gif; q=1.0, image/jpeg; q=1.0, image/png; q=1.0

HTTP_ACCEPT_CHARSET => ISO-8859-1,utf-8;q=0.7,*;q=0.7

HTTP_ACCEPT_LANGUAGE => en HTTP_ACCEPT_ENCODING => identity;q=1.0, gzip;q=0.1, *;q=0

HTTP_X_NOVARRA_DEVICE_TYPE => 0

HTTP_X_DEVICE_USER_AGENT => Nokia6288/2.0 (05.94) Profile/MIDP-2.0 Configuration/CLDC-1.1

HTTP_X_MOBILE_GATEWAY => Novarra-Vision/6.1 (VF-UK; Server-Only)

HTTP_VIA => HTTP/1.1 glgwsl11 (XMG 724Solutions HTG XMG-31_VODAFONE_M3_B030 20070724.091400), 1.1 Novarra (Vision/6.1), 1.1 frankenstein1:3128 (squid/2.6.STABLE9-20070214)

HTTP_CACHE_CONTROL => max-age=259200

HTTP_CONNECTION => Keep-Alive

You see what’s wrong? Because the page you’re visiting is served by Novarra’s platform, the original user agent is lost. So if you receive a visit on your mobile site from a Vodafone UK customer, there’s no way for you to notice it since the user agent is masked.

Thankfully, there are workarounds for this issue. If you look at the HTTP Request you will see this field: HTTP_X_DEVICE_USER_AGENT, which contains the actual mobile device’s user agent.

It is also possible to be whitelisted by Vodafone, if you have a mobile site and don’t want visitors from Vodafone UK to see a transcoded version of your site, you need to let Vodafone know (send an email to whitelist@vodafone.com , but due to the amount of requests, it may take time before they include you in their whitelist).

So, you may ask, if there are workarounds to this issue, should we still complain? The answer is YES. People active in mobile commerce, mobile web, mobile content delivery etc must make sure that third parties do not change the way their site is supposed to work, because this is going against standards and it is not fair for both users and sites’ owners (who are of course losing money when that kind of issue arise).

AdMob vs. Adsense

Russell Beattie recently dropped AdMob to try Adsense Mobile and he’s sharing a few figures with us:

Type Page Impressions Clicks Page CTR Page eCPM Earnings
AdSense for Mobile 13,819 16 0.12% $0.10 $1.33
AdMob 22,283 256 1.15% $0.81 $17.98

Clearly, AdMob paid much more than AdSense. There may not be enough sites in Adsense Mobile’s inventory yet, which would explain the poor CTR et CPM.

Have you been able to try Adsense Mobile as well? If so, please be kind to share your experience with it vs. AdMob or other mobile advertising platforms.

Google Officially Launches Adsense For Mobile

We’ve been waiting for it, and it’s finally here. Google just announced that Adsense will be availalable for Mobile. Mobile sites owners will now have the possibility to display Adsense ads and will be paid per click.

Google Adsense will directly compete with AdMob, which is currently the leader in this arena. Let’s see how long they will remain at the top…

AdSense for Mobile will be available in : US, England, France, Italy, Germany, Spain, Ireland, Russia, Netherlands, Australia, India, China, and Japan.

New York Times Archives Now Free To Enjoy Traffic Generated By SEO

The New York Times announced that they will put an end to TimeSelect, their premium access, in order to grant free access to all visitors.

From now on, New York Times’ archives from the past 20 years, and the public domain years of 1851-1922 will be accessible to all. Some content from 1923-1986 will also be available for free.

This move was motivated in a large part by the surge of traffic that the NYT website is receiving from search engines, thanks to Search Engine Optimization efforts. According to Staci Kramer from PaidContent, who interviewed Vivian Schiller, SVP and GM of NYTimes.com, the site was applying SEO techniques since their acquisition of About.com in 2005:

“The SEO dates back to NYTCo’s acquisition of About.com. Since About.com’s SEO technology and expertise was applied, NYTimes.com has experienced triple-digit growth in unique visitors. From July 2005 to July 2006, internal logs showed an 87-percent increase in unique visitors compared with 21 percent the previous year over year.”

NYTimes.com realized that it will make sense for them to index even more content and enjoy the traffic coming from organic search, so they’ve decided to open up their million archive pages.

With TimeSelect, the NYT was making $10M a year. Since the site has about $13M unique visitors per month, it will generate much more revenue with an ad-based advertisiting than it could from online subscription sales.

I guess Google AdWords sells guys have immediately called the NYT to suggest AdSense 🙂