Hostipfox

This extension is pretty intrusive in the sense that it will lookup IP and query location for every link you point at. Only for geo-geeks :-).

IP2LocationFox

Rightclick on a link and choose the “IP2Location Lookup” option:

Here you also get the ISP.

Why is the location of a web hoster relevant?

It’s relevant if you’re into SEO :-). Search engine results depend on the national version of Google you’re using, and the location from where you search. For example: Belgian web sites will be favoured strongly in searches on Google.be, even if the search is not restricted to Belgium.

Both Peter Forret and I write in English and have an international .info/.com extension. Pascal.vanhecke.info is recognised as Belgian, whereas Peter Forret’s blog isn’t. I get a proportionally high number of search hits via Google.be, for Peter’s blog there will be no such effect, but there will be for some other subdomains of his that are hosted in Belgium.

When you have a country domain extension (.be, .nl…), Google will recognize your “nationality” anyhow (example: the Second Life forum I started last month, is Belgian, even if it is hosted by the same US company (referral) as Peter’s blog). However, experiences like this indicate that the location of your hosting still matters even if you have a country domain extension (Dutch traffic on a domain with a Dutch .nl extensions dropped dramatically when the hosting moved out of the Netherlands).

The international and French, UK… etc version still show no change.

The search for the disappeared pages shows a link to a copy of the verdict, but the Belgian Google News and search home fail to formally mention the ruling as demanded by the court:

The court ruling did not order to take them out of the Search Engine Result Pages… just the removal of the cached pages and the excerpts from Google News…  One might be tempted to interpret “all their sites” as including the SERPs, but the rest of the verdict really makes it clear it’s about the copies and excerpts of the newspaper articles respectively in Google Cache and News.

Is Google “taking revenge” by removing them from the index alltogether?  And is it just a matter of time before they disappear from the the non-Belgian versions as well, or does Google consider the Belgian court not to have any authority on their non-Belgian sites?

To be continued…

Extra:

• Further international reporting: PaidContent, Matthew Ingram,  GMSV, SEW Blog
• I corrected my statement in yesterday’s posting on a necessary prior consent from a publisher before its content can be indexed.  The wording of the verdict really focuses on excerpts (in Google News) or copies (in the Google cached pages) of content.  It does however still ignore the practice of the NOCACHE, NOARCHIVE robot tags and again raises questions on what is considered fair use for news aggregators.

The plaintiff’s copy of the ruling is available (pdf) in French and in English translation (from p. 19 on). Here’s the conclusion (p20):

The overview of the facts (p21), describes the Google News service:

According to the report (p21-22) of the expert (Luc Golvers, who is university lecturer and president of the Belgian Association for Computer security), Google News is not a search engine, but an information portal, implying that Google has created a derivative work without respecting the original copyright holders.

But now follows a strange twist in the motivation:  a description of the Google cache.  There are no “in cache” links in the Google news sites, so without further clarification, arguments are drawn that refer to the way all regular search engines operate.

 … and in its wrap-up, Google News is being held responsible for the effects of the Google cache, more in particular the fact that a withdrawn article does not disappear (immediately) from the Google cache (bullets 1 and 4). 

These points are combined with arguments that traditionally have been used against news aggregators and deeplinking (so far unsuccessfully, see the Kranten.com case as discussed in a previous posting): the news aggregator disintermediates news publishers by making people skip the home page, reducing the stickiness of the news site, etc…

There’s a lot more to be said about:

• the apparent arrogance of Google (not even notifying news publishers officially of their inclusion in what is indeed a news portal) on the one hand… **
• and on the other hand: the (pretended) ignorance of news publishers, and their inability to monetize the traffic generated by Google News***. 

But let’s now stick to the legal case:

The mixup of the working of a news aggregator and of general search search engines is regrettable because it fails to deliver a clear verdict on the legitimacy of news aggregators.

I doubt the same wording of the verdict will be held up in appeal.  If that were the case however, it would be a dangerous precedent.  According to the verdict’s logic, a search engine should ask permission explicitly from any publisher or website before indexing  and caching its content,  ignoring the well-established practice on the web of robots.txt and robot (NOCACHE, NOARCHIVE) metatags.  That could spark a flurry of opportunistic court cases against Google and other search engines, seeking compensation for copyright infringement, and make the search engines to actually establish a procedure for publishers to have their content indexedcached.

Please note that there’s nothing specifically Belgian in the argumentation, since all copyright and intellectual property legislation is based on the same principles in the European Union and in the US.

Sources: NL: lvb.net,  FR: Le Soir, EN: Google Blognewschannel
(Updated 18 Sept: EN: Reuters, Google Blogoscoped, Virtual Economics, Tech Dirt)

* As represented by Copiepresse, the company that had filed a complaint.  The Dutch-language Belgian Press, as represented by REPROCOPY had asked and obtained removal of their articles from Google news (not from the general index) but had not taken any judicial steps.

**They do have a Dutch engineer (audio interview) in the Google News team, and also have a legal representation in Belgium

***For a friend of mine, an independent journalist on EU affairs, Google News is his main source of incoming traffic ànd income by both subscriptions and article purchases

[tags]google, syndication, aggregation, Belgium, googlenews, publishing, media, copyright, fairuse, news.google.be[/tags]

That’s no longer the case, as you can see in the screenshot, whereas an explicit link:pascal.vanhecke.info still gives the expected result **:

This made my Google blog search bookmarklet stop linksearching – so if you installed the bookmarklet, go and get the updated bookmarklet …

** I did some googling to verify the behaviour really had changed over time, but this early Google Blog Search review confirms:

Interestingly, Google Blog Search automatically prefixes a URL search with the link: operator, defaulting the search to incoming link sources.

Language

The evaluated search query runs on Google.nl, the localised Dutch version – which makes it sensible to write in Dutch and to link from Dutch-language pages (localised Google versions favour pages in the corresponding language – even your location partly determines the search results). But the second word is English (not sure whether Google recognizes the word “sitedeals” in the proper noun “sitedeals15”), which makes it sensible to write in or at least to link from English language pages (since they’re more relevant to the keyword “released” – see the explanation on “multivector pagerank” at Sitepoint).

Strategy

At first sight, the top current positions seem to follow 3 different strategies:

On-page optimization

Using keyword placement and frequency in the URL, title, and links, and applying “meaningful markup” such as headings, emphasis in the body text. Examples: Sporthumor, Vrollie.

Blogging

Regularly adding fresh and relevant content to the page: Sven Bergiers, Pebbels, the Telenet blog and Bart’s blog,

Backlinks

Making sure to gather as many links to the page with the keywords “sitedeals15” and “released”: some huge link factories like startplezier and opishop perform surprisingly well. Surprisingly, because it’s contrary to Google’s own propaganda they value quality (relevance) over quantity (such as site-wide or domain-wide links at the bottom the startplezier link pages – nothing wrong with that BTW, but Google pretends to weigh this kind of links a lot less). And notwithstanding the publicity the crackdown on “black hat” techniques got (such as the BMW ban), some tricks still seem to pass under the radar, as this experiment showed.

Another surprise is that “fresh” domains, such as Bart’s .be domain (registered on the day of the contest launch), or svenbergiers.be make it to the top 10, despite the “Google Sandbox” (BTW: I’m doubting it is helpful to have a URL with the concatenated keywords sitedeals15released – using hyphens, slashes or %20 separators seems to make more sense – or am I wrong?).

Behind the scenes

If you try the same query on ask, (the forgotten AskJeeves search engine that is desperately – though with a lot of money – trying to recapture its share of the market), a lot becomes clear. The former butler is so helpful to show the referring pages instead of the referred pages (hey, what a coincidence they’re only fourth!), and shows us goodbuy.be, Multiwave.nl and vakantielokaties.nl (pointing to startplezier.nl, but from different IP addresses, so that the sites seem independent to Google), belg.be, pointing to Pebbles, winprijzen.be and winprijzen.net pointing to svenbergiers.be, tijd.net pointing to lvb, Bart obviously points to himself. The lesson: venturing on the web is like venturing in real estate: it takes patience and long work to build up a portfolio, but once done that, you can harvest the fruit of your labour for a long time to come. No high rank in SEO games if you cannot rely on plenty of “virtual real estate”.

My bets

I still believe in Google. I still believe they’re trying to optimize their algorithms in such a way that trusted sites, who patiently built their reputation, win over link gaming. That is why it is not a coincidence Luc’s entry and Bart Noppens’ blog posting shot up so quickly in the results, after their late entry in the game (Feb 27 and March 6 respectively). Both on sites with high pageranks (they didn’t make the mistake – I think – of quickly creating something new). Both have  some impressive virtual real estate (blogium and Brussels Journal – you’ll have to search in the noscript tag at the latter) or Pagerank-rich friends. And if the first Bart 301’s, his blog into a subdomain, or even better a subdirectory of netlash.com, I ‘d wager he has even better location (pagerank) and backing (referring sites)! May at least one of you catch up with the Dutch link factory :-).

Watching the engines

Being the prime target of SEO, Google is protecting itself to such an extent it makes their timeliness suffer: new sites tend to show up earlier in msn search, for example. So it’ll be fun to watch how msn, yahoo and ask differ from Google. I don’t know anything of engine-specific optimization, if you have ideas and resources, please share them in the comments.

And finally…

But hey, wasn’t I going to stage my own attempt to storm the SERPs? Wasn’t I going to try the impossible with almost two thirds of the race run? Of course I am!
So, if you:

• enjoyed at least one of my posts and regretted I don’t have PayPal
• want to prove that even sluggish search engines like Google cannot ignore fast-linking bloggers
• want to fight Flemish and Dutch Big Online Media with the help of the blogging masses
• sympathize with desperate causes…

… then copy this link in one of your postings: “sitedeals15 released”. You will not only show up in the nifty Technorati graph (get your own here), there’s more:

You can win an Ipod Shuffle!

Indeed, if I win the contest, you can have my Ipod Shuffle… if you post on Friday March 24th at the latest (that is in 2 weeks) and on April 1st show up highest on this Technorati support list (of postings referring “sitedeals15 released”, ranked by authority). OK, I know it’s not the V7ndotcom Elursrebmem $4000, but at least it is something!

So start linking and enjoy the contest, I’ll post follow-ups as comments to this posting, you can subscribe to the comment feed!

PS1: none of the links in this posting has a rel nofollow – ain’t that generous :-) ?
PS2: if you didn’t notice “tongue-in-cheek-mode” of last paragraph, please note that none of this posting bears any moral judgement. I do not state that this or that technique is to some extent “unfair” or “wrong” – search engines have created a playing field and you have to play by the rules of the game…

• In a one-click action, you can remove a result for a specific search query, with an additional click-and-submit you remove the page or an entire domain for all future searches.
• Requirement: Google’s personalised search, for which you need to be logged in with your Google account. There’s no other way than an account in some way or another if you want to make your ban list persistent
• Privacy concerns: you can switch of the “search history” feature of personalised search if you don’t want Google to record and store your search requests all day (which are, sort of, the footprints of your thoughts..)
• For the time being, removing pages from you search results has only consequences, for your own future personalised searches, it is not a collaborative spam tagging effort yet. I assume Google will investigate whether this can be used as a collaborative spam tagging effort, but obviously they’ll have to tackle possible abuses first (such as blacklisting competitors etc…)

Update Aug 27th 2006: see here

Bookmarklet installation**

Firefox

• Right-click this link: Google Blog Search and choose “Bookmark This Link”
• Click the “Create in” dropdown and choose “Bookmarks Toolbar Folder”

Internet Explorer

• Right-click this link: Google Blog Search and choose “Add to Favorites”
• Click “Yes” if there is a “May not be safe” popup
• Choose “Links” in the “Create in” pane and click “OK”

Now you’re off keywordsearching (when a few keywords in your browser window are selected) and linksearching (for blog links pointing to your current location).

What you get is the default search… you can adapt the “blogsearch?q=” phrase at the end of the javascript to any of the advanced options by test-and-replace the url.

BTW, if you use the Firefox search plugin, there’s already a version available for Blog search.

* which is how Technorati calls bookmarklets
** instructions copied and adapted from del.icio.us, Safari and Opera users are probably intelligent enough to do it on their own ;-)