Friday, 24 August 2007

Acupressure point for nausea, anyone?

Not as easy to find as I would have expected. I was confident there was one and I did find a nice diagram in the end on a site on a New Zealand site on nausea during labour. It is a point a bit (2-3 inches) above the wrists, between the two veins on both arms.

You can massage it gently or even apply a wristband with a plastic button available through New Zealand chemists.

Thursday, 23 August 2007

Who ranks who with Personalized Pagerank?

San Jose is pretty far from Bratislava so there is no way for me to make it to search engine strategies this year. Instead, I have been reading the amazingly useful live coverage at the Search Engine Roundtable.

In the coverage of Personalised Search this assertion attributed to Dave Davies of Beanstalk Search Engine Positioning genuinely surprised me:
Google will be assigning a value to you as a search user. This value will be used to grade your activities on web sites. This will affect the rankings of sites based on an individuals activities on that site. Users who are trusted by Google will affect the results to a greater degree than those who are not trusted.

I had thought of personal pagerank as individualised pagerank assigned to a website just for you (and people like you). E.g. Google knows I get annoyed with Web 2.0 social sites and love all things Russian. Therefore if I searched for dating sites something like Consumating would have a low personal pagerank for me. would on the other hand appear pretty high up.

The Davies interpretation would mean something quite different: his peronalised pagerank would rank my personal worth as a searcher whose search behaviour is, I guess, generic enough to be useful to influence the general pagerank for other. Goofy people, who surf all over the web, would be discounted, while systematic, savvy searchers would carry more influence.

I think if you juxtapose the two ideas, it is clear which one makes more sense: since everyone's results will behighly personalised there is not so much point in finding trusted users to watch their behaviour. Rather, in a fashion similar to how Amazon works in recommending CDs you may like, search engines can group users according to a variety of characteristics: "If you liked Consumating and Digg, you will love Sphinn."

Goofy users won't have low personal pagerank. Rather, if Google knows you are goofy, it will modify your personalised pagerank algorithm to rank highly sites other goofy people like.

Friday, 17 August 2007

Do you pay when Google employees click on ads?

I am wondering, would it be possible to geotarget an Adwords campaign to Googleplex? The land area is almost 1 sq. km so it should be enough to meet the Adwords minimum area requirements. That way you could see whether and how clicks by Google employees count.

It would be worthwhile to check with Google PR people whether there is a "Google policy on employees clicking Adsense and Adwords ads". Do these clicks get filtered? Google CEO Eric Schmidt let it slip at an August 2006 Search Engine Strategies that he clicks ads 'all the time'. This may seem like a questionable practice, especially if Schmidt's 10,000+ employees follow suit...

Several people replied in Comments that Google supposedly filtered out all clicks from its offices, and from the whole Google network (employees on laptops connecting through VPNs).

It would also be interesting to know whether Google employees are allowed to click ads outside of work. This could be a conflict of interest for some: it may be tempting to do some late night clickin' and boost ad revenues after a quarter when a hiring binge dented profits in Q2 2007.

But seriously. I deal with various Google ads a lot in my day job. Generally, I try to avoid clicking on them both on my own sites and other sites. When I search on Google, in the ads displayed you can select the URL, copy it and visit the site without incurring its owner costs.

With Adsense ads Google places on other people's sites, the standards seem somewhat looser: on many sites if you tried to click to select the URL you actually click the ad. The clickable area is much larger and often invisible when the color matches the surrounding box. It's curious why reduction of the clickable area of ads on Google's search results pages wasn't followed in ads on Adsense and Google's other advertising properties.

But no doubt this will be self-regulating: when advertising on Google becomes crap and advertisers see lower returns other networks will grow. People learn and at the tip of the pyramid stats are being scrutinised closely.