Skip to Main Content (press enter)

Archive for October, 2004

Case studies, blog postings, research papers, tools info and eBook notes

Google Search Appliance Launches

According to ZDNet Google have let their enterprise search appliance loose on the UK market. The Google Search Appliance starts at £19,000 and scales up to 150,000 documents, and slots into your file server stack in bright yellow. Desktop search = FREE, Google Search Appiance = £19,000, there really should be something in the middle.

(Filed in Blog, October 25th, 2004)

Postcodes vs. Online Profiles

A friend recently asked me about my opinion on the use of postcodes as a predictor of social grouping, and their subsequent effectiveness as targeting method for direct marketing. He had been reading Dave Weinberger’s Small Pieces and had been debating their lack of granularity with a postcode guru, particularly when compared with the accuracy of online groups.

The problem I see moving forward for *traditional* marketers is that the self-defining online-interest groups, which definitely makes them better targets to market at (whether this defines them better socially, or not, I can’t help you with) are very granular, making location (of said groups) and marketing deals (with whoever hosts the shared interests) labour intensive. And naturally these groups are also deeply opposed to unsolicited communication. These two conditions break the mould of the direct marketeer, who would otherwise select geographically and use the one distributor (the Royal Mail, in the UK) to deliver the message, and have got very efficient at it.

I haven’t read Dave Weinberger’s book, but have found him compelling in the past, particularly in the guise of Cluetrain co-conspirator, so am not sure how he proposes these social groups are reached, although I have a reasonable idea. Which takes me back to the point above.

It strikes me that the postcode guru is chasing the dinosaurs of the marketing world. At the moment post codes are a blunt but not ineffective method, and no doubt their tools sharpens it up some, but just suggest to him the power of eBay‘s data relating user login’s with their browsing, watching, buying and selling classifications. These are examples very small, but highly powerful pieces of data, which in due course will wipe the floor with post codes as an effective means of marketing.

These dinosaurs also have absolutely no reliable means by which they can associate post codes with email addresses or user identities online. Consequently, as the reach, effectiveness and productivity of online marketing increases – which *won’t* be based on post code – these dinosaurs will lose their way in the world because they will be too expensive and too ineffective.

(Filed in Blog, October 19th, 2004)

Google Clusters – no not a new breakfast cereal!

Clusters. This month’s small-to-middling topic on the search horizon. Google has been showing off it’s clustering technologies at the recent Web2.0 conference (Danny Sullivan’s write up and an MP3 of the session) and Vivisimo has taken its search grouping to the next level by launching the somewhat tweely-named Clusty search site.

Clustering is based on Bayesian (A method of combining the likelihood ratio with additional information to produce an overall estimate of the strength of a piece of evidence, named after the Reverent Bayes) principles and is particularly useful when search terms can have source material across completely different areas of interest, as it seeks to group the results in their most likely overlaps first. I am not convinced, however, that clustering will make a significant difference the hunt for spam pages.

(Filed in Blog, October 19th, 2004)

October 1st, 2004

Froogle UK appears to be just around corner, as it has quietly started asking for applications for feeds to the UK site. I have worked with Froogle on the site for a US client, and found it incredibly simple to work with. Optimisation proves to be a separate challenge, but not an insurmountable one. Conversion rates from Froogle seem to be universally low, though. Having a UK site to use may help understand why this is the case, as I have never used the US site as a consumer. More on this when I have a client feeding the UK engine.

(Filed in Uncategorized, October 1st, 2004)