You'll love cheesy peas!

OK, this Fast Show sketch never managed the ubiquity (thank god!) of 'Suits you sir', and doesn't manage the gentle pathos of Ted and Ralph, but it still makes me laugh.

It came to mind reading an article on advertising in this week's Media Guardian which describes a couple of new initiatives to target consumers while they are shopping in store. The first uses facial recognition to work out what demographic you belong to, and then shows an advert on a screen targeted at that demographic. The second, detects the products you take from the shelf using RFID, and uses this to display an advert for a related product. In the article the example used is if you buy Shampoo for a specific hair type, the screen might suggest you get the related conditioner.

 It was this second example that made me think of cheesy peas – the assumption that if I like both cheese and peas, then there will be nothing I will like more than them both together is the kind of reasoning you probably want to avoid with this type of system.

I'm also a bit skeptical of the scenario described – people generally aren't stupid, and know if they use conditioner or not. Perhaps more likely I could see a situation where picking up one brand of shampoo led to a screen saying 'have you thought about trying this rival brand instead – cheaper, better, and altogether shinier' or something like that – so kind of aggressive advertising.

What is all this doing on a blog that claims to be about libraries etc? The other thing this article reminded me of was a discussion on NGC4LIB following the initial publicity about the Microsoft 'Surface'. At the time I suggested the idea of using a book that you already had, to show other books that might be relevant. Since quite a few libraries already have RFID or are going to in the near future, perhaps we ought to be looking at something similar to the system described above (being developed by Proctor and Gamble). Of course, you could argue that this is the whole point of having the books shelved by a classification scheme in the first place – so that the other relevant books are nearby – but clearly any physical ordering is inherently limited. I also wonder if there is a similar argument for books as I've made above for Shampoo – would this type of approach open the possibility of showing books that weren't related to the one you'd just picked up? ("That looks like a really heavy read – how about the latest chick lit for when you want a break?")

Isn't technology brilliant?

btw, I really recommend the Media Guardian supplement (every Monday, and also at http://www.guardian.co.uk/media) as a source of interesting developments relevant to tech and libraries – the way publishing is trying to come to terms with the Internet (or sometimes hoping it will go away) is very interesting, and I think libraries have a lot to learn from looking at their experience (as well as needing to deal with whatever publication formats and models come out next). This Monday was a case in point where they had a supplement (sadly and ironically not available online) about e-publishing including an piece on how Penguin are expanding into e-books, especially in Asian markets, and how their thinking about how books sales work are changing – i.e. think of a million e-book sales as a small title, hundreds of millions is what you want to aim for)