Personalisation

In the mid nineties I was fortunate to be the project manager of a large scale interactive television trial[1] (we weren’t allowed to call it a trial as the initial deployment was to 50,000 households, but as we were making some fairly fundamental decisions on technology and usability I think it fair to call it a trial)

A key part of the commercial underpinning was advertising.  We were offering a wide range of advertising options including conventional TV style commercials, strip advertisements running across the bottom of a programme and some fairly full on product placement options including the ability to dynamically change the brand being shown in the programme[2].

Unlike conventional broadcast television interactive TV allowed us to uniquely identify the individual household and to target advertisements specifically for that household.  The advertisers loved this proposition and were happy to pay significant amounts of money to deliver an advertisement to a highly targeted audience – the holy grail being to deliver the advertisement for the expensive German car to the consumer when they were about to buy a new car.

We built a sophisticated suite of data mining tools that analysed in some detail what a consumer was doing (every click on the remote for example), put this together with their profile – we had signed them up to a physical service so knew who they were plus where they lived and also blended into the mix their on line purchasing habits (we offered online fulfilment).  All very cool stuff and gave a fairly accurate view of the consumer[3] enabling us to offer relevant content as well as relevant advertisements.

This trial was cancelled, primarily for cost reasons, before we fully deployed.

Fast forward 15 years and personalisation of content on the internet is again a big deal.  Amazon offers “Customers Who Bought This Item Also Bought” and “Today’s recommendations For You”.  Google serves up advertisements I might be interested in and also adjusts the search results to suit[4]  my previous searching and usage patterns.

As the amount of online activity become larger and larger the ability of content aggregators such as Google and destinations such as Amazon to provide personalised services is ever increasing.  According to public statements Google is now tracking users’ social networking behaviour[5]  when logged in as well as a wide range of other signals such as location, search behaviour regardless of whether the user is logged in.  This level of understanding of a user’s behaviour allows very specific targeting of advertisements and searches.

The collection and use of this personalisation data is now at a level we saw as pure science fiction back in the nineties and allows a highly tailored experience to be generated for a user within the commercial world of Amazon, Google and the like.

So what has this got to do with libraries?  Well we are in the amazing position of being able to generate personalised content on a large scale within the library world.  Modern discovery style interfaces often include the ability to take account of some aspect of the library’s collection and the borrower’s behaviour.

Sorcer[6] for example provides a wide range of containers that offer highly specific result sets.  The ‘Others have Read’ container shows items that have been borrowed by borrowers in the same category as the logged in borrower, ‘Books for you’ containers derives recommendations based on the borrowers specific past reading habits.

This style of personalisation provides two great advantages to library users.  On the one hand a user can go down a path of ‘I know what I like’ and find new items that meet their profile.  Importantly there is also a broadening approach where the loan behaviour of others is brought into the mix of recommendations.


[1] 1995 Press release describing launch of iTV service. http://www.highbeam.com/doc/1G1-16009435.html

[2] Description of dynamic product placement http://www.althos.com/tutorial/TV-advertising-tutorial-embedded-ads.html

[3] General discussion on Interactive Television Advertising. http://jcmc.indiana.edu/vol9/issue2/giaglis.html

About Nigel

Product Executive Director, Civica Library & Learning
This entry was posted in Product and tagged , . Bookmark the permalink.

2 Responses to Personalisation

  1. Polyxena says:

    Hmmmm bonfire is looking good 🙂

  2. Nigel says:

    Well. It was time for a change, but on reflection normal service to resume shortly 🙂

Comments are closed.