Improving the Relevance of Content

Content on websites is generally seems to get organised in one of two ways;

1. Some form of Information Architecture design gets produced and all siter content gets categorized under that IA; the more clever ones are able to accommodate a single content item in multiple locations. The IA provides a strict, fixed structure; better designed IA's allow for an element of extensibility.
2. The content evolves organically and gets put where the author thinks is the right place at the time of publication. Again, a fixed structure is the result.Every now and then content may be reviewed, refreshed and minor changes made to structure. Often this is a time consuming, expensive and error prone process and so happens very infrequently.

The upshot of structuring content as above is that it more often than not fails to meet the needs and expectations of the consumers of said content. One of the key challenges is that for a given site there will be many different consumers with many different wants, needs and motivations.

The practice of developing persona's to help with IA desig n is intended to help address this challence. Unfortunatley, even the best persona modelling is going to fall short. There is also the fact that consumers evolve and their wants and needs are constantly evolving, as is the content they are interested in.How then, can content be delivered to the consumer in a way that meets their specific requirements and how can that content be managed to address the evolution of the consumer?One answer could be Smart Agents. What if we built sites that could constantly optimise their structure based on the behaviour of its consumers? We could even take this a step further and enable products and services to reconfigure and repackage (bundle) themselves based on consumer behaviour.

For example, take a site that sells PC's. At launch it presents an array of standard products organised around a specific IA and enables customers to configure and customise their PC. What if the site could learn what the most popular products and customisations are? What if the site could learn what types of user go for what products and customisations. The site could then create new standard offerings and reorganise its content to ensure that the most popular offerings are higher up the IA and offerings are grouped based on the learnings of consumer type and behaviour. As the product offerings and consumers evolve so does the site.

We can go even further and start to think about how external content that is relevant to the site content can be drawn in and structured within the site as an integral part; being given prominence based on consumer behaviour.