The Problem With Letting Users Decide
Card Sorting exercises are one of the methodologies that fill the standard “toolbox” of most user experience designers. Here at UsabilityOne we use them frequently. to assist with developing site IAs.The concept of using card sorts to observe how users group, arrange and label content gives amazing insight into how users view your content and what relationships they see between items.
Considering how successful a card sort can be, it logically follows that a website that can dynamically support users sorting, grouping and labeling content would successfully evolve into the perfect IA, right?
Not really.
Recently, UsabilityOne has helped with development of several sites that have offered such a feature. These sites included areas where users could upload varying types of content, from business case studies to new inventions. During the upload process, users would give each item a category, sometimes making their own category, whilst other sites had several predefined categories.
The issue that we saw across all of these sites was that users’ perceptions of where content belonged, differed. An item that the uploading user felt belonged in one category would be sought in a different category by other users. Frequently we heard users complain that content was not where they expected it to be, and that they felt they were wasting their time trawling through multiple categories.
What had been overlooked is that a successful card sorting exercise also includes a very large amount of analysis. Labeling trends are isolated, card placement across multiple categories is calculated to see which placement was the strongest, outliers are isolated. Most importantly, the data arising from a sort is used to alongside the context of known user behavior and attitudes, as well as business goals and requirements.
The recommended inclusion of a SOMETHING search tool, as well as user generated tags for each item went some way to resolving the issue. However, the question still remains; is it appropriate to have users decide where content will be placed on your site? In these days of Youtube, Flickr, Etsy and a thousand other sites that adopt this process, it may be tempting to jump on board the user generated content bandwagon. However, these successful sites have been built from the ground up to support this sort of behaviour; their very strategies are based around user generated content and dynamic, loosely defined architectures.
If you are considering allowing users to develop content for your site, it pays to sit and consider some points before leaping in. Will it matter if some content is overlooked by users? Will your users be seeking specific items of content? Are your users likely to be invested enough to properly categorise and tag content? And the bigger question, is this really the right strategy for your site, or would your users prefer a more solid, reliable architecture that has been developed and validated by people who properly understand the content?