Desire paths
Posted on 01 June 2009
Desire paths in the real world
From Wikipedia:
A desire path (or desire line) is a path developed by erosion caused by animal or human footfall. The path usually represents the shortest or most easily navigated route between an origin and destination.
Desire paths show how people think and interact with an environment or a product in a real world situation. They highlight differences between what a designer expected to happen and what people really want. People won't follow a path if they can find a better way.
There are a number of anecdotal examples, often featuring university campuses, where architects and developers allow people to wander freely, creating their own paths across the landscape, before adopting the most popular routes and adding lighting and a walking surface.
Desire paths on the Web
Just as on university campuses, people want to do the things they want to do and get to where they're going as quickly and as easily as possible. But how do we spot the desire paths in the absence of actual wear and tear? In a controlled usability testing environment we can use eye tracking to see where users are looking and record video and audio of their behaviour. Outside the lab, we can record clicks to produce a heat map and analyse traffic statistics to understand movement throughout the website.
If we have an understanding of how users behave in the wild, we can adapt our design to match this behaviour. On a very basic level, we could remove items that are rarely used or change the order of menu items based on the frequency with which they are clicked, putting the most popular items first. This could be done based on a trend demonstrated by all users or for individual users, personalising each user's experience to match their own behaviour.
E-commerce
So how can we apply these principles to, for example, e-commerce? We want to provide an easy route to the products that our customers want to buy. As well as observing how users move through a website, we have the benefit of being able to analyse their purchases. This means that in addition to listing products in order of most viewed, we could also present the most popular based on number of sales. For each product, we could also show which other products were viewed or purchased by other users. On a very basic level, we are making dynamic, short term changes to the website based on users' desire paths — albeit very abstract paths.
Folksonomy
It's not just about traditional navigation. The concept of desire paths can be extended to other forms of collective intelligence. One prominent example to emerge from Web 2.0 is tagging — attaching keywords to chunks of information to build a social taxonomy. By exposing the tags chosen previously, emphasising those used most frequently, we can encourage others to follow the same pattern and allow a set of conventions and a controlled vocabulary to develop naturally.
When creating a new bookmark at Delicious, the user is presented with a list of relevant tags from their own collection, together with a list of popular tags used by other members with the same bookmark. They can see what they've done before and they can see what other people are doing, but crucially they still have the freedom to go their own way and create new tags – to create their own paths to the information.
Learning from users
In the early days of Twitter, there was no way to send a reply. Without instruction, users found a solution. They started prefixing their tweets with @username if they wanted to direct a message to a particular person. The pattern became a convention and was ultimately adopted as an official feature which has now evolved to be one of the most fundamental parts of the service. Can you imagine Twitter without conversations today?
The point being, our users are a great source of intelligent design ideas. There is no better usability test than releasing a product into the wild, observing the problems that real users discover and the ingenious solutions they devise – particularly when people work together as a team.
In the short term we can design interfaces that adapt dynamically in response to the behaviour of our users. In the long term, we can analyse their behaviour and draw conclusions which can then influence our designs for the better at a more fundamental level.
This blog entry was writen by Matt.
If you would like to discuss this post why not follow us on twitter?
