Text scanning patterns
When a user engages with content on a web page, they are unlikely to spend time reading every word of text. They are more likely to scan the content for headings, keywords, images or videos which are of interest to them for the task they are looking to complete.
Text scanning enables a user to be more efficient with their time and quickly source the information they are looking to find, which can often be surrounded by unnecessary additional text. A user’s scanning pattern will likely change for each web page they visit depending on their task, previous online experience, page layout and page content.
The Neilson Norman Group article on Text Scanning Patterns: Eyetracking Evidence discusses four common text scanning patterns which I have summarised below before discussing it in the context of the University’s web pages. Examples of each type of pattern can be found in the original Neilson Norman article.
Types of patterns
The F-pattern is a common but somewhat inefficient way of scanning. F-pattern scanning is most likely to take place when a page is not built using basic formattings such as subheadings or bulleted lists. Users who follow the F-pattern are found to mainly read the text which sits on the top left of the page, leaving the text to the bottom right unread. F-pattern is so-called as eye-tracking heatmaps reveal an ‘F’ style map.
The spotted pattern is seen as a more effective way of scanning as users will be drawn to certain words on the web page they are on. Users are likely to be attracted to particular words which stand out because they are formatted differently. This might be because they are bolded, hyperlinked or use a different heading style from the rest of the page. Words associated with a particular task a user is wanting to complete are also likely to stand out. For example, if a user is looking for contact details, they will be drawn to any numerical numbers which may represent a phone number or the @ symbol which may represent an email address.
Layer-cake scanning pattern
The layer-cake scanning pattern is another common and efficient way of scanning compared to F-pattern. Layer-cake scanning is based on a user methodologically moving down the web page looking at headings and subheadings until one is reached that is of relevance to them. Once they have located a section of text which helps them complete their task, they are likely to spend time reading the accompanying text.
The most comprehensive yet least efficient scanning pattern is the commitment pattern. The commitment pattern is similar to traditional reading where, instead of scanning, a user will read every word on the page. The commitment pattern is most likely to occur when a user has a particular interest or motive for reading the content. Examples of this could be instructions, news articles or reading for enjoyment. Commitment pattern users are also more likely to trust the source of the information and be loyal to the brand.
When a page is designed with commitment pattern in mind, it is still important to correctly format the page using headings, subheading and lists where suitable. This allows the user to still have the ability to move between different sections and pause when required.
University web pages and text scanning
University web pages are built in a way which is more likely to support spotted pattern or layer-cake scanning. F-pattern scanning is not desirable for a web page as it is often a result of poor design and risks leaving a large portion of text unread.
We try to avoid F-pattern scanning by ensuring that all new web content is correctly divided into appropriate headings, sub-headings, accordions and lists where possible. T4 users are also trained to ensure that their writing is concise, scannable and objective to allow users to reach the information they are looking for in a quick and efficient manner.
The University’s digital standards provide information on our digital content policy and links policy. These policies also help to ensure that all content and links meet a high standard which has the user in mind and ensures consistency across the website.