The Value in the Process: Building Smokefree.Gov's New Information Architecture

Background: Smokefree.gov is the National Cancer Institute’s initiative designed to provide mobile and web resources to help people who want to quit smoking, which includes Smokefree.gov web resources and 5 mobile-optimized websites that provide smoking cessation information, resources, and support to all smokers and custom content to women, teens, Spanish speakers, LGBT, veterans, pregnant women, and smokeless users.

Program background: Solid research literature supports the use of web-based and other mHealth platforms as a means to deliver interventions for a variety of health behaviors, including smoking cessation. Additionally, substantial research has been conducted on the organization, labeling, and prioritization of web content as these are key elements of Information Architecture (IA) (i.e. the underlying structure of a website). However, little research has been done to examine how IA can improve the effectiveness of smoking cessation interventions. As part of a continuous evaluation and a larger content strategy realignment, the Smokefree.gov Initiative reexamined the IA of Smokefree.gov.

Evaluation Methods and Results: The IA project employed a multi-staged process. First, a content inventory and metrics analysis were completed to assess the site’s structure and its performance. Next, a team of interdisciplinary subject matter experts (user experience, content strategy, and tobacco cessation) performed iterative card sorting exercises to create two new proposed IAs. These new IAs were evaluated through moderated tree testing, a usability technique for determining the findability of topics within a website’s structure. The team developed ten online smoking cessation tasks (e.g., You're trying to put together a step-by-step quit plan. How would you do that?). Current smokers, recruited via Facebook, were asked to complete these tasks for both IAs. Afterwards, users were asked to complete a labeling exercise to assess the clarity of the chosen category names. The findings revealed how the expanding offerings and increased resources from Smokefree.gov had evolved into a flat site structure (i.e. large amounts of content with few hierarchical categories). Tree test data (e.g., first click, success rates, completion time) indicated the sections where users expected to find specific content, highlighted strengths and weaknesses in each IA, and showed where user behavior differed from initial assumptions.

Conclusions: Through user and expert insights, a new IA has been developed for Smokefree.gov to optimize findability and navigability of smoking cessation content and resources. This user-centered multi-phased process was beneficial for not only arriving at a better IA, but also informing content strategy and management. For example, the initial content inventory identified a set of “core intervention” content, and label testing provided actionable insights to improve search engine optimization. Continuous optimization of Smokefree.gov will be measured through monitoring, user testing, and evaluation of page performance.

Implications for research and/or practice: Similar practices can be implemented with web resources across a variety of health topics. In addition to making content findable, IA improvements can lead to content that more deeply engages users. Additional research is needed to better understand how different types of content structures and labels might support the established benefits of online interventions for practice areas such as tobacco abstinence.