This disclosure describes techniques to measure user friction in search tasks performed by a user using a search engine. User friction can be defined as the inverse of total time spent in refinement before a user obtains a desired search result. User friction is indicated by the amount of user actions taken before the user obtains a desired search result. Based on user-permitted metadata indicating queries and actions input by the user, a user search session is divided into user visit segments classified based on the query intent of the user, each segment indicating a search task. User friction scores are determined for completed search tasks based on counts of user friction interactions, such as returning to a search results page after selecting a search result or refining a query. The described techniques can measure friction in user search tasks involving multiple interactions in a user journey and can enable improvement in the experience of using a search engine.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.