We propose to use reading time as a metric to report progress in language learning applications. As a case study we use a web-based application that enables learners of a foreign language to read texts from the web and practice vocabulary with interactive exercises generated based on their past readings. The application captures generic interactions with the web page (e.g. switching to a different tab) but also interactions directly related to language learning (e.g. clicking on a word to get a translation). We propose two metrics for approximating reading times based on user interactions with the web application. We analyze the correlation between these metrics and other interaction metrics and show that active time is the best metric for estimating the user's actual involvement with the texts and that it can be approximated from interaction metrics. [For the complete volume, "Intelligent CALL, Granular Systems and Learner Data: Short Papers from EUROCALL 2022 (30th, Reykjavik, Iceland, August 17-19, 2022)," see ED624779.]