Extensive research on process-oriented guided inquiry learning (POGIL) has addressed its impact on student performance but little attention has been devoted to approaches to learning and performance on delayed tests. The use of POGIL in online education must also be explored. This research concerns changes in students' approaches to learning and their performance on delayed tests after the implementation of POGIL and demonstrates the challenges faced in implementing POGIL online. Various changes that could be made to adapt this methodology to online learning are also outlined. In this quasi-experimental exploratory mixed-method study, two cohorts in an undergraduate architecture program formed the control (n=79) and treatment (n=91) groups. The results of surveys, performance tests, and interviews showed that student performance on delayed tasks was significantly better (t=3.719, p=0.000, d=0.50) following the implementation of POGIL and that there was a higher increase in deep approaches and a lower increase in surface approaches to learning in the treatment group after its application. We also provide recommendations for overcoming these challenges using a POGIL framework with some modifications to its eight typical characteristics.