Many students struggle with algebra and often turn to instructional explanatory videos (IEVs) as learning resources. This feasibility study explored whether students implicitly recognize quality indicators in algebra videos and how these influence their choices. Using Conjoint analysis in a 2×2 design, we examined the focus on subject-matter didactical and multimedia features among 34 students aged 14–15 years. We presented authentic IEVs on solving algebraic equations varying in didactical and multimedia features. Bayesian analysis revealed that both didactical and multimedia features influenced video choices with no notable differences in impact strength. Student characteristics (mathematics achievement, interest, and self-concept) affected preferences only for the didactical features. The study concludes that students can identify quality indicators in math videos, although this recognition is limited and may be further enhanced through targeted guidance. Methodologically, conjoint analysis demonstrated its efficacy in mathematics education research, effectively addressing social desirability bias and implicit behaviors.