This study examined the effect of a Canva-assisted deep learning approach on the anecdotal-text writing skills of tenth-grade students at SMK YPK Medan, Indonesia. A quantitative quasi-experimental method with a pretest–posttest control-group design was employed. The population comprised 99 tenth-grade students, while two intact classes were selected through cluster random sampling: X AKL as the experimental group and X MPLB as the control group, with 19 students in each group. The research instrument was an analytical writing test assessing content, text structure, linguistic features, mechanics, and creativity in communicating social criticism. Data were analyzed using descriptive statistics, the Shapiro–Wilk test, Levene’s test, and an independent-samples t test. The experimental group’s mean score increased from 44.74 at pretest to 83.42 at posttest, whereas the control group’s mean increased from 36.84 to 57.89. The hypothesis test yielded t(36) = 11.557, p <.001, indicating a statistically significant difference between the two groups.  The findings suggest that integrating meaningful, mindful, and engaging learning processes with Canva-supported multimodal composition can substantially improve students’ ability to generate ideas, organize anecdotal-text structures, apply appropriate language features, and communicate social criticism creatively. The approach is therefore recommended as an alternative for teaching Indonesian writing in vocational secondary education.