Digital technologies are increasingly being used to support school learning, but few studies have assessed the effectiveness of these new teaching methods for very young students. The aim of the present study was to assess the impact of implementing a digital notebook application designed for a stylus-oriented tablet in kindergarten classrooms. This digital notebook was dedicated to the acquisition of handwriting skills by beginning writers. Using artificial intelligence to finely analyze the spatiotemporal characteristics of handwriting (i.e., shape, order and direction of the segments), the exercises were personalized, and extrinsic feedback was delivered at the end of each trial to inform learners of their results. A total of 22 kindergarten classes participated in a 12-week teacher-implemented program, half working exclusively with paper and pencil, and half partially undertaking their handwriting training with the digital notebook. A paper-and-pen writing task was administered as a pre-test and post-test to assess the progress of all the children. Data analysis showed that learning outcomes with the digital notebook were contingent upon the students’ initial handwriting level, as the benefits of training with the app were only demonstrated for children with a medium level at the start of the study. The results are discussed in the light of the literature on the impact of extrinsic feedback and learners’ initial levels.