From the result there are three novelty, first The construct Energizing Value has more influenceTo WOM than the functional value.  Second, after the student make decision to intention to enroll only few to make recommendation to other friend because R² is only 15.4 %, the model in how topredict the intention to enroll, is produced in thi research by using the machine learning modelwith R2 is 24 %. This research using two different methodology approach. The first is usingThe quantitative method by using PLS-SEM, in order to determine with indicator have reliableAnd valid , and to test how much the influence factor between functional value or epistemicValue or Energizing Value to intention to enroll and also the influence the intention to enrollTo wom. After that in this research will produce the regression model to determine toIntention to enroll This research uses quantitative methods, by taking primary data 437 primarydata from high school students. The founding from this research is the Energizing value issecond strong path will influence to intention to enroll after the functional value, but thefirst influence to WOM, Beside that in this research is produce the model to predict tointention to enroll . intention to enroll = 0.996 +0.087 Functional Value + 0,13 EpistemicValue + 0,15 Energizing Value, so The Energizing Value is the strongest factor to predict the intention to enroll