If we compare the efficacy of "telling people to lose weight" vs taking ozempic, there is a stark contrast. Pooled results show that education and counselling[sic] did not significantly reduce weight (SMD –0.73, 95% CI –1.89 to 0.42, n = 3 studies; I 2 = 98%)[1]. The mean change in body weight from baseline to week 68 was −14.9% in the semaglutide group as compared with −2.4% with placebo[2].
Telling people to just lose weight is not an evidence-backed intervention to affect weight loss. Continuing to practice it is equivalent to practicing folk medicine or alternative medicine. Despite us wanting it to work, it does not. Like all interventions, pharmaceutical interventions have the possibility of side effects, and it's up to the patient and doctor to weight risk vs reward. There is variation in both for each case, but we have to keep in mind that the mere existence of side effects is not typically a reason for categorically deciding against using a drug.