Not knowing much about the specifics of this business, my initial thought is that the users in October 2023 may be different from later cohorts. Given the lower user volume, could these October users have been internal or test users? That might explain why the retention rate was so high.
It could be helpful to segment the users from different cohorts (like Oct vs. later months) to see how their behavior differs. Additionally, as you bring more users into the application, retention rates could naturally fluctuate. Over time, once your user base grows and stabilizes, you'll likely get a more accurate reflection of long-term retention trends.
