Opencare metric investigation: finding why patient conversion fell 25%

Opencare is an online service where patients find a dentist and book an appointment. Conversions on the platform have fallen by 25%. How would you work out why? ```hint Rate or count Before listing causes, check whether bookings themselves fell or only the rate did, and whether something changed in the traffic you are dividing by. ``` ### Clarifying Questions - Which conversion fell: visitors who book an appointment, booked appointments that patients actually attend, or dentists who sign up to list their practice? - Did the 25% drop appear over a day or two or build up over several weeks, and which baseline period is it compared against? - Is the drop in the conversion rate, in the absolute number of bookings, or both? - Is it spread across every city and platform (web and mobile apps), or concentrated somewhere? ### What a Strong Answer Covers - Pinning down the metric definition, the time window, and whether the rate or the volume moved - Ruling out tracking and data-pipeline problems before investigating patient behavior - Separating causes outside Opencare (seasonal dental demand, competitors, market changes) from causes inside it (releases, experiments, marketing spend, dentist availability) - Locating the drop with a booking-funnel breakdown and segmentation by platform, geography, channel and patient type - Prioritizing hypotheses, confirming the root cause with data, and proposing a fix with ongoing monitoring ### Follow-up Questions - Suppose the whole drop sits at the step where patients pick a time slot, and only in two cities. What would you look at next? - If a recent booking-flow change caused the drop but also raised the share of booked appointments that patients attend, how would you decide whether to roll it back? - Which dashboards or alerts would you set up so a drop like this is caught within a day?