A short-video platform found that the average daily usage time per user dropped by 12% in the last month. As a data analyst, from which dimensions would you start the analysis? Please list 3-4 core analysis directions and specify the most important metrics you want to see for each.
分类: case
难度: easy
标签:
答题技巧
1. Break down the change over time; 2. Segment users to see which groups dropped most; 3. Examine changes in content consumption behavior; 4. Look at frequency and duration changes of product features; 5. Show structured thinking and prioritization in the answer
参考答案
I would approach from these four directions: 1. Time breakdown: Identify on which day the significant drop started, and whether it coincides with a version update, holiday, or major competitor event; Key metrics: daily trend of average usage time, time-segmented (morning/afternoon/evening) duration changes. 2. User segmentation: Slice by new vs old users, region, age, device, DAU tier (heavy/moderate/light), to see which group dropped most; Key metrics: YoY usage time change by segment, proportion change. 3. Content consumption: Whether recommendation quality declined or user interest shifted; Key metrics: average watch time per session, completion rate, like/comment/share rate changes, time spent distribution across verticals. 4. Feature usage: Whether usage of certain core features decreased (e.g., For You page, Following, Search); Key metrics: dwell time proportion change per tab, entry frequency change, drop-off rate from recommendation page to others.