Interpret a cohort retention table and surface key insights

You are a product analytics expert. I will paste a cohort retention table and I need you to extract the most important insights from it.

Retention table (paste raw data or describe it): {{RETENTION_TABLE}}

Context:
- Product type: {{PRODUCT_TYPE}}
- Cohort definition (e.g., users who signed up in a given week/month): {{COHORT_DEFINITION}}
- Retention metric being tracked (e.g., DAU, feature usage, purchase): {{RETENTION_METRIC}}

Follow these steps:
1. Identify the overall retention curve shape (e.g., steep early drop, flattens at a floor, gradual decline) and what it typically signals for this product type.
2. Compare the three best-performing cohorts to the three worst-performing cohorts. List the specific percentage-point differences at Day 7 and Day 30 (or the closest intervals available).
3. Flag any cohort that shows an unusual pattern — either a sudden improvement or a sudden drop — and name a plausible cause.
4. State the approximate long-run retention floor if the data shows one, or note if the data does not extend far enough to determine it.
5. Suggest two specific hypotheses worth testing to improve retention at the biggest drop-off point.

Note: This prompt works best when you have at least 6 cohorts and retention data through Day 30. Shorter windows will produce less reliable trend analysis.