Retention based on Registration Time

Does retention across .eth length tiers change dependent on registration date?

Two views of the same question - how registration date influences 12 month retention.

What this tells you: Is ENS subject to speculative cycles, and does retention rebound?

Methodology

12m retention = current domains.expiry_timestamp >= first_registered_ts + 12 × 30 days.
Approximate - doesn't distinguish original holders still renewing from lapsed-and-re-registered names.
Bubble scatter (top) plots every cohort-month older than 12 months. Dot size scales as sqrt(count / max-count) × 12 + 2px floor - area ∝ count, but the floor keeps tiny cohorts hoverable.
Per-bucket panels (bottom) aggregate cohorts to year-level for a smoother trend. Year cohorts younger than 12 months are excluded.

All buckets overlaid (monthly granularity)