“meta ads library” — who sticks, who vanishes
Best organic position per site in each monthly Semrush snapshot (US, top 20), Aug 2025 → Aug 2026. One row per site, sorted by months present. A long unbroken row = a site that sticks.
Who sticks: buffer.com all 13/13 months (never worse than #11) ·
reddit.com 11 months, parked at #3 ·
mida.so 9 months and climbing (#8 → #4) through three core updates ·
youtube.com 8 · foreplay.co 7 straight, then flushed by the Mar 2026 core ·
shopify.com 6 straight, flushed after the Dec 2025 core.
Who vanishes: each core update flushed roughly two-thirds of the third-party top 20
(Dec 2025: 12 of 19 gone · Mar 2026: 12 of 17 gone) — and 35 sites appeared for one or two months, ever.
So: third parties can hold a branded SERP for a year; most don’t survive their first core update.
1–3
4–10
11–20
Meta Meta’s own properties
· = not in top 20 that month
▮ core update
▮ spam update