---
name: "social_archive_strategy"
description: "Runs a quarterly archive audit on old posts and pages: inventory the library, rank each URL by decay and value, assign keep / refresh / merge / retire, and retire content with redirect mechanics that preserve link equity. Use when a content library needs a scheduled cleanup, evergreen refreshes, pruning dead pages, or consolidating overlapping posts."
---

# Social Archive Strategy

## Purpose

Turn a pile of old posts into a quarterly decision list: evergreen winners get refreshed before they finish decaying, dead weight gets retired, and no link equity leaks out in the process. The audit is the product, not a report that sits in a folder.

## Workflow

1. **Build the inventory.** Export every published URL with: publish date, last updated, 90-day sessions, traffic trend vs the prior quarter (up/flat/down), referring domains, internal inlink count, target keyword or intent. Start from `assets/content-audit-tracker.csv` as the column template. Sources: analytics platform, Search Console, and a backlink/crawl tool.

2. **Rank the decay.** Run `bin/triage.py` against the tracker. It scores each URL so decaying high-value pages float to the top: pages with real traffic and links but a declining trend rank highest; stable performers rank next; dead pages with no links rank last. The score only ranks, it never decides. See `references/decision-matrix.md` for what each score band means.

3. **Assign one disposition per URL.** Walk the ranked list and assign keep, refresh, merge, or retire using the decision matrix. Work top-down: refreshed winners are where the return is. HubSpot's own analysis found roughly three-quarters of their blog views and over 90% of their leads came from old posts, and refreshed posts roughly doubled their traffic. Do the high-return work first.

4. **Execute the dispositions.**
   - **Refresh:** update facts, stats, examples, screenshots, and product references; add the sections the current top results cover that yours lacks; fix broken links; answer the follow-up questions readers now ask. Republish with a real changelog. A date swap with no new substance is not a refresh.
   - **Merge:** move any unique content into the survivor page, then point a permanent server-side redirect (301 or 308) from each retired URL to the survivor, one hop, no chains. Update internal links and the sitemap to the survivor.
   - **Retire:** if the URL has backlinks or traffic, 301/308 it to the closest living page that serves the same reader. If it has neither, return 410 (or 404). Only keep a dead page live with `noindex` when humans still need to reach it.

5. **Record the baseline.** For every URL you touched, snapshot traffic, position, and backlinks before the change, and set a review date for next quarter. The next audit starts from these numbers, not from memory.

## Output Contract

A finished quarterly audit delivers:
- The filled tracker: every URL with a disposition, and every refresh/merge/retire item with a named owner and deadline.
- A refresh list ordered by the triage score, each item naming the specific gaps the refresh must close.
- A merge/retire list with exact redirect targets or status codes, plus a pre-change traffic and backlink snapshot per URL.
- A next-review date on the calendar. An audit without a follow-up date is a one-off cleanup, not a strategy.

## Operating Rules

1. One URL, one disposition, one owner, one deadline. An audit that assigns no owners is shelfware.
2. Refresh substance, never timestamps. If you cannot name what got better, the refresh is not done.
3. Never redirect every retired URL to the homepage. Each one goes to the closest living page that serves the same reader, or gets a 410.
4. Keep redirects live at least 12 months, and update internal links and the sitemap to point at the survivors.
5. Do not use robots.txt to hide pages you want consolidated (it blocks crawlers from seeing the noindex or canonical), and do not use the Search Console Removals tool as a permanent fix (requests last about six months).
6. Do not use `rel=canonical` to point an old page at a newer, different page. Canonical means the pages are equivalent; a redirect says one replaced the other. Google reads them differently.
7. Every factual claim in the audit report must trace to the inventory data or a cited source. Unverifiable claims get cut or framed as opinion.