You audit your content. Half your articles are 3+ years old. Many reference tools that have changed radically — “ChatGPT” articles from when the model was GPT-3.5, “Vercel deployment” articles from when their UI looked nothing like today’s, “Tailwind v2” articles when v4 is current. They still rank — some still drive traffic — but readers click, see outdated info, and bounce.
Refreshes outperform new content for SEO and trust. A 2024 article ranking position 8 + a thorough 2026 refresh becomes a position 4 article. Same URL, same backlinks, more value. The trick is doing real refreshes (Google detects superficial date-bump tricks) and prioritizing by traffic so effort goes where it matters most.
Common causes
Ordered by hit rate, highest first.
1. Editorial focus shifted to publishing new instead of refreshing old
The team feels productive when shipping new articles. Refreshing feels like “going backward.” The dashboard tracks publish count; nobody tracks refresh count.
How to spot it: Your editorial calendar has 90% new articles, 10% refreshes. Compare to an industry leader’s ratio (often 50/50 or weighted toward refresh).
2. No system to identify what needs refreshing
You have 300 articles. You can’t read all of them every quarter. There’s no automated way to surface “these specific articles are stale.”
How to spot it: When asked “which articles need refreshing?” the answer is “I don’t know without reading them.” No prioritization tooling exists.
3. Older articles still rank, so they feel “fine”
Article still on page 1 — feels OK. But it’s position 8 with 1.2% CTR when a refresh could make it position 4 with 6% CTR. “Still ranking” hides a lot of opportunity cost.
How to spot it: Search Console: filter to articles older than 18 months. Look at their average position. If many cluster in positions 6-15, they’re underperforming their potential.
4. Refreshes are perceived as “boring”
Authors prefer writing new things. Refreshes (re-screenshot, update facts, update examples) feel like maintenance. Without explicit assignment, they don’t happen.
How to spot it: Your team’s “interesting work” backlog is full; “maintenance” backlog is empty or stagnant. Refreshing went underground.
5. Fear of breaking what’s working
“This article ranks well — what if I change it and lose the ranking?” — common fear. Often wrong: substantial refreshes with same URL usually improve, not lose.
How to spot it: Articles that haven’t been edited in 24+ months with substantial traffic. Fear preserved them.
6. Date-bump tricks were tried and discounted
You ran a script to update all dateModified fields to today. CTR didn’t improve; Google noticed. Now there’s burnout on “refresh” as a tactic.
How to spot it: Many articles have dateModified newer than their last real git edit. Date-bump was applied without real edits.
Shortest path to fix
Ordered by ROI. Step 1 prioritizes; Steps 2-4 refresh correctly.
Step 1: Sort articles by traffic × age — refresh the high-value stale
For each article:
- Get monthly impressions from Search Console
- Get last modified date
- Calculate "refresh value" = impressions / months_since_last_update
Sort desc. Top 20 are your highest-value refresh candidates.A 2-year-old article with 5,000 monthly impressions has more refresh value than a 4-year-old article with 500.
Step 2: For each, do a real refresh
Real refresh means:
1. Read the article carefully — identify what's outdated
2. Update screenshots (UI changed)
3. Update version numbers / pricing (often outdated)
4. Update example links (often 404 or moved)
5. Add 1-2 new examples / data points from current context
6. Rewrite the conclusion if the recommendation has changed
7. Update internal links to point to current canonical URLsPlan 1-2 hours per refresh. If a “refresh” takes 10 minutes, you didn’t do enough.
Step 3: Update dateModified honestly
---
title: "..."
publishedAt: 2023-08-15 # preserved
lastModified: 2026-05-22 # updated to today
---Visible to readers (“Updated: May 22, 2026”) and to Google via JSON-LD. Both should match the real edit date, not a script-bumped date.
Step 4: Make the refresh visible
Add a banner at the top of refreshed articles:
> **Updated May 2026**: Refreshed for the latest UI redesign, updated pricing,
> added a new section on the v4 features.Visible refresh signals to readers + makes Google’s content-change detection easier.
Step 5: Schedule recurring refresh cycles
- Top 20 articles by traffic: refresh every 12 months
- Top 50: every 18 months
- Long tail: every 36 months or never (decide per article)
- Tools / UI articles: refresh whenever the underlying platform redesignsCalendar this. Without scheduling, refreshes never happen.
Step 6: Build a “stale check” into quarterly audits
# scripts/audit-stale.mjs
# List articles where:
# - lastModified > 18 months ago AND
# - has any monthly traffic AND
# - mentions tools/platforms in the rapidly-changing list
# Output to dashboard as the refresh queueQuarterly cadence + visible queue = systematic instead of accidental.
Prevention
- Refresh cadence by traffic tier: top 20 yearly, top 50 every 18mo
- Real refreshes only — date-bump tricks are detected and discounted by Google
- Refresh queue in your quarterly audit; visibility prevents “fear of breaking” excuse
- For tools/UI articles, refresh triggers off platform redesign events, not the calendar
- Editorial KPI: refresh count tracked alongside new-article count
- A refreshed article on the same URL usually outperforms a new article; URL equity matters
Related
Tags: #Content ops #Site quality #Site audit #Troubleshooting #Stale content