Content Strategy
The Content Refresh Framework: When to Update, When to Rewrite, When to Kill
Every page on your site is either compounding or decaying. A 3-action decision model for content managers running 100+ page libraries who need to stop guessing and start triaging with data.
What Is a Content Refresh Framework?
A content refresh framework is a decision system that tells you exactly what to do with every page on your site: update it (minor fixes, keep the structure), rewrite it (new angle, new structure, same URL), or kill it (redirect or remove entirely). Every piece of content on your site falls into one of these three buckets. The framework eliminates the debate and replaces it with a scoring model.
Most content teams treat their published library like an archive. They publish, move on, and never look back unless someone complains. The result: 60-70% of pages on a typical B2B site generate zero organic traffic after 12 months, according to Ahrefs’ 2025 Content Study of 14 million pages. Those pages aren’t just sitting there harmlessly. They dilute crawl budget, confuse topical authority signals, and make your site harder for both search engines and AI models to understand.
The teams that win organic growth aren’t the ones publishing the most. They’re the ones maintaining what they’ve already published. A 2024 HubSpot analysis found that updated blog posts generated 106% more traffic than newly published posts targeting the same keywords. The old content had domain authority, backlinks, and indexing history working in its favor. It just needed better information.
This post gives you the complete framework: the signals that trigger each action, the scoring model to prioritize your backlog, and the execution process to run this as a quarterly system instead of a one-time project.
What Are the Three Actions in a Content Refresh?
Every page evaluation ends with one of three decisions. The distinctions matter because they determine the level of effort, the expected return, and the timeline for results.
Action 1: Update
An update preserves the existing structure, URL, and core argument. You’re fixing what’s stale, not rebuilding what’s broken. Updates typically take 1-3 hours per page and show ranking improvements within 2-4 weeks. What an update includes:- Replacing outdated statistics with current data (swap 2023 numbers for 2025 numbers)
- Adding 1-2 new sections to cover subtopics that competitors now address
- Fixing broken internal and external links
- Updating meta title and description to improve CTR
- Refreshing images, screenshots, or examples that reference old UI or products
Action 2: Rewrite
A rewrite keeps the URL but replaces the content substantially. The original angle, structure, or depth is no longer competitive. You need a fundamentally different approach to the topic, not just fresher data. What a rewrite includes:- New outline based on current SERP analysis and search intent
- Different content format (turning a listicle into a framework, a how-to into a decision guide)
- Expanded word count to match competitive depth (if competitors average 2,800 words and you have 900, an update won’t close that gap)
- New internal linking structure to connect with content published since the original
- Updated schema markup to match the new format
Action 3: Kill
A kill means the page no longer deserves to exist as a standalone URL. You either 301-redirect it to a stronger page on the same topic or remove it entirely with a 410 status code. When to kill:- The page targets a keyword you’ve covered better elsewhere (cannibalization)
- The topic is no longer relevant to your business or audience
- The page has zero backlinks, zero traffic, and no ranking potential even after a rewrite
- Multiple thin pages on similar subtopics should be consolidated into one comprehensive page
Which Signals Tell You What Action to Take?
Five data signals drive the decision. You don’t need all five to be conclusive. Three out of five pointing in the same direction is enough to act with confidence.
Signal 1: Traffic Trend (Last 6 Months)
Pull organic sessions from Google Search Console or GA4 for every page, comparing the most recent 6-month period to the prior 6 months. A page that dropped from 500 sessions/month to 300 is in decay. A page that dropped from 500 to 50 is in collapse. The velocity of the decline matters as much as the absolute number.- Mild decline (10-30%) = Update signal
- Steep decline (30-70%) = Rewrite signal
- Near-zero or 90%+ decline = Kill signal (unless the keyword still has strategic value)
Signal 2: Current Ranking Position
Where a page sits in rankings tells you how much effort is required to recover. Position 4-10 means the page is competitive and likely needs only an update. Position 11-30 means the page has relevance but has been outcompeted structurally, which usually means a rewrite. Position 30+ or unranked means the page has lost the fight entirely.Signal 3: Content Age and Freshness
Age alone doesn’t make content bad. A definitive guide to financial ratios can rank for 5 years with minor updates. But content that references specific tools, statistics, screenshots, or industry benchmarks has a shelf life. For most B2B content, that shelf life is 12-18 months before some element becomes visibly outdated. For tech and SaaS content, it’s 6-9 months.Signal 4: Competitive Gap
Pull the top 5 ranking pages for your target keyword. Compare their depth, format, and recency to yours. If the top results have 2,500-word guides with tables, frameworks, and embedded tools while your page is an 800-word overview from 2023, an update won’t close that gap. You need a rewrite or a kill-and-consolidate.Signal 5: Backlink Profile
Pages with 10+ referring domains have equity worth preserving. Even if the content is poor, the URL carries authority. These pages should almost never be killed. They should be rewritten to capitalize on the link equity they’ve earned. Pages with zero backlinks and no traffic have nothing to lose from a kill decision.“The biggest mistake content teams make is treating every page as equally worth saving. It’s not. Your top 20% of pages drive 85-90% of your organic traffic. The framework exists to protect that 20% and make honest decisions about the rest.”
Hardik Shah, Founder of ScaleGrowth.Digital
How Do the Signals Map to Decisions?
This table is your quick-reference cheat sheet. For each signal, it shows the threshold that points toward Update, Rewrite, or Kill. When 3 or more signals align on the same action, that’s your answer.
| Signal | Update | Rewrite | Kill |
|---|---|---|---|
| Traffic Trend | Down 10-30% over 6 months | Down 30-70% over 6 months | Down 90%+ or near-zero sessions |
| Ranking Position | Position 4-10 (striking distance) | Position 11-30 (page 2-3) | Position 30+ or not indexed |
| Content Age | 6-18 months, core is solid | 18-36 months, structure outdated | 36+ months, topic shifted entirely |
| Competitive Gap | Missing 1-2 subtopics vs. top 5 | Wrong format or half the depth | Completely outclassed, no angle |
| Backlink Profile | 5+ referring domains | 10+ referring domains (preserve URL) | 0-2 referring domains |
One override rule: if a page has 15+ referring domains, it should never be killed regardless of what the other signals say. Redirect that equity to a rewritten version on the same URL. Backlink equity takes years to build and seconds to destroy.
How Do You Score and Prioritize Hundreds of Pages?
When you’re managing 100, 500, or 2,000 pages, you can’t evaluate each one by hand every quarter. You need a scoring model that automates the triage and surfaces the highest-impact opportunities first.
Here’s the Content Refresh Score (CRS), a 0-100 model built on the five signals above. Each signal is scored on a 0-20 scale, and the total determines the recommended action and priority tier.
Scoring Each Signal (0-20 Points)
Traffic Trend Score- Traffic growing: 0 points (no action needed)
- Flat (within 10%): 5 points
- Down 10-30%: 10 points
- Down 30-70%: 15 points
- Down 70%+ or near-zero: 20 points
- Position 1-3: 0 points
- Position 4-10: 8 points
- Position 11-20: 14 points
- Position 21-50: 18 points
- Position 50+ or not ranking: 20 points
- Updated in last 6 months: 0 points
- 6-12 months: 6 points
- 12-24 months: 12 points
- 24-36 months: 16 points
- 36+ months: 20 points
- Your page matches or exceeds top 5: 0 points
- Missing 1-2 subtopics: 8 points
- Significantly less depth (50%+ word count gap): 14 points
- Wrong format for current intent: 18 points
- No competitive angle available: 20 points
- 20+ referring domains: 0 points (high preservation value)
- 10-19 referring domains: 4 points
- 5-9 referring domains: 8 points
- 1-4 referring domains: 14 points
- 0 referring domains: 20 points
Interpreting the Total Score
- CRS 0-25: No action required. Page is healthy. Review again next quarter.
- CRS 26-50: Update. Page has early decay signals. Fix what’s stale, add what’s missing, and re-index.
- CRS 51-75: Rewrite. The page needs a structural overhaul. New outline, new angle, same URL. Allocate 6-12 hours.
- CRS 76-100: Kill. The page is beyond recovery or not worth the investment. 301-redirect to the strongest related page or consolidate into a larger piece.
How Do You Prioritize When Everything Needs Attention?
The scoring model tells you what to do. The prioritization matrix tells you what to do first. On a 300-page site, you might have 45 pages flagged for updates, 20 for rewrites, and 15 for kills. You can’t do all 80 at once. You need a priority sequence.
The matrix uses two axes:
- Current organic value (traffic + conversions the page generates today)
- Recovery potential (how likely the action is to produce measurable results)
Quadrant 1: High Value, High Potential (Do First)
Pages that still drive meaningful traffic but are losing ground. These are your position 4-10 pages that dropped 15-25% in the last 6 months. An update here protects revenue you already have. A page generating 800 sessions/month that’s declining at 20% will lose 160 sessions next month if you do nothing. Updating it typically stops the bleed and recovers 50-80% of the lost traffic within 30 days.Quadrant 2: Low Value, High Potential (Do Second)
Pages that aren’t driving much traffic today but have the ingredients for recovery: strong backlink profiles, high-volume keywords, or topics central to your content strategy. These are your rewrite candidates. The page at position 18 with 12 referring domains and 4,400 monthly searches on its target keyword is a goldmine waiting for better content.Quadrant 3: High Value, Low Potential (Monitor)
Pages that drive traffic but have limited upside from changes. A page ranking #2 for a moderate-volume keyword doesn’t need a rewrite. It might benefit from a minor update to defend its position, but it shouldn’t consume resources that could go to Quadrant 1 or 2. Check it quarterly. Act only if it starts declining.Quadrant 4: Low Value, Low Potential (Kill)
Pages with no traffic, no backlinks, no ranking, and no strategic keyword. These are your kill list. Don’t agonize over them. 301-redirect to the closest relevant page and move on. Every hour spent deliberating over a page that gets 3 visits per month is an hour stolen from a page that could generate 300. For a typical 200-page B2B site, the distribution looks roughly like this: 15-20 pages in Quadrant 1, 25-35 in Quadrant 2, 30-40 in Quadrant 3, and 40-60 in Quadrant 4. That means 20-30% of most sites should be killed or consolidated. That’s not a failure of content production. It’s the natural result of publishing without a maintenance system.What Does the Quarterly Refresh Process Look Like?
Running this as a one-time project is useful. Running it as a quarterly system is transformational. Here’s the 4-phase process that turns content refresh from an occasional project into a growth engine.
Phase 1: Audit (Week 1)
- Export all indexed pages from Search Console (URL, clicks, impressions, average position for the last 6 months)
- Pull backlink data from Ahrefs or Semrush (referring domains per URL)
- Map each page to its target keyword and keyword volume
- Calculate the Content Refresh Score for every page
- Sort into the four quadrants
Phase 2: Triage (Week 2)
- Review the top 10 Quadrant 1 pages manually. Confirm the scoring model’s recommendation with a human review of each page’s content, SERP layout, and business value.
- Review Quadrant 4 kill list. Confirm no pages have sentimental or strategic value that the data doesn’t capture (e.g., a low-traffic page that’s referenced in sales decks or linked from partner sites).
- Assign each action to a writer with a brief: for updates, specify what to change; for rewrites, provide a new outline and competitive analysis; for kills, specify the redirect target.
Phase 3: Execute (Weeks 3-8)
Run the updates, rewrites, and kills according to the priority sequence. Batch by type for efficiency:- Week 3-4: Execute all kills and redirects (fastest to complete, immediate crawl budget recovery)
- Week 3-5: Execute all updates (highest ROI per hour invested)
- Week 4-8: Execute rewrites (highest effort, longest time to results)
Phase 4: Measure (Weeks 9-12)
Compare performance 30, 60, and 90 days after the refresh. The metrics that matter:- Organic sessions (before vs. after, per page)
- Average position (did the page recover, improve, or continue declining?)
- Click-through rate (did the updated meta title/description improve CTR?)
- Conversions attributed to refreshed pages (the metric your CFO cares about)
“We audited a 340-page B2B site last year. Killed 87 pages, rewrote 34, and updated 52. Total organic traffic increased 38% in 90 days. They didn’t publish a single new page. They just made the existing library work harder.”
Hardik Shah, Founder of ScaleGrowth.Digital
What Are the Most Common Content Refresh Mistakes?
At ScaleGrowth.Digital, a growth engineering firm that builds organic acquisition systems, we’ve run this process across 50+ client sites through our SEO practice. Here are the 6 mistakes that derail content refresh programs most frequently.
1. Updating When You Should Rewrite
Adding a paragraph and swapping a statistic won’t save a page that’s structurally wrong. If the search intent has shifted from informational to transactional, or the top results are all comparison tables while your page is a narrative essay, you need a new structure. An update on a structurally broken page wastes 2 hours and changes nothing.2. Refusing to Kill
Content teams develop attachment to their work. We’ve seen teams spend 8 hours rewriting a page that gets 11 visits per month with no backlinks and no conversion path. That’s 8 hours that could have updated 4 pages in the top 10, each driving 500+ sessions. The math doesn’t support sentimentality.3. Refreshing by Age Instead of Performance
Some teams set a rule: “refresh everything older than 12 months.” That’s a calendar-based system, not a performance-based system. A 3-year-old page ranking #1 with growing traffic needs nothing. A 6-month-old page that already dropped from position 5 to position 19 needs immediate attention. Always triage by signals, not by publication date.4. Ignoring Cannibalization
When two pages target the same keyword, Google splits authority between them and neither ranks well. The refresh framework should include a cannibalization check: group all pages by target keyword, identify duplicates, and consolidate. We’ve seen single consolidation moves (merging 3 thin pages into 1 comprehensive page) produce position jumps of 8-15 spots within 4 weeks.5. Skipping the Re-Index Request
You updated the content but forgot to tell Google. On large sites, Google may not recrawl an individual page for 30-60 days unless prompted. Submit the updated URL in Search Console immediately after publishing. The difference between a 5-day re-index and a 45-day re-index is 40 days of lost recovery traffic.6. No Tracking System
If you don’t record what you changed and when, you can’t measure the impact. Every refreshed page needs a log entry: URL, action taken, date, and the specific changes made. Without this, your quarterly measurement phase becomes guesswork. A simple spreadsheet with 6 columns is sufficient. Sophisticated tooling is not required.How Do You Turn This Into a Permanent System?
A one-time content audit produces a one-time result. A quarterly refresh system produces compounding returns. The difference is infrastructure.
Here’s what the infrastructure looks like for a content team managing 200-500 pages:
The Content Inventory
A master spreadsheet or database containing every indexed URL on your site with these columns: URL, target keyword, keyword volume, last refreshed date, current CRS score, assigned action, status, and owner. This is the single source of truth. If a page isn’t in the inventory, it doesn’t exist in your system.The Quarterly Calendar
- Month 1, Week 1: Pull data, calculate scores, sort into quadrants
- Month 1, Week 2: Review, triage, assign briefs
- Month 1-2: Execute kills, updates, rewrites
- Month 3: Measure 30/60-day results, feed into next cycle
The Team Allocation
For a 300-page site doing quarterly refreshes, plan for approximately 40-60 hours of content work per quarter: 15 kills (2 hours total), 25 updates (50 hours), and 5 rewrites (40 hours). That’s roughly 15-20 hours per month, or about 25% of one full-time content person’s capacity. The other 75% goes to new content production. The ratio matters. Teams that spend 100% of their capacity on new content build a library that decays faster than it grows. Teams that spend 70-75% on new content and 25-30% on maintenance build a library that compounds. After 12 months of this split, the maintenance-inclusive team will have fewer total pages but significantly more organic traffic because every page is working. Our content strategy practice builds this exact system for every client engagement. The framework, the scoring model, the inventory template, and the quarterly cadence all transfer to the client’s team by month 3. The goal is to make the system self-sustaining, not dependent on external support. If your analytics setup already feeds data into a central dashboard, connecting it to the CRS scoring model takes less than a day. The signals are the same ones your SEO reporting already tracks. The framework just gives those signals a decision layer.Your Content Library Is Either Compounding or Decaying. Which Is It?
If you’re managing 100+ pages and haven’t run a structured refresh in the last 6 months, the decay is already happening. The framework is here. The scoring model is here. The question is whether you’ll run it yourself or want a team that’s done it across 50+ sites to run it with you. Start Your Content Audit →