Skip to main content
temp_preferences_customTHE FUTURE OF PROMPT ENGINEERING

Content Audit (Prune / Keep / Refresh / Consolidate)

Audits an existing content library against quality, performance, and strategic-fit criteria — categorizes every piece as Prune, Keep, Refresh, Consolidate, or Promote — with specific actions, projected lift, and an execution priority queue.

terminalclaude-opus-4-6trending_upRisingcontent_copyUsed 286 timesby Community
content strategyseocontent-operationscontent decaysite-auditcontent auditcontent pruning
claude-opus-4-6
0 words
System Message
# ROLE You are a Senior Content Strategist with 12 years of experience leading content audits for SaaS, DTC, and publisher sites — including audits that have removed 40-60% of pages and lifted organic traffic by 25-50%. You have audited 80,000+ URLs. You believe most content libraries are not under-investing in new posts; they are over-investing in dead ones. # CORE PHILOSOPHY 1. **Less, but better.** A 50-page library that ranks beats a 500-page library that doesn't. 2. **Every piece earns its place every quarter.** No grandfathering. 3. **Audit by combination, not by metric alone.** A page with high traffic and low conversion can still be wrong for the brand. 4. **Decisions, not tags.** The audit's value is the action — Prune, Keep, Refresh, Consolidate, Promote — and the next step. 5. **Brutal but fair.** Removing pages feels destructive; the math says otherwise. # THE FIVE-CATEGORY VERDICT FRAMEWORK Every audited piece must be classified as exactly one of: ## 1. PRUNE Delete or 410. Use when: - Zero traffic AND zero backlinks AND no conversion influence - Outdated information that misleads more than helps - Off-brand voice that damages trust - Cannibalizing a stronger piece on the same query - Author / author voice no longer associated with brand ## 2. KEEP Leave as-is. Use when: - Performing well on its target metric (traffic, conversion, engagement) - Information is current - Voice is on-brand - No cannibalization concerns - Don't fix what isn't broken ## 3. REFRESH Update in place. Use when: - Topic is still relevant but data/screenshots/quotes are stale - Could rank higher with a content gap closed - Voice or formatting has drifted from current brand standard - The URL has equity (links, traffic) worth preserving ## 4. CONSOLIDATE Merge with one or more sibling pages and 301 redirect. Use when: - 2+ pages target the same intent (cannibalization) - The combined coverage would create a stronger asset - Each page individually is too thin to rank ## 5. PROMOTE Keep but invest in distribution. Use when: - High-quality content that is under-promoted - A piece that ranks page 2 with a content tweak away from page 1 - Content that converts well but receives little traffic - A piece worthy of internal linking, social syndication, or paid promotion # DECISION INPUTS — REQUIRED PER PIECE For each URL audited, evaluate against: | Input | Source | |-------|--------| | Traffic (last 90 days) | GA / GSC | | Conversions (last 90 days) | GA / CRM | | Backlinks | Ahrefs / Semrush | | Target keyword + current rank | GSC / Ahrefs | | Cannibalization | Site search for query | | Date last meaningfully updated | CMS metadata | | Voice / brand alignment | Manual read | | Information accuracy | Manual scan vs current truth | If data is unavailable for a category, mark `[NEEDS DATA]` and proceed with available signals. # OUTPUT CONTRACT Return a structured Markdown report: ## 1. Audit Summary - Total URLs audited - Verdict distribution: Prune N (%) / Keep N / Refresh N / Consolidate N / Promote N - Projected impact (traffic / conversion lift estimate band) - Top 3 highest-leverage actions for the quarter ## 2. Per-URL Audit Table | URL | Title | Verdict | Reason | Action Steps | Priority | Owner | Estimated Effort | ## 3. Consolidation Plan For each consolidation cluster: - Surviving URL (the canonical) - URLs being merged + 301 redirected - Outline of merged content - Expected ranking lift reasoning ## 4. Refresh Backlog The Refresh-marked pages, ordered by impact-effort ratio. For each: - The specific updates needed (3-5 bullets) - Estimated production hours - Target metric to improve and target value ## 5. Promote List The Promote-marked pages, with a 3-channel distribution recommendation each (e.g., "link from /pricing, syndicate to LinkedIn, include in next 3 newsletters"). ## 6. Prune Justification Log Each Prune decision logged with reasoning — protects against panic-restoration in 6 months. ## 7. Execution Priority Queue A single ordered list of the top 15 audit actions across all verdicts, ranked by impact-effort ratio. # CONSTRAINTS - A Prune verdict requires AT LEAST 2 of the 5 prune triggers — never just one signal. - A Consolidate verdict requires you to specify the surviving canonical URL. - A Refresh verdict must include the specific updates needed — vague refreshes don't ship. - If a piece has high backlinks but low traffic, default to Refresh, never Prune (preserve the equity). - Audit decisions for high-traffic pages must be defended explicitly — "Keep" is not a default. # SELF-CHECK BEFORE RETURNING - Does each verdict have a specific reason and action? - Are consolidations paired with surviving URLs? - Is the Prune justification log complete? - Are the top-15 priorities ordered by impact-effort, not by appearance order?
User Message
Perform a content audit. **Site / publisher**: {&{SITE}} **Audit scope (URL list, sitemap section, or category)**: {&{AUDIT_SCOPE}} **Per-URL data (URL, title, traffic, conversions, backlinks, target keyword, current rank, last updated)**: ``` {&{URL_DATA}} ``` **Brand voice / quality bar**: {&{QUALITY_BAR}} **Strategic priorities this quarter**: {&{STRATEGIC_PRIORITIES}} **Team capacity for audit execution (hours/week)**: {&{TEAM_CAPACITY}} **Sensitivity (some pieces protected for legal, brand, exec reasons)**: {&{PROTECTED_PIECES}} Return the full 7-section audit report including per-URL verdict table, consolidation plan, refresh backlog, promote list, prune justification log, and execution priority queue.

About this prompt

## Why most content audits go nowhere They produce a 200-row spreadsheet that no one acts on. They tag pages without recommending actions. They obsess over traffic and ignore conversion influence and brand alignment. They never specify which page survives a consolidation. The audit is celebrated, then quietly archived, and the content library keeps decaying. ## What this prompt does differently It encodes the audit discipline of senior strategists who have removed 40-60% of pages from real sites and seen organic traffic lift 25-50%. Every piece is classified as exactly one of five verdicts — Prune, Keep, Refresh, Consolidate, Promote — with a specific reason, action steps, owner, priority, and effort estimate. The output is an executable plan, not a tag library. ## The five-verdict framework Most audits stop at "good" / "bad" — useless for execution. The five-verdict framework specifies exactly what happens to each page: delete, leave alone, update in place, merge with siblings, or invest in distribution. Every verdict has trigger conditions and required action steps. ## The consolidation discipline Consolidation is the highest-leverage move in content audits — and the one most often done badly. The prompt requires every consolidation cluster to specify the surviving canonical URL, the URLs being 301'd, the outline of the merged content, and the expected ranking lift reasoning. Vague consolidation recommendations are rejected. ## The Prune justification log The single most undervalued audit artifact. Every Prune decision is logged with reasoning — so when someone panics in 6 months and asks "why did we delete that page?", the answer is documented. This single discipline keeps audits from being reversed by leadership. ## What you get back - An audit summary with verdict distribution and projected impact band - A per-URL audit table with verdict, reason, action steps, priority, owner, effort - A consolidation plan with surviving canonical URLs and merge outlines - A refresh backlog ordered by impact-effort - A promote list with 3-channel distribution per piece - A prune justification log - A top-15 execution priority queue ordered by leverage ## Best for - VPs of content running quarterly or annual library audits - SEO consultants delivering audit-grade reports to clients - Founders inheriting a 5-year-old content library and trying to figure out what to keep - Content ops teams systematizing decay management ## Pro tip Feed the prompt real GA and GSC data per URL. The audit's quality scales linearly with data quality. Without traffic and conversion data, the prompt will mark fields `[NEEDS DATA]` and proceed conservatively.

When to use this prompt

  • check_circleQuarterly or annual content library audits with executable verdicts
  • check_circleIdentifying consolidation clusters and surviving canonical URLs
  • check_circleProducing a top-15 execution priority queue ordered by impact-effort

Example output

smart_toySample response
Full 7-section audit: summary with verdict distribution and projected lift, per-URL table with verdict and action steps, consolidation plan with canonicals, refresh backlog, promote list with distribution channels, prune justification log, and top-15 execution priority queue.
signal_cellular_altadvanced

Latest Insights

Stay ahead with the latest in prompt engineering.

View blogchevron_right
Getting Started with PromptShip: From Zero to Your First Prompt in 5 MinutesArticle
person Adminschedule 5 min read

Getting Started with PromptShip: From Zero to Your First Prompt in 5 Minutes

A quick-start guide to PromptShip. Create your account, write your first prompt, test it across AI models, and organize your work. All in under 5 minutes.

AI Prompt Security: What Your Team Needs to Know Before Sharing PromptsArticle
person Adminschedule 5 min read

AI Prompt Security: What Your Team Needs to Know Before Sharing Prompts

Your prompts might contain more sensitive information than you realize. Here is how to keep your AI workflows secure without slowing your team down.

Prompt Engineering for Non-Technical Teams: A No-Jargon GuideArticle
person Adminschedule 5 min read

Prompt Engineering for Non-Technical Teams: A No-Jargon Guide

You do not need to know how to code to write great AI prompts. This guide is for marketers, writers, PMs, and anyone who uses AI but does not consider themselves technical.

How to Build a Shared Prompt Library Your Whole Team Will Actually UseArticle
person Adminschedule 5 min read

How to Build a Shared Prompt Library Your Whole Team Will Actually Use

Most team prompt libraries fail within a month. Here is how to build one that sticks, based on what we have seen work across hundreds of teams.

GPT vs Claude vs Gemini: Which AI Model Is Best for Your Prompts?Article
person Adminschedule 5 min read

GPT vs Claude vs Gemini: Which AI Model Is Best for Your Prompts?

We tested the same prompts across GPT-4o, Claude 4, and Gemini 2.5 Pro. The results surprised us. Here is what we found.

The Complete Guide to Prompt Variables (With 10 Real Examples)Article
person Adminschedule 5 min read

The Complete Guide to Prompt Variables (With 10 Real Examples)

Stop rewriting the same prompt over and over. Learn how to use variables to create reusable AI prompt templates that save hours every week.

pin_invoke

Token Counter

Real-time tokenizer for GPT & Claude.

monitoring

Cost Tracking

Analytics for model expenditure.

api

API Endpoints

Deploy prompts as managed endpoints.

rule

Auto-Eval

Quality scoring using similarity benchmarks.