performance_decay

Classify web entity traffic curves into winning, losing, or crashing categories.

17|5|Updated Jun 25, 2026

Search Tags:#performance #analytics #marketing #trends #seo #decay

Ready to run

npx skills add https://github.com/portermetricsample/marketing-skills --skill performance-decay

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: performance_decay
Source: https://github.com/portermetricsample/marketing-skills/tree/main/analytics/performance_decay
Command: npx skills add https://github.com/portermetricsample/marketing-skills --skill performance-decay

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires numpy, matplotlib, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This skill solves the problem of superficial performance analysis by providing a source-agnostic engine that reads the shape of traffic curves like a human, distinguishing between normal noise and actual growth or decay.

Core Features & Use Cases

Human-like Trend Classification: Automatically labels pages, ads, or content as Winning, New, Healthy, Volatile, Losing, Crashing, or Crashed based on curve shape rather than simple percentage changes.
Hierarchical Drill-down: Allows users to navigate from high-level performance (e.g., language) down to specific topics or platforms to identify the exact levers driving traffic changes.
Use Case: Use this to proactively detect when a critical landing page or ad campaign begins to lose traction, allowing for timely intervention before the performance fully crashes.

Quick Start

Run the performance decay analysis on your Google Search Console data by executing the analyze script with the appropriate source environment variable.

Frequently Asked Questions about performance_decay

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I detect content decay in Google Search Console or Google Analytics 4 data?▼

To detect content decay, you analyze traffic curves from Google Search Console or Google Analytics 4 datasets to classify performance trends into winning, losing, or crashing categories. This process applies mathematical signal processing to distinguish normal noise from actual growth or decay patterns at scale.

What is the best way to classify web page traffic trends automatically?▼

The best way to classify web page traffic trends is using a source-agnostic engine that reads the shape of traffic curves like a human. It automatically labels pages as Winning, New, Healthy, Volatile, Losing, Crashing, or Crashed based on curve shape rather than simple percentage changes.

Do I need numpy and matplotlib to analyze performance trends?▼

Yes, you need numpy and matplotlib to analyze performance trends. Numpy is required for the mathematical signal processing used to evaluate the traffic curves, and matplotlib is necessary for the visual rendering of those performance trends.

How can I identify which specific topics are driving traffic changes at scale?▼

You can identify which specific topics are driving traffic changes by using hierarchical drill-down. This allows you to navigate from high-level performance metrics, such as language, down to specific topics or platforms to pinpoint the exact levers driving traffic changes.

Why does my landing page traffic curve show as volatile instead of losing?▼

Your landing page traffic curve shows as volatile instead of losing because the analysis engine distinguishes between normal noise and actual growth or decay. A volatile classification indicates fluctuating traffic patterns that do not exhibit the sustained downward trend required for a losing or crashing label.