metehan.ai Research & SOPs

Appearance

Query Fan-Out Analysis ​

Fresh

This content reflects Metehan Alp's Query Fan-Out methodology as of 2025-2026.

Query Fan-Out Analysis maps how AI search engines decompose a single user query into multiple sub-queries. Understanding this decomposition is the foundation for RRF-based optimization -- you cannot optimize for sub-queries you have not identified.

Core Concept ​

When a user asks Google AI Mode (or any AI search engine) a question, the system does not run a single search. It breaks the query into multiple sub-queries, retrieves results for each, and fuses the ranked lists using RRF or similar fusion methods.

Example: A user queries "best CRM for small consulting firms."

AI Mode might decompose this into:

  • "CRM software features for consulting"
  • "small business CRM comparison 2025"
  • "consulting firm client management tools"
  • "CRM pricing for small teams"
  • "consulting workflow CRM integration"

Each sub-query returns its own ranked list. Your content needs to appear across enough of these lists to cross the citation threshold.

Screaming Frog Integration ​

Metehan built a Screaming Frog custom script that automates query fan-out analysis. The script integrates with Screaming Frog's Custom JavaScript feature to extract and analyze sub-query patterns.

Setup ​

  1. Open Screaming Frog
  2. Navigate to Configuration > Custom > JavaScript
  3. Import the Query Fan-Out script
  4. Configure your target queries
  5. Run the crawl against Google AI Mode results

5 Key Script Functions ​

1. Query Decomposition Extractor ​

Captures the sub-queries generated by AI Mode for a given input query. Outputs a structured list of all sub-queries with their relationship type (related, implicit, comparative, procedural, contextual).

2. SERP Position Mapper ​

For each sub-query, maps which URLs appear in the results and at what position. Creates a position matrix: URLs on the Y-axis, sub-queries on the X-axis, positions in the cells.

3. RRF Score Calculator ​

Applies the RRF formula to the position matrix. Computes the fused score for each URL. Flags URLs that meet or exceed the citation threshold (tau = 0.020).

4. Coverage Gap Analyzer ​

Identifies sub-queries where your domain does not rank at all. These are the gaps that prevent your content from reaching the citation threshold. Prioritizes gaps by the weight of the sub-query.

5. Competitor Overlap Scanner ​

Maps which competitor domains appear across multiple sub-queries. Identifies competitors that are already meeting RRF thresholds and the specific sub-queries they dominate.

Implementation Steps ​

  1. Select target queries -- Start with 5-10 high-value queries in your niche
  2. Run the Screaming Frog script -- Extract sub-queries and SERP positions for each
  3. Build the position matrix -- Map your domain's presence across all sub-queries
  4. Calculate RRF scores -- Determine which pages meet the citation threshold
  5. Identify coverage gaps -- Flag sub-queries where you have no presence
  6. Prioritize optimization -- Focus on sub-queries where small ranking improvements push you over the threshold
  7. Create or optimize content -- Build pages that address coverage gaps
  8. Re-run analysis -- Validate improvements after content changes

Real-World Case Study Results ​

Metehan tested the Query Fan-Out methodology across three content categories:

Technical SEO Content ​

  • Primary query: "technical SEO audit checklist"
  • Sub-queries generated: 12
  • Initial coverage: 5 of 12 sub-queries (42%)
  • Post-optimization coverage: 10 of 12 sub-queries (83%)
  • Result: Content crossed citation threshold after addressing 5 gap sub-queries

E-commerce Content ​

  • Primary query: "best headphones for working from home"
  • Sub-queries generated: 9
  • Initial coverage: 2 of 9 sub-queries (22%)
  • Post-optimization coverage: 6 of 9 sub-queries (67%)
  • Result: Coverage improved significantly but citation threshold required additional supporting content

Local Service Content ​

  • Primary query: "emergency plumber [city]"
  • Sub-queries generated: 6
  • Initial coverage: 3 of 6 sub-queries (50%)
  • Post-optimization coverage: 5 of 6 sub-queries (83%)
  • Result: Local content reached citation threshold faster due to fewer sub-queries and less competition

Optimization Framework ​

Sub-queries fall into five categories. Each requires a different optimization approach:

CategoryDescriptionOptimization
RelatedDirectly related to the primary queryCover these in your main content piece
ImplicitAssumed context (user intent not explicitly stated)Add sections that address unstated assumptions
Comparative"vs" queries, alternatives, comparisonsCreate comparison sections or dedicated comparison pages
Procedural"How to" variations of the topicAdd step-by-step sections or linked how-to guides
ContextualBackground, prerequisites, related conceptsBuild supporting content pages that link to the main piece

GitHub Repository ​

The full Screaming Frog script with documentation is available on GitHub:

github.com/nicche/query-fan-out-analysis

Visual Version Available

A Visual Query Fan-Out script is also available that generates graphical representations of sub-query relationships. See Screaming Frog AI Scripts for details.

Further Reading ​