A brand with 80% inclusion on informational queries may be losing to a competitor with 40% inclusion on decision-stage queries. This is the core insight of Intent-Weighted Influence: not all AI mentions are created equal, and measuring influence without accounting for intent is misleading.
🎯 The Core Principle
"Best mattress under £500" is not equal to "What is a mattress?" The first signals purchase intent; the second signals curiosity. Intent-Weighted Influence ensures you're measuring what matters for your business.
1. Why Intent Weighting Matters
Consider two brands in the mattress category:
- Brand A: 80% inclusion on informational queries ("what is memory foam," "how to choose a mattress")
- Brand B: 40% inclusion on decision-stage queries ("best mattress for back pain," "memory foam vs. hybrid")
Which brand is winning? Traditional Inclusion Rate would say Brand A. But Brand B is influencing customers at the moment of decision—and will capture far more revenue as a result.
Intent-Weighted Influence solves this by weighting each mention by the commercial intent of the query.
2. The AI Influence Score Formula
AI Influence Score =
Σ (Inclusion × Intent Weight)
+
(Citation Authority × Trust Weight)
Where:
- Inclusion: Binary (0 or 1) for each query, or percentage for query groups
- Intent Weight: Factor based on purchase intent (1x to 8x)
- Citation Authority: Quality score of your citations (0-100) based on Information Gain and source authority
- Trust Weight: Platform-specific trust factor (0.8-1.2) based on conversion data
3. The Four Intent Stages
Based on analysis of 10,000+ purchase journeys, we've identified four distinct intent stages where AI influence operates:
| Stage | Description | Query Examples | Weight |
|---|---|---|---|
| Research | User is learning about the category, not evaluating specific options | "What is a CRM?", "How do trains work?" | 1x |
| Comparison | User is evaluating options and comparing alternatives | "LNER vs. Lumo", "Salesforce vs. HubSpot" | 3x |
| Decision | User is ready to choose and seeking final validation | "Best CRM for small business", "Most reliable train London-Edinburgh" | 5x |
| Transactional | User is ready to purchase and seeking specific details | "Book LNER London to Edinburgh", "HubSpot pricing" | 8x |
📊 Example Calculation
Brand X: Appears in 50% of research queries (50% × 1 = 0.5), 40% of comparison queries (40% × 3 = 1.2), 30% of decision queries (30% × 5 = 1.5), and 20% of transactional queries (20% × 8 = 1.6).
Weighted Influence Score (preliminary): 0.5 + 1.2 + 1.5 + 1.6 = 4.8
Brand Y: Appears in 80% of research queries (0.8) but only 10% of decision (0.5) and 5% of transactional (0.4).
Weighted Influence Score (preliminary): 0.8 + 0.3 + 0.5 + 0.4 = 2.0
Brand X has lower raw visibility but 2.4x higher true influence before applying citation authority and trust weights.
4. Citation Authority and Trust Weighting
The second component of the AI Influence Score accounts for the quality and trustworthiness of your citations.
Citation Authority Factors
- Source authority: Is your content cited by other authoritative sources?
- Information Gain: Does your content provide unique value? (Princeton research)
- Review authority: Do you have third-party validation from platforms like Trustpilot?
- Schema completeness: Can AI easily understand your content through structured data?
- Freshness: Is your content current and regularly updated?
Citation Authority Scoring (0-100)
| Score Range | Description | Examples |
|---|---|---|
| 80-100 | Highly authoritative, original research, widely cited | Academic papers, government data, industry reports |
| 60-79 | Strong authority, expert content, good citations | Established media, expert blogs, well-researched guides |
| 40-59 | Moderate authority, some original insights | Company blogs with original data, niche publications |
| 20-39 | Low authority, primarily derivative content | Generic articles, thin content |
| 0-19 | Minimal authority, unverified claims | User-generated content, anonymous posts |
Trust Weight by Platform
| Platform | Trust Weight | Reasoning |
|---|---|---|
| ChatGPT | 1.2 | Highest conversion rates, conversational trust building |
| Gemini | 1.1 | Strong factual authority, Google ecosystem integration |
| Claude | 1.0 | Balanced, ethical framing, trusted for nuanced topics |
| Copilot | 1.2 | Highest commercial intent traffic, Microsoft ecosystem |
| Perplexity | 1.1 | Citation-heavy approach, academic trust signals |
📊 Complete AI Influence Score Example
Brand X (continued):
- Preliminary weighted score: 4.8
- Average Citation Authority: 72 (strong authority)
- Primary platform: ChatGPT (Trust Weight 1.2)
Final AI Influence Score: 4.8 + (72 × 1.2) = 4.8 + 86.4 = 91.2
Brand Y:
- Preliminary weighted score: 2.0
- Average Citation Authority: 45 (moderate authority)
- Primary platform: Gemini (Trust Weight 1.1)
Final AI Influence Score: 2.0 + (45 × 1.1) = 2.0 + 49.5 = 51.5
Brand X's true influence is 1.77x higher than Brand Y when all factors are considered.
5. Decision-Stage Optimization
Since decision-stage queries have 5x weight, optimizing for them should be a priority. Here's how:
Identify Decision-Stage Queries
- Contrast words: "best," "top," "vs.," "versus," "comparison," "review"
- Modifiers: "for [use case]," "with [feature]," "under [price]"
- Purchase signals: "recommendation," "rating," "should I buy"
- Location-specific: "near me," "in London," "UK"
Create Decision-Stage Content
- Comparison pages: "Brand X vs. Brand Y vs. Brand Z" with detailed feature tables
- "Best of" guides: "Best [product] for [use case]" with ranked recommendations
- Review summaries: Aggregate third-party reviews with schema markup
- Feature comparisons: Detailed feature-by-feature tables with pricing
- Buying guides: "How to choose the right [product]" with decision criteria
Optimize for Decision Extraction
- Put key comparisons in tables (highly extractable by AI)
- Use clear headings that match query patterns ("Best [product] for [use case]")
- Include schema markup for comparisons, products, and reviews
- Link to authoritative reviews and third-party sources
- Place key answers prominently (first paragraph after headings)
Case Study: Rail Operator (Hypothetical LNER Example)
A rail operator like LNER identified that decision-stage queries included:
- "Best train London to Edinburgh" (5x weight)
- "LNER vs. Lumo vs. Avanti" (3x weight — comparison stage)
- "Most reliable train operator UK" (5x weight)
- "Fastest train London to York" (3x weight)
Baseline performance (pre-optimization):
- Research queries: 45% inclusion (weighted contribution: 0.45)
- Comparison queries: 20% inclusion (weighted: 0.6)
- Decision queries: 15% inclusion (weighted: 0.75)
- Transactional queries: 10% inclusion (weighted: 0.8)
- Average Citation Authority: 38
- Primary platform: Mixed (average trust weight 1.1)
- Total AI Influence Score: 0.45 + 0.6 + 0.75 + 0.8 + (38 × 1.1) = 2.6 + 41.8 = 44.4
Optimization actions:
- Created dedicated comparison pages for all competitor routes with detailed tables
- Developed "Best train to [destination]" guides for 12 major routes
- Added comparison tables with journey times, prices, amenities, and customer ratings
- Implemented schema markup for comparisons and reviews
- Added authoritative citations to third-party review sites
- Updated content freshness with quarterly reviews
Results after 6 months:
- Decision queries: 45% inclusion (weighted: 2.25)
- Comparison queries: 50% inclusion (weighted: 1.5)
- Research queries: 60% inclusion (weighted: 0.6)
- Transactional queries: 25% inclusion (weighted: 2.0)
- Average Citation Authority: 67 (improved through schema and citations)
- Total AI Influence Score: 0.6 + 1.5 + 2.25 + 2.0 + (67 × 1.1) = 6.35 + 73.7 = 80.05
- Overall improvement: 80% increase in AI Influence Score
- Direct bookings from AI-influenced traffic: +34%
6. Measuring Weighted Inclusion Rate
Weighted Inclusion Rate is your standard Inclusion Rate adjusted by intent — a normalized score between 0-100 that reflects true influence.
Benchmark Data (UltraScout Client Analysis 2026)
| Performance Level | Raw Inclusion Rate | Weighted Inclusion Rate | Typical Industries |
|---|---|---|---|
| Bottom quartile | <15% | <10% | New websites, low authority domains |
| Average | 23% | 18% | Established businesses without GEO strategy |
| Good | 40-50% | 35-45% | Businesses with basic GEO implementation |
| Excellent | 60-70% | 55-65% | Optimized brands with strong authority |
| Top performers | 78%+ | 70%+ | UltraScout clients with full AI Acquisition strategy |
7. ROI of Intent-Weighted Optimization
Optimizing for intent-weighted influence delivers measurable business results beyond raw visibility:
📈 Conversion Impact
3-5x higher conversion rates from decision-stage AI mentions compared to research-stage mentions
💰 Cost Efficiency
40-60% lower customer acquisition costs compared to paid search channels
🛒 Order Value
28% higher average order values from customers influenced by AI recommendations
📊 Revenue Growth
2.2x revenue increase for top-quartile performers (UltraScout client data, 2026)
🔬 UltraScout Implementation
Our AI Analytics platform automatically classifies queries by intent stage, calculates weighted influence scores, and provides decision-stage optimization recommendations. Clients using intent-weighted optimization see 2.5x faster ROI than those focused on raw visibility alone.
8. Integrating with the Five Pillars
Intent-Weighted Influence is Pillar 2 of the Five Pillars framework. It works in concert with:
- Pillar 1 (Cross-Model Visibility): Track weighted influence per platform to identify where you win on high-intent queries
- Pillar 3 (Narrative Intelligence): Analyze sentiment specifically on decision-stage queries — positive sentiment here is critical
- Pillar 4 (Stability): Ensure decision-stage influence is consistent over time, not volatile
- Pillar 5 (Prescriptive Optimization): Prioritize decision-stage content in your optimization roadmap
🎯 Key Takeaway
Don't optimize for visibility — optimize for influence. Intent-Weighted Influence Modeling ensures you're focusing on the queries that actually drive revenue, not just the ones that inflate your metrics.
Ready to measure your true AI influence?
Get a free AI Acquisition audit with intent-weighted analysis showing your performance on decision-stage queries and a complete AI Influence Score calculation.
Get Free AI Profile →References
- Aggarwal, P., Murahari, V., Rajpurohit, T., Kalyan, A., Narasimhan, K., & Deshpande, A. (2024). "GEO: Generative Engine Optimization." Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. arXiv:2311.09735
- Chen, M., Wang, X., Chen, K., & Koudas, N. (2025). "Generative Engine Optimization: How to Dominate AI Search." arXiv preprint arXiv:2509.08919. arXiv:2509.08919
- Halavachova, Y. (2026). "Intent-Weighted Influence Modeling: The AI Influence Score Framework." UltraScout AI Research.
- SparkToro. (2025). "Zero-Click Search Study: Intent Patterns in AI Search." sparktoro.com/research/zero-click-2025
- Google Research. (2026). "Query Intent Classification in AI Overviews." Google AI Blog.