LLM Visibility
Track and optimize how your products appear in LLM shopping assistant responses.
Overview
LLM Visibility tracks how your products are represented by AI shopping assistants — tools like ChatGPT Shopping, Google Gemini, and Perplexity AI that increasingly influence purchasing decisions. When a shopper asks "what's the best blue widget under €20?", NexusCommerce measures whether your product appears, where it ranks, how it is described, and whether the description is accurate.
As LLM-mediated commerce grows, LLM visibility becomes a new optimization frontier alongside SEO and Buy Box. NexusCommerce's llm_visibility_score worker is the measurement and optimization tool for this channel.
Key Concepts
LLM Query — A shopping-intent question submitted to an LLM assistant. NexusCommerce generates queries based on your product's category, key attributes, and competitive landscape. Example query: "What are the best portable Bluetooth speakers under €50 in Germany?"
Visibility Score — A composite score (0–100) measuring how prominently your products appear in LLM responses:
- 0 — Product not mentioned in any tested queries
- 50 — Product mentioned in some queries, not always prominently
- 100 — Product mentioned first or prominently in all relevant queries
Citation — An instance where an LLM assistant mentions or recommends your product. Citations include the query, the full LLM response, your product's position in the response, and the description used.
Accuracy Score — A score measuring how accurately the LLM describes your product. Inaccurate descriptions (wrong price, outdated features, incorrect availability) are flagged for attention.
Competitor Gap — Queries where a competitor's product appears but yours does not. These represent visibility opportunities.
Getting Started
Navigate to LLM Visibility under Advanced in the left sidebar.
Configure LLM Targets
- Click Configure Queries
- Select the LLM assistants to monitor (ChatGPT, Gemini, Perplexity, or all)
- Select the SKUs to monitor
- NexusCommerce auto-generates queries based on product attributes, or add custom queries
- Set the scan frequency
- Click Save
Run Your First Scan
Click Run Scan Now to trigger an immediate visibility scan. The llm_visibility_score worker submits generated queries to each configured LLM assistant and processes the responses. Results are available within 5–15 minutes depending on the number of SKUs and LLMs being scanned.
Features
Visibility Dashboard
Portfolio Overview:
- Average visibility score across all monitored SKUs
- Visibility trend (30-day chart)
- Breakdown by LLM assistant (ChatGPT vs. Gemini vs. Perplexity)
- Breakdown by product category
Top Performers:
- SKUs with highest visibility scores
- SKUs with most citations in the scan period
Attention Required:
- SKUs with visibility score below threshold
- SKUs with declining visibility trend
- SKUs with accuracy issues in LLM responses
Per-SKU Visibility Detail
Click any SKU to see its full visibility profile:
Score Card:
- Composite visibility score
- Per-LLM scores (ChatGPT score, Gemini score, Perplexity score)
- Score trend (90 days)
- Accuracy score
Citations: A list of all citations for this product in the scan period:
| Column | Description |
|---|---|
| Query | The question submitted to the LLM |
| LLM | Which assistant responded |
| Position | Where in the response your product appeared (1st, 2nd, etc.) |
| Sentiment | How the LLM described the product (positive / neutral / negative) |
| Accurate | Whether the description matches your current listing |
| Date | When the citation was recorded |
Click a citation to see the full LLM response.
Competitor Gap Analysis: Queries where a competitor appeared but your product did not. Clicking a gap shows the competitor citation and the LLM's description of the competitor, which often explains why your product was not chosen.
Accuracy Issues
When the LLM mentions your product with inaccurate information, NexusCommerce flags it:
- Price inaccuracy — LLM states a price that is different from your current price
- Feature inaccuracy — LLM mentions a feature your product does not have, or misses a key feature
- Availability inaccuracy — LLM says the product is unavailable when it is in stock
- Brand inaccuracy — LLM attributes the product to the wrong brand
Accuracy issues are tracked over time. As LLMs update their knowledge bases, inaccuracies often self-correct — but NexusCommerce can accelerate this by improving the product's data footprint on the web (via better marketplace listings, manufacturer websites, and press coverage).
Query Optimization
Based on visibility scan results, NexusCommerce recommends query-specific optimizations:
Listing Optimization — Adjust your listing copy to include the exact phrases that LLMs associate with high-relevance recommendations in your category.
Attribute Completion — LLMs often mention specific attributes (e.g., "IP67 waterproof rating") in their responses. Ensure all relevant attributes are in your listings and product pages.
Price Positioning — LLMs frequently group products by price tier. Visibility improves when your product is clearly positioned within a common price tier (e.g., "under €30" rather than €29.50).
LLM Visibility Scores Over Time
Track visibility score trends per SKU, per LLM, and per query category. Use this to measure the impact of:
- Listing changes
- New marketplace connections (more indexed pages = more LLM citations)
- Review volume increases (LLMs often prefer products with many positive reviews)
- Press coverage or external mentions
Scheduled Scans
Configure automated weekly or monthly scans via AI Studio:
Trigger: Scheduled (Monday 5 AM)
Step 1: llm_visibility_score (skus: "all_monitored", llms: "all")
Step 2: notify_email (
recipients: ["[email protected]"],
subject: "Weekly LLM Visibility Report",
template: "llm_visibility_report"
)Configuration
| Setting | Description | Default |
|---|---|---|
| LLM assistants | Which LLMs to query (ChatGPT, Gemini, Perplexity) | All enabled |
| Scan frequency | How often to run visibility scans | Weekly |
| Query count per SKU | Number of queries per SKU per scan | 5 |
| Visibility threshold | Score below which a SKU is flagged for attention | 40 |
| Accuracy alert | Alert when accuracy score drops below threshold | 70 |
| Query languages | Languages to generate queries in | Marketplace languages |