Prompt
I need to measure AI mentions across product lines
Latest observation
Oct 26, 2026
OpenAI APIWeb search: off
Sure — to measure AI mentions across product lines, the key is to define what counts as a mention, where you’re measuring it, and how you want to compare products.
A practical way to do it
1) Define the product lines
Make a list of the product lines you want to track, for example:
- Product A
- Product B
- Product C
2) Define what counts as an “AI mention”
Decide whether you want to count:
- Exact keyword matches:
AI,artificial intelligence - Related terms:
machine learning,genAI,LLM,copilot,automation - Product-specific AI features: e.g. “smart assistant,” “recommendation engine”
A good approach is to create a keyword dictionary with:
- Core AI terms
- Synonyms
- Brand/product-specific AI phrases
3) Choose your data sources
Common sources:
- Website pages
- Product documentation
- Press releases
- Sales collateral
- Support articles
- Social posts
- Call transcripts
- Reviews / app store listings
4) Count mentions by product line
You can measure:
- Raw count: number of AI mentions per product line
- Normalized count: mentions per 1,000 words or per document
- Share of voice: % of all AI mentions attributed to each product line
- Trend over time: mentions by month/quarter
5) Use a simple scoring method
For each document:
- Detect product line
- Detect AI mentions
- Assign the mention to the relevant product line
- Aggregate totals
Example table:
| Product Line | AI Mentions | Docs Analyzed | Mentions per Doc |
|---|---|---|---|
| Product A | 42 | 10 | 4.2 |
| Product B | 18 | 8 | 2.25 |
| Product C | 73 | 12 | 6.08 |
6) Watch out for ambiguity
Some mentions may not refer to your product specifically. For example:
- “AI-powered” may be generic marketing language
- “AI” may appear in competitor comparisons
- “Machine learning” may be part of an unrelated technical article
You may need rules like:
- Count only when AI terms appear within X words of product name
- Count only in product-specific documents
- Exclude competitor/industry articles
If you want to automate it
You can use:
- Spreadsheet formulas for a small dataset
- Python/R for scalable analysis
- NLP/LLM tagging for more accurate classification
A common workflow:
- Collect text
- Tag product line
- Search for AI terms
- Deduplicate repeated mentions
- Aggregate results
If you want, I can help you build one of these:
- A keyword list for AI mentions
- A spreadsheet template to track them
- A Python script to automate counting
- A dashboard metric definition for reporting
If you share your product lines and data source, I can propose a concrete measurement method.