Prompt
How do I build a list of prompts that trigger brand mentions?
Latest observation
To build a list of prompts that reliably trigger brand mentions, start by defining what counts as a brand mention and then generate prompts across a few common trigger patterns.
1) Define the target
Decide whether you want mentions of:
- Any brand
- Specific brands (e.g., Nike, Apple, Starbucks)
- Brand families (parent companies, subsidiaries, products)
- Implicit mentions (e.g., “the iPhone company”)
Also define the output you expect:
- Direct naming
- Comparison with brands
- Recommendation of brands
- Use-case prompts that make brand examples likely
- Questions that ask for brand alternatives
2) Use trigger categories
Create prompts that naturally elicit brand names. Good categories include:
A. Recommendation prompts
These often produce brand mentions.
- “What are the best brands for running shoes?”
- “Recommend a good laptop brand for students.”
- “Which coffee brands are popular?”
B. Comparison prompts
- “Compare Apple and Samsung phones.”
- “Is Nike better than Adidas for running?”
- “What’s the difference between Coke and Pepsi?”
C. Category-to-brand prompts
- “Name a few toothpaste brands.”
- “What are some popular car brands?”
- “List streaming service brands.”
D. Purchase-intent prompts
- “I need a budget-friendly vacuum—what brands should I consider?”
- “What brand of headphones has the best noise cancellation?”
- “Where can I buy a Toyota in my area?”
E. Brand association prompts
- “What company makes the iPhone?”
- “Which brand is known for the golden arches?”
- “What brand is the ‘Just Do It’ slogan for?”
F. Review and opinion prompts
- “Is Patagonia a good brand for outdoor gear?”
- “What do people think of Tesla?”
- “How does Dyson compare to Shark?”
G. Trend/news prompts
- “Which brands are trending in skincare this year?”
- “What brands are in the latest smartphone ads?”
- “Which fast food brands are expanding internationally?”
3) Include multiple phrasings
Brands may be mentioned more often when prompts vary in wording. Use:
- “What brands…”
- “Which companies…”
- “Name some…”
- “Can you recommend…”
- “Tell me about…”
- “Compare…”
- “What’s the best…”
4) Make prompts specific enough
Vague prompts can produce generic answers. Add a category, audience, budget, or context:
- “What are good sportswear brands for marathon training?”
- “Recommend smartphone brands for older adults.”
- “What are popular skincare brands for sensitive skin?”
5) Build a structured list
A useful list can be organized like this:
| Category | Prompt |
|---|---|
| Recommendation | “What are the best brands for noise-canceling headphones?” |
| Comparison | “Compare Sony and Bose headphones.” |
| Category listing | “Name three popular laundry detergent brands.” |
| Purchase intent | “What mattress brands are worth considering?” |
| Brand association | “Which brand uses the swoosh logo?” |
| Opinion | “Is Lululemon a good brand for workout clothes?” |
6) Expand systematically
Once you have a seed list, expand it by:
- Swapping in categories: shoes, phones, cars, food, cosmetics, tools
- Swapping in adjectives: best, cheapest, premium, durable, eco-friendly
- Swapping in audiences: students, families, professionals, beginners
- Swapping in intent: buy, compare, review, recommend, identify
7) Test and refine
Run the prompts through your model or dataset and check:
- Did it mention a brand?
- Was the mention direct or indirect?
- Did it mention the right brand?
- Was the response consistent across paraphrases?
Then keep prompts that reliably trigger brand mentions.
Example starter list
- “What are the best brands for wireless earbuds?”
- “Compare Nike and Adidas shoes.”
- “Name some trusted car brands.”
- “Which brand makes the best budget laptops?”
- “What company is behind the iPhone?”
- “Recommend skincare brands for dry skin.”
- “What are popular soda brands?”
- “Which streaming service brands should I consider?”
- “Is Dyson worth the price?”
- “What brand is known for ‘Finger Lickin’ Good’?”
If you want, I can also help you build:
- a large prompt bank by category,
- a CSV/JSON template for testing, or
- brand-mention prompts tuned for a specific industry.