The Discovery Prompt Pattern
Find a restaurant for [occasion] in [location].
Party: [details]. Budget: $[X]/person [including/excluding drinks].
Dietary: [restrictions].
Vibe: [atmosphere preference].
Cuisine: [preference or "open to suggestions"].
Logistics: [parking/transit/walking distance from X].
[Any other constraints: outdoor seating, kid-friendly, private dining room, etc.]
Suggest 3 restaurants ranked by fit. For each include:
- Name, cuisine, price range
- Why it fits this specific group/occasion
- What to order (2-3 standout dishes)
- Reservation strategy (how far in advance, best time)
- One potential drawback to be aware of
Pro Tips for Restaurant Discovery
Describe the feeling, not just the cuisine. "I want to feel like I'm eating in someone's grandmother's kitchen in Tuscany" produces better results than "Italian restaurant." AI can translate emotional descriptions into concrete restaurant attributes.
Ask for the anti-recommendation. "What should I NOT order at this restaurant?" or "What type of diner would NOT enjoy this place?" reveals honest limitations that pure recommendations gloss over.
Use comparative framing. "The last restaurant I loved was [name] because of [reasons]. Find me something with that same energy but different cuisine." This gives AI a taste profile to work with rather than abstract preferences.
Request the contrarian pick. "Also suggest one restaurant I'd probably never find on Yelp — a local favorite that doesn't market well online but has incredible food." AI can synthesize the difference between rating-volume-optimized restaurants and genuine hidden gems.