Quality · How good is it?
Core product/service excellence. High ratings + positive review language push this up; sparse or poor ratings pull it down.
Signals: Star rating (1–5), review volume trust, review sentiment, category-language boost
Methodology
How scoring works
Every business gets an overall score from 0–100 and six dimension scores. The math is deterministic from place data — same inputs always produce the same scores. No paid boosts, no secret hand-tuning per brand.
Demo data vs live Places
- Without
GOOGLE_PLACES_API_KEY: search uses the seeded demo catalog. Footers and API responses saymode: "demo". - With the key set (e.g. on Vercel): live Google Places fields feed the same scorer. Mode becomes
"places". - The scoring formulas do not change between modes — only the source of rating, review count, price, status, and review text does.
Overall labels
Overall is a category-weighted blend of the six dimensions, then banded:
Bands
Elite
85–100
Top-tier vibe — standout across weighted dimensions.
Strong
70–84
Clearly above average. Safe bet for most people.
Mixed
50–69
Solid in places, weaker in others. Check the breakdown.
Risky
0–49
Below average overall — weak dimensions are dragging it.
Six dimensions
Quality · How good is it?
Core product/service excellence. High ratings + positive review language push this up; sparse or poor ratings pull it down.
Popularity · How well-known?
Social proof from review volume and rating lift. Log-scaled so mega-chains don't automatic-win over strong smaller places.
Value · Worth the price?
Quality relative to price tier. Cheap + high rating = great value; luxury categories (hotel/services) get a softer price penalty.
Service · Staff & experience
People and process. Review sentiment is the main driver; retail, health, and services weight this higher in overall.
Atmosphere · Feel of the place
Vibe / ambiance inferred from review tone, open energy, and price-tier cues. Coffee, hotels, and food lean on this more.
Reliability · Can you count on it?
Operational trust — open status, review depth, rating consistency. Gyms and health weight reliability higher.
- Elite (85–100)— Top-tier vibe — standout across weighted dimensions.
- Strong (70–84)— Clearly above average. Safe bet for most people.
- Mixed (50–69)— Solid in places, weaker in others. Check the breakdown.
- Risky (0–49)— Below average overall — weak dimensions are dragging it.
Six dimensions
Each dimension is scored 0–100 from the signals below, then clamped.
Popularity · How well-known?
Social proof from review volume and rating lift. Log-scaled so mega-chains don't automatic-win over strong smaller places.
Signals: Review count (log scale), rating lift, soft penalty at extreme volume
Value · Worth the price?
Quality relative to price tier. Cheap + high rating = great value; luxury categories (hotel/services) get a softer price penalty.
Signals: Rating vs price level, cheap/mid bonuses, gym premium when expensive but excellent
Service · Staff & experience
People and process. Review sentiment is the main driver; retail, health, and services weight this higher in overall.
Signals: Rating + sentiment blend, category-language hits, open-now weak signal
Atmosphere · Feel of the place
Vibe / ambiance inferred from review tone, open energy, and price-tier cues. Coffee, hotels, and food lean on this more.
Signals: Review sentiment, quality base, open-now, mid-price ambiance, category boost
Reliability · Can you count on it?
Operational trust — open status, review depth, rating consistency. Gyms and health weight reliability higher.
Signals: Business status, review volume, rating band, open-now, sparse-review penalty
Category-aware weights
Overall = weighted sum of dimensions. Coffee leans quality; gyms lean reliability; retail and services lean service. Weights sum to ~1.0.
| Category | qua | pop | val | ser | atm | rel |
|---|---|---|---|---|---|---|
| Coffee | 30% | 10% | 18% | 16% | 16% | 10% |
| Gym | 22% | 10% | 14% | 18% | 12% | 24% |
| Food | 32% | 12% | 18% | 16% | 12% | 10% |
| Retail | 22% | 12% | 18% | 28% | 10% | 10% |
| Health | 30% | 6% | 12% | 24% | 8% | 20% |
| Services | 26% | 8% | 14% | 28% | 8% | 16% |
| Hotel | 26% | 10% | 12% | 24% | 18% | 10% |
| Default | 28% | 12% | 15% | 18% | 12% | 15% |
Column headers: quality · popularity · value · service · atmosphere · reliability
Pipeline
- 1
Resolve category
Name, types, cuisine, and category map to a weight bucket (coffee, gym, food, …).
- 2
Score dimensions
Rating, review count, price level, open status, business status, and review text → six 0–100 scores.
- 3
Weight & label
Category weights produce overall; band → Elite / Strong / Mixed / Risky.
- 4
Explain
Ranking hint + one-line summary call out strongest and weakest dimensions.
What this is not
- Not a paid ranking product — scores are formula-driven.
- Not a substitute for visiting, calling, or reading full reviews yourself.
- Not live inventory, wait times, or real-time quality inspection — signals are public place metadata + review text.
- Demo mode is honest: curated sample businesses until Places is unlocked.