November 14, 2025
As squash fans, we all know that squash courts can be found in all sorts of places – from dedicated squash clubs to hotel fitness centres, from university sports facilities to military bases. But how do we know which category each venue belongs to?
With over 6,500 venues in our database, manually categorizing each one would be impossible. That’s why we’ve built an intelligent, automated categorization system that uses cutting-edge technology to get it right.
Our 16 Venue Categories
Before we dive into how we categorize, let’s look at what we’re categorizing into.
Our system recognizes 16 different venue types:
-
Dedicated facility – Venues that are exclusively for squash (like “Squash Club” or “Squash Centre”)
-
Leisure centre – Multi-sport recreation facilities (sports complexes, recreation centers)
-
School – Primary, secondary, and other educational institutions
-
Gym or health & fitness centre – Fitness centers and health clubs
-
Hotel or resort – Hotels and resorts with squash facilities
-
College or university – Higher education institutions
-
Military – Military bases and facilities
-
Shopping centre – Shopping malls with squash courts
-
Community hall – Community centers and civic facilities
-
Private residence – Private homes with courts
-
Business complex – Office buildings and corporate facilities with sports amenities
-
Private club – Sports-focused private membership clubs
-
Country club – Golf or country clubs (more social, less sports-focused)
-
Industrial – Factories, warehouses, and industrial sites with staff facilities
-
Other – Venues that don’t fit any of the above (we try to avoid this!)
-
Don’t know – Venues we haven’t categorized yet (this is what we’re working on!)
How We Categorize: A Multi-Layered Approach
Our categorization system uses a multi-step process that gets smarter at each stage.
Think of it like a detective solving a case – we start with the most obvious clues and work our way to more complex analysis.
Step 1: The Name Says It All (Highest Priority)
The first thing we check is the venue’s name.
If a venue is called “Squash Club” or “Leisure Centre,” that’s a strong hint.
Our system analyzes names in multiple languages, recognizing patterns like:
-
“Squash Club” → Dedicated facility
-
“Leisure Centre” or “Recreation Center” → Leisure centre
-
“Hotel” or “Resort” → Hotel or resort
-
“School” or “École” → School
…and many more in English, Spanish, French, German, and other languages.
Special case – Dedicated Facility vs Leisure Centre
If a venue name contains “squash” but also mentions other sports (like tennis, swimming, or multi-sport), we know it’s not a squash-only facility – it’s a multi-sport leisure centre.
Step 2: Combination Logic (The Smart Detective)
Sometimes a single Google Places type isn’t enough.
We look at multiple types together to make smarter decisions:
-
Gym + Swimming Pool = Leisure centre
-
Gym + Multiple Sports Facilities = Leisure centre
-
Golf Club + Restaurant = Country club
-
Private Club (without golf) = Private club
-
Office Building + Sports Facilities = Business complex
-
Factory + Sports Facilities = Industrial
This helps distinguish between similar categories – for example, a Private club is sports-focused, while a Country club is more social.
Step 3: Google Places Type Mapping
The Google Places API gives detailed “types” for each venue.
We translate those into our own categories using a mapping table:
| Google Type | Our Category |
|---|---|
gym, fitness_center, health_club |
Gym or health & fitness centre |
sports_complex, sports_club, recreation_center |
Leisure centre |
hotel, resort_hotel |
Hotel or resort |
school, primary_school, secondary_school |
School |
university, college |
College or university |
community_center |
Community hall |
private_club |
Private club |
country_club, golf_club |
Country club |
military_base |
Military |
shopping_mall |
Shopping centre |
office_building |
Business complex |
We check the primary type first (highest confidence), then use secondary types if needed.
Step 4: Language Translation (Breaking Down Barriers)
When a venue name is in Chinese, Arabic, Indonesian, or any other language, we automatically translate it to English using the Google Translate API before re-analyzing.
Examples:
-
“网球俱乐部” → Tennis Club → analyzed
-
“نادي الرياضة” → Sports Club → analyzed
This lets us categorize venues globally – language is no barrier.
Step 5: AI-Powered Fallback (The Expert Consultant)
When the data is unclear, we call in our AI expert (GPT-5). It receives the venue’s name, address, Google Places data, and our category definitions, then recommends the most likely match – with reasoning.
AI excels at:
-
Distinguishing between Private club and Country club
-
Handling edge cases
-
Suggesting new categories where needed
Note: The AI knows that “Dedicated facility” means squash-only – not just “has squash courts.”
Confidence Levels: How Sure Are We?
Each categorization includes a confidence level:
-
HIGH – e.g. venue name is “Squash Club” or type is
hotel -
MEDIUM – name contains “squash” but might be multi-sport
-
LOW – only matched a secondary type
We only auto-update venues when confidence is MEDIUM or HIGH, reviewing low-confidence cases manually.
Special Handling: Place ID Management
Google Place IDs can expire. Our system automatically:
-
Tries Google’s Place ID refresh method
-
If that fails, searches by name + address via the Text Search API
-
If found, updates the Place ID and retries
-
If not found, flags the venue for deletion (likely closed)
The Human Touch
Automation with transparency:
-
Dry-run mode – Preview recommendations before applying
-
Detailed reports – See exactly why each venue was categorized
-
Audit logging – Every change logged with reasoning
-
Batch processing – 5–50 venues per day for review
-
Export options – Download CSV or JSON reports
Why This Matters
Accurate categorization helps us:
-
Understand the squash landscape – Where courts are, and what kinds
-
Improve search & filtering – e.g. “Show me all hotels with squash courts”
-
Generate insights – e.g. “Are more courts in leisure centres or dedicated facilities?”
-
Enhance user experience – Helping players find the right venue fast
Continuous Improvement
Our system keeps learning by tracking:
-
Unmapped Google types – New ones we haven’t seen
-
AI suggestions – Potential new categories
-
Edge cases – Venues that don’t fit neatly
We refine mappings and add categories over time.
The Bottom Line
Manually categorizing 6,500 + venues would take months! Our automated system processes them intelligently, combining data sources and AI to get it right – while keeping human oversight.
Behind every venue is a real place where people play squash, and getting the category right helps connect players with the courts they love.
