Citi Bike
2026
Citi Bike Rider Demand & Membership Analysis
An end-to-end cloud analytics project analyzing nearly 1 million NYC 311 service requests to uncover demand patterns, service performance, geographic trends, and operational opportunities across New York City.
Business Problem
Understanding Rider Behavior and Network Demand
Citi Bike serves both members and casual riders whose usage patterns may differ across time, location, trip duration, and bike type. Understanding these patterns can help identify opportunities to improve bike availability, station planning, and rider strategy.
Key Question
How do Citi Bike members and casual riders differ in their usage behavior, and when and where is demand concentrated across the bike-share network?
Analysis Focus
When is rider demand highest?
How do members and casual riders differ?
Which stations experience the most activity?
Where is demand geographically concentrated?
How are electric and classic bikes used?
Dataset & preparation
Preparing Citi Bike Trip Data for Analysis
Dataset
Source: Official Citi Bike NYC Trip History Data
Working Sample: 25,000 trips from the July 2026 source file
Final Analytical Dataset: 24,993 rides
The dataset included trip timestamps, rider type, bike type, start/end stations, and geographic coordinates. The raw data was imported into PostgreSQL for validation, cleaning, and transformation before visualization in Tableau. The original schema contained 13 trip, station, geographic, and rider fields.
Data Preparation
Quality Checks
Checked duplicate ride IDs and missing values
Validated trip start/end timestamps
Reviewed station and coordinate completeness
Examined ride-duration outliers
Transformations
Calculated Ride Duration
Extracted Ride Date and Ride Hour
Created Day of Week
Classified trips as Weekday / Weekend
Removed invalid/out-of-period records
Workflow
Raw Citi Bike CSV → PostgreSQL → Data Quality Checks → Cleaning & Transformation → SQL Analysis → Tableau
Sampling Note: The analysis uses a 25K-trip working sample from the beginning of the July 2026 source file. Findings therefore describe the analyzed sample rather than the complete Citi Bike network.
SQL Analysis
Exploring Demand and Rider Behavior
SQL was used to analyze the cleaned dataset across four primary areas:
Temporal Demand
Analyzed hourly and weekday/weekend patterns to identify periods of concentrated trip activity.Rider Behavior
Compared members and casual riders by trip volume, share of rides, and average ride duration.Station Demand
Ranked starting and destination stations separately to identify high-activity locations.Bike Type Usage
Compared electric and classic bikes by trip share and average duration.
The analysis identified a strong evening demand period, substantial differences between member and casual ride duration, different start and destination hotspots, and dominant e-bike usage.
Interactive Tableau Dashboard
Turning Analysis Into an Interactive Decision-Support View
The SQL findings were translated into an interactive Tableau dashboard that brings together overall performance, rider behavior, temporal demand, station activity, geographic demand, and bike usage.
Core KPIs
24,993
Total Rides
82.26%
Member Share
12.3 min
Avg Ride Duration
70.98%
E-Bike Share
Dashboard Analysis
The dashboard includes:
Hourly demand by rider type
Member vs casual trip volume and duration
Top start/destination stations
Geographic station-demand map
Electric vs classic bike usage
Interactive controls allow the analysis to be filtered by date, rider type, bike type, and station type.
Key Findings
What Did the Data Reveal?
Membership drives trip volume
Members accounted for 82.26% of analyzed rides, indicating that member activity represents the majority of usage within the sample.
Casual riders take longer trips
Casual rides averaged 18.21 minutes, compared with 11.03 minutes for members—approximately 65% longer.Evening hours show the strongest demand
Trip activity was highest during approximately 5–7 PM, with the largest hourly volume occurring at 6 PM with 2,147 rides. A smaller morning concentration appeared around 8 AM.Start and destination hotspots differ
The highest-volume start station and destination station were different, showing that station demand should be evaluated directionally rather than assuming departures and arrivals follow identical patterns.E-bikes dominate usage
Electric bikes represented 70.98% of analyzed trips, compared with 29.02% for classic bikes.
Recommendations & Limitations
Translating Insights Into Action
Optimize Peak Availability
Prioritize bike and dock availability during the evening demand window, particularly around 5–7 PM.
Improve Station Rebalancing
Evaluate high-volume departure and destination stations separately to better identify potential directional rebalancing needs.
Protect E-Bike Availability
Given the high e-bike share within the sample, maintain strong e-bike availability at high-demand stations and periods.
Explore Casual-to-Member Opportunities
Casual riders show distinctly longer ride behavior. Future customer-level analysis could identify repeat casual users who may be suitable for targeted membership strategies.
Limitations
The analysis uses a 25K working sample from the beginning of the July 2026 source file rather than the complete monthly dataset. Results should therefore be interpreted as patterns within the analyzed sample and not as system-wide July estimates.
The trip data also does not directly capture trip purpose, weather, events, station capacity, or rider intent, so behavioral interpretations should be treated as analytical hypotheses rather than causal conclusions.
VIEW DASHBOARD →
VIEW GITHUB REPOSITORY →
