I'm a data person. When something costs me money repeatedly, I track it.
After two years of flying out of Boston Logan Airport from Rhode Island, I had a feeling I was overspending on transportation. But feelings aren't data.
So I built a spreadsheet and tracked every single trip for 24 months.
The results completely changed how I think about "cheap" vs "expensive."
The setup
Location: Providence, RI to Boston Logan Airport
Distance: ~50 miles
Frequency: 15-20 trips per year (business + personal)
Time period: January 2023 - December 2024
Total trips tracked: 37
Transportation options available:
- Uber/Lyft
- MBTA Commuter Rail
- Professional car service (pre-booked)
- Drive and park at airport
I tried all of them. Here's what the numbers actually said.
Methodology
For every trip I logged:
- Transportation method
- Total cost (including tips, parking, gas, fees)
- Time of departure/arrival
- Door-to-terminal duration
- Stress level (subjective 1-10 scale)
- Whether I made my flight comfortably
I pulled data from:
- Bank statements
- Email confirmations
- Calendar entries
- Rideshare app history
- Parking receipts
Not perfect data, but good enough to spot real patterns.
The raw numbers
Over 24 months, here's what I actually spent:
| Method | Trips | Total Cost | Average |
|---|---|---|---|
| Uber/Lyft | 19 | $2,247 | $118.26 |
| Pre-booked car service | 12 | $1,176 | $98.00 |
| Commuter rail | 4 | $186 | $46.50 |
| Drive + park | 2 | $287 | $143.50 |
Wait.
The "expensive" pre-booked service was 20% cheaper on average than the "affordable" rideshare option?
That's not what I expected.
The Uber problem: variance
The issue with rideshare wasn't the average—it was the unpredictability.
Uber/Lyft cost distribution:
- Minimum: $62
- Maximum: $241
- Median: $97
- Standard deviation: $47
Nearly 4x difference between cheapest and most expensive ride for the exact same route.
Surge pricing breakdown:
| Surge Level | # of Trips | Avg Cost |
|---|---|---|
| No surge (0-20%) | 7 | $76 |
| Moderate (21-50%) | 6 | $112 |
| High (51-100%) | 4 | $167 |
| Extreme (100%+) | 2 | $227 |
The pattern I missed: I kept checking prices during normal hours when planning trips, seeing $70-80 quotes, and assuming that's what I'd pay.
But I wasn't traveling during normal hours.
Most of my flights were:
- 6-7 AM departures (need pickup at 4-5 AM)
- 9-11 PM arrivals (peak demand time)
- Friday-Sunday (weekend surge)
Exactly when surge pricing hits hardest.
Pre-booked car service: consistency
Meanwhile, the 12 pre-booked car service trips were boring:
- Minimum: $95
- Maximum: $110
- Median: $98
- Standard deviation: $4
Every trip within a $15 range regardless of:
- Time of day
- Day of week
- Weather
- Demand
Why? Flat rate pricing quoted in advance.
Additional data point: 0 cancellations out of 12 trips.
Compare to Uber: 3 cancellations out of 22 booking attempts (13.6% failure rate).
Each cancellation = rebooking at higher surge price + stress.
Commuter rail: cheapest but...
The train wins on pure cost: $46.50 average
(That's $12.75 for the train + ~$15-20 Uber to/from Providence Station on each end)
But the hidden costs:
Time cost:
- Average door-to-terminal: 2h 45min
- vs 1h 10min for direct car options
- That's 1.5 hours lost per trip
Stress data:
- Average stress rating: 7.2/10
- Made flight comfortably: 75% (vs 100% for car)
- One near-miss due to train delay
Physical cost:
- Multiple flights of stairs
- Carrying luggage through stations
- Standing on crowded trains
- Walking through terminals after already exhausted
For a solo traveler with a backpack and a midday flight? Train works fine.
For my typical scenario (laptop bag + roller bag, 6 AM flight)? The $46.50 wasn't worth it.
Driving: the hidden costs
I drove myself exactly twice before the data convinced me to stop.
Trip 1 (3 days):
- Logan economy parking: $29/day × 3 = $87
- Gas: $14
- Tolls: $8
- Total: $109
Trip 2 (5 days):
- Logan economy parking: $29/day × 5 = $145
- Gas: $14
- Tolls: $8
- Total: $167
Not included in those numbers:
- Time spent in traffic (unpredictable)
- 25 minutes searching for my car when I returned
- Stress of navigating Boston traffic
- Arriving exhausted before trip even started
- Still having to drive home after landing at midnight
The spreadsheet screamed "stop doing this."
Time value analysis
This is where it got interesting.
I calculated opportunity cost using my freelance hourly rate ($85/hr):
Time spent door-to-terminal:
| Method | Avg Time | Time Cost |
|---|---|---|
| Uber/Lyft | 1h 10m | $99 |
| Car service | 1h 10m | $99 |
| Train | 2h 45m | $234 |
| Driving | 1h 30m + return | $179 |
Real cost = Actual cost + Time cost:
| Method | Actual | Time | Total |
|---|---|---|---|
| Uber/Lyft | $118 | $99 | $217 |
| Car service | $98 | $99 | $197 |
| Train | $47 | $234 | $281 |
| Driving | $144 | $179 | $323 |
When I factored in time value, pre-booked car service was the clear winner.
Even if you value your time at minimum wage ($15/hr), the train's extra 1.5 hours costs $22.50, bringing total to $69—still not dramatically cheaper than car for the added hassle.
Day of week patterns
I broke down rideshare costs by day:
| Day | Uber/Lyft Avg | Car Service |
|---|---|---|
| Monday | $89 | $98 |
| Tuesday | $76 | $98 |
| Wednesday | $82 | $98 |
| Thursday | $91 | $98 |
| Friday | $143 | $98 |
| Saturday | $128 | $98 |
| Sunday | $152 | $98 |
Takeaway: If traveling Friday-Sunday, flat-rate service is almost always cheaper. Tuesday-Wednesday it's close, but consistency still wins.
Stress correlation
I rated stress level 1-10 for each trip:
| Method | Avg Stress | Stress Sources |
|---|---|---|
| Drive yourself | 8.5 | Traffic, parking, navigation |
| Train | 7.2 | Tight connections, timing, stairs |
| Uber/Lyft | 5.8 | Cancellations, price uncertainty |
| Pre-booked car | 3.1 | Minimal—everything confirmed |
Key finding: Stress strongly correlated with uncertainty, not cost.
- Will my driver cancel?
- Will surge pricing spike?
- Will I miss my connection?
- Will traffic make me late?
Pre-booking eliminated most uncertainty.
Departure vs return trip differences
My behavior changed based on trip direction:
Departures (time-sensitive):
- Pre-booked car: 10 trips
- Uber: 6 trips
- Train: 2 trips
- Drive: 0 trips
Returns (flexible):
- Uber: 13 trips
- Pre-booked car: 2 trips
- Train: 2 trips
- Drive: 2 trips
Why? Missing your outbound flight is catastrophic. Missing a ride home is annoying but manageable.
I'm willing to gamble on surge pricing or train timing when I've already reached my destination.
What the data changed
Based on 24 months of tracking, my current approach:
For departures:
- Book transportation 2-3 days ahead
- Use pre-booked car service
- Confirm night before
- Budget $95-105
- Arrive calm
For returns:
- Check rideshare when landing
- If price reasonable (<$100), take it
- If surging (>$120), pre-book pickup
- If exhausted, pre-book regardless
Never again:
- Train with luggage for early flights
- Driving for trips over 2 days
- Booking rideshare morning-of for 6 AM flights
Unexpected findings
1. "Cheapest" depends entirely on when you travel
Train is cheapest for midday solo trips. For early morning with bags, it's expensive when you factor in time and stress.
2. Consistency beats occasional wins
I'd rather pay $98 every time than average $118 with high variance.
3. Cancellation rate matters
0% vs 13.6% cancellation rate isn't just stress—it's real cost when you rebook at surge prices.
4. Time has actual monetary value
That extra 1.5 hours on the train isn't "free" if you could be working, sleeping, or literally anything else.
5. Optimization varies by person
A student with flexible schedule and no luggage should take the train.
A business traveler with early flights and bags should pre-book.
Your optimal solution depends on your constraints.
The spreadsheet template
If you want to track your own data:
Core columns:
- Date
- Flight time
- Transportation method
- Base cost
- Tips/fees/extras
- Total cost
- Door-to-terminal time
- Stress rating (1-10)
- Made flight comfortably (Y/N)
- Day of week
- Notes
Useful calculations:
=AVERAGE(cost_by_method)
=STDEV(cost_by_method)
=COUNTIF(method, "Uber")
=AVERAGEIF(day, "Friday", cost)
Track for 6+ months. Patterns emerge clearly.
Tools used
- Google Sheets (free, collaborative)
- Email search for historical confirmations
- Bank statement export for verification
- Basic formulas (AVERAGE, MEDIAN, COUNTIF, STDEV)
- Conditional formatting for visual patterns
Nothing fancy. Just consistent tracking.
Limitations of this data
Sample size: 37 trips isn't huge
Single route: Providence to Logan only
Personal context: Your constraints differ
Time period: 2023-2024 pricing
Subjective metrics: Stress ratings are personal
Your data will differ based on:
- Distance from airport
- Flight timing patterns
- Luggage requirements
- Schedule flexibility
- Time valuation
- Stress tolerance
But the methodology applies anywhere.
What I'd track differently
If I started over:
- Weather conditions (does rain affect surge pricing?)
- Booking lead time (does advance booking lower Uber prices?)
- Airport parking availability (does lot fullness vary?)
- Traffic delay minutes (quantify the variance)
- Return on investment for different methods
More data = better optimization.
Final numbers
Total spent over 24 months: $3,896
If I'd used pre-booked service every trip: ~$3,626 (7% less)
If I'd used Uber every trip: ~$4,376 (12% more)
If I'd used train every trip: ~$1,721 (but unrealistic with my schedule)
For my specific use case—frequent business travel, early flights, luggage, high value on reliability—pre-booked transportation was optimal.
But I wouldn't have known without tracking the data.
The bigger lesson
This isn't really about airport transportation.
It's about assumptions vs data.
I assumed rideshare was cheapest because individual trips felt cheap.
I assumed pre-booking was expensive because the upfront cost felt high.
I assumed driving was free because I already owned the car.
All wrong when I actually tracked it.
What other assumptions am I making without data?
Probably a lot.
Discussion
Am I overthinking airport transportation? Absolutely.
Did tracking change my behavior and save money? Yes.
Do you track transportation costs? What patterns have you found?
Drop your data in the comments. Especially curious about:
- Other cities/airports
- International travel patterns
- Public transit optimization
- Parking cost analysis
Let's compare spreadsheets.
Top comments (0)