01 · Fleet Yield Analysis

Your fleet.
One question per panel.

Vantage Fleet OS is the operating layer for a Turo rental business of any size. Every section below answers a specific question with a chart, a finding, and the math behind it. Drop in your own Turo CSV export to run the same analysis on your real data, or explore the synthetic demo first.

OperatorVantage Ventures Rentals

Fleet3 Tesla · 2 ICE

History26 mo trailing

MethodTrip-level

SourceSynthetic demo data

/ What this dashboard tracks

A working operator's view of fleet yield.

Every Sunday I sit down with the same question: which vehicle in my fleet is actually earning its operational slot? Each car costs roughly the same to insure, list, photograph, message about, and maintain. The slot is the constraint. This dashboard turns trip-level data into the five sub-questions an operator needs to answer that, regardless of fleet size.

VEHICLE

A car in the fleet. 5 in this dataset. Identified by short name; carries type (Tesla / ICE), acquisition cost, and listing identity.

TRIP

A single Turo booking. Fields: start_date, days, distance_mi, earnings_usd. The atomic record in the source CSV.

BOOKED-DAY

The atomic unit of utilization. One day a vehicle was actively rented. Out of 365 possible per year, this is the slot-use metric.

METRICS

$/booked-day (yield), utilization (capacity), CV (consistency), z (per-trip outlier flag).

SOURCE Production: Turo CSV exports plus Tessie API for vehicle telemetry. This demo: synthetic data, calibrated against the real TTM numbers from my fleet. Drop your own CSV in the upload panel below to run the same analytics on your data.

Fleet Rev TTM

--vs prior 12mo

Rev / Booked-Day

σ = $0

Utilization

booked / fleet days

Trips / Month

avg, last 12mo

Trend R²

0.00

monthly rev fit

Run on your data

0 rows loaded

Drop a Turo CSV export below. Parsing happens entirely in your browser · nothing leaves your machine. Schema detection handles the common column-name variants automatically.

Click to upload or drag CSV here

.csv · utf-8

Expected schema · auto-detected aliases

vehicle	car / vehicle_name
start_date	trip_start / pickup_date
end_date	trip_end / dropoff_date
days	trip_days / duration
distance_mi	miles / mileage
earnings_usd	earnings / payout

Missing columns are ignored. Currency symbols, commas, and quotes are stripped automatically.

/ 01

Per-vehicle yield

Distribution & significance

Q1. Which vehicle generates the most revenue per booked-day, and is the spread between top and bottom statistically meaningful? box plot · t-test

Finding

Loading analysis...

METHODPer-trip revenue / booked-days, grouped by vehicle. Box shows IQR with median; whiskers at 1.5 × IQR; dots beyond whiskers are statistical outliers (z > 2). Welch's two-sample t-test on top vs bottom yielder.

/ 02

Weekday premium

Mean & 95% CI by day

Q2. Renters say weekends are the premium days. Is that true at the per-day rate level once you control for trip length? 95% CI bars

Finding

Loading analysis...

METHODTrip start day-of-week vs revenue per booked-day. Bars are sample means, error bars are 95% CI. Sample sizes shown above each bar.

/ 03

Revenue trajectory

Monthly + linear fit

Q3. Is the fleet actually growing, or is recent strength just seasonality? Decompose the trend and report goodness of fit. OLS · R²

Finding

Loading analysis...

METHODMonthly revenue rollup. OLS regression of revenue on month index. Slope reported per month. R² reported as goodness of fit. Stat grid shows fleet-wide $/day distribution properties.

/ 04

Outlier book

|z| > 2 trips

Q4. Which trips broke the model? List the outlier bookings so they can be investigated for pricing errors, damage claims, or expansion patterns. z-score · review queue

Vehicle	Start	Days	$/Day	Z	Earnings	Flag

Finding

Loading analysis...

METHODPer-trip $/day standardized within vehicle (z-score from vehicle-specific mean & std). Trips with |z| > 2 are flagged. Top 12 absolute z-scores shown.

/ 05

Vehicle scorecard

Stock-screener style

Q5. If each vehicle were a security, how would they rank by revenue, yield consistency, and risk? screener · CV ranking

Vehicle	Trips	Booked-days	Revenue	$/Day Mean	Median	σ	CV	Rank

METHODCV (coefficient of variation) = σ / mean of per-trip $/day. Lower CV = more consistent yield. Composite rank = 50% revenue + 30% mean $/day + 20% inverse CV.

/ 06

Issues & recommendations

What to do next, ranked

Q6. If I only had time to investigate three things this week, what should they be? triage · ranked actions

✓Loading...

METHODEach issue is generated from the same aggregated dataset feeding the panels above. Severity is assigned by impact (revenue at risk, deviation from fleet median, statistical significance). High = act this week. Med = investigate this month. Low = monitor.

Why this work

Operating a 5-vehicle Turo fleet by spreadsheet means every fleet question is a manual exercise.

I built this so the daily question, "what's the lowest-yield asset and what should I do about it?", has a one-screen answer with the supporting evidence already laid out. It's not a BI tool. It's a decision-support layer that turns raw bookings into ranked, actionable issues.

Craft on display

Schema design
Derived metrics
Aggregation pipeline
Interactive Plotly filtering
Scorecard heuristics (CV, yield gap)
Automated issue detection
Severity ranking
Responsive layout
Client-side, zero backend

Your fleet.One question per panel.

A working operator's view of fleet yield.

Run on your data

Expected schema · auto-detected aliases

Per-vehicle yield

Weekday premium

Revenue trajectory

Outlier book

Vehicle scorecard

Issues & recommendations

Operating a 5-vehicle Turo fleet by spreadsheet means every fleet question is a manual exercise.

Your fleet.
One question per panel.