Famous ideas — Uber — as pitched in 2009
Weak Fit
Something is there. Key dimensions hold it back.
THE DAY-ONE PITCH — WHAT THE INSTRUMENT WAS GIVEN
In San Francisco you can never find a taxi when you need one. Our iPhone app lets you press one button and a professional black-car driver comes to you. You watch the car arrive on a map. Payment happens automatically on the card you have on file — no cash, no tipping. It costs about 1.5x a taxi, but it actually shows up. We start with limo drivers who sit idle between corporate bookings and want to fill that time.
Traction at grading: Prototype app works. Three limo drivers have agreed to take test rides. We've ordered cars with it ourselves around the city.
Graded from these facts only. No hindsight — the instrument knew nothing after 2009.
THE FULL REPORT — ALL SEVEN DIMENSIONS
The pain point — SF taxis that don't show up — is real and specific, and the 1.5x taxi premium is a concrete pricing hypothesis, but you have zero paying customers cited; the three drivers and self-ordered rides confirm supply-side interest, not willingness to pay at scale. Post a price and charge a stranger this week to turn this into a 7.
Ride-hailing margins are structurally thin — at 1.5x taxi with a small fleet and 5–15 hours a week, the founder's version of this business tops out as a modest local service, not a category-defining company; the ceiling is real but the path from three drivers to meaningful revenue volume is long. Signing 20 drivers and modeling commission take-rate against realistic weekly ride volume would clarify whether this is a real business at this scale.
The prototype works, three drivers are committed, and you've already ordered rides with it — you are one willing paying stranger away from dollar one, which is genuinely close. The constraint is 5–15 hours a week, which slows iteration but doesn't block a first transaction.
iPhone app development and limo fleet negotiation are precisely the two skills this idea needs in its earliest phase — you built the prototype and signed three drivers, which is direct evidence of fit, not a claim. The 5–15 hours per week is the real constraint: it limits how fast you can respond to driver issues or iterate on the app, so say so explicitly to yourself before you grow.
Supply side is partially solved (three drivers under agreement), the app is prototyped, and the regulatory environment for black-car/limo services in 2010 SF is lighter than for taxis — these are licensed commercial drivers, which sidesteps the hardest compliance risk. The gap between 'prototype' and 'reliable enough that a stranger trusts it for a paid ride' is real but not large.
The model scales geographically — adding cities adds supply and demand independently — but at 5–15 hours a week it is operationally fragile; a driver no-show or app bug becomes a reputation crisis with no one to fix it fast. This is a team problem wearing a scalability costume, and it will surface the moment volume grows beyond what three drivers can absorb.
The core insight — idle limo drivers as supply, automatic card payment, live map — is genuinely novel in practice even if the pieces exist; no one has packaged it this way in SF at launch. That said, 'press a button, car arrives' is a concept any well-funded competitor could clone the moment it shows traction, and you've named no moat beyond first-mover timing.
THE SUMMARY
This is an early but concrete execution of a real and specific problem — SF taxi unreliability — with a working prototype, three committed drivers, and self-validated rides already taken, which puts it meaningfully ahead of most submissions at this stage. The structural risks are the founder's 5–15 hours per week against an operationally demanding two-sided marketplace, and thin margin economics that make the solo-founder version of this business hard to scale before a competitor with more resources notices. The single change that would most raise this grade is charging a real stranger for a real ride this week and recording what breaks.
WHAT'S WORKING
THE RISKS
THE NEXT STEP MERIT GAVE THEM
This week, find one person you know by name — not a friend doing you a favor — who agrees to pay the stated price for a ride using the prototype app. Charge their card, complete the ride, and write down every single thing that broke or felt uncertain. That is your real product spec.
What happened next: a verb, and one of the most valuable startups in history — after billions in venture capital bought the city-by-city supply the grade said the day-one pitch couldn't afford.
THE DEEP DIVE — THE $29 REPORT, PUBLISHED IN FULL
This is the actual paid product: we took the 2009 Uberpitch and resubmitted it today. The research maps the market a founder walking in with this idea right now would face — because that's what every Deep Dive does: fresh research, at purchase, on your exact submission.
THE EXPANDED READ
Your report scored this 64 because it graded what you have, not what the field looks like — and the field is more open than you think. Research shows the incumbent apps in SF ride-summoning (Taxi Magic, Cabulous) both route requests to the same unreliable taxi fleet you're competing against, which means nobody has actually solved the pickup-reliability problem you're pricing against. Your idle-limo supply model is the only one in the market that doesn't depend on taxi medallion holders showing up. The real risks the free report flagged — 5 to 15 hours a week against a live operations business, and three drivers as your entire fleet — are still the risks. But the demand side is less contested than a 6/10 marketDemand score implies, and the fastest way to prove that is to charge strangers at 1.5x meter this month and record fill rate.
THE TEARDOWN — WEAKEST DIMENSIONS
Your 6 exists because every rider so far has been you. The pain is documented — SF has roughly 1,500 taxi medallions for a city of 800,000, dispatch fill rates on phone-ordered cabs are notoriously poor, and both Taxi Magic (launched 2008) and Cabulous (SF, 2009) built businesses on that exact frustration — but neither proves anyone pays a 50% premium for reliability, because both apps are free layers over metered taxis. Your open question is not 'is the pain real', it's 'will a tech worker pay ~$15 for a ride the meter prices at $10'. That is answerable in one week with the prototype you already have.
The 6 reflects a real structural fact: at 1.5x taxi fares with three cars, your gross ride volume is capped at whatever three drivers can serve in their idle windows, and your take is a slice of that. But the unit economics are better than a typical marketplace because your supply cost is near zero — these drivers are already paid for standby time by corporate accounts, so any fare you route them is margin they didn't have. What's missing is a model: you have not named a commission rate, a rides-per-driver-per-week estimate, or a break-even fleet size, so the ceiling is unquantified rather than low.
The report called this correctly: it's a team problem in a scalability costume. A two-sided live-logistics marketplace generates incidents in real time — driver no-shows, GPS drift, card declines mid-ride — and at 5 to 15 hours a week you are not available for most of them. Cabulous and Taxi Magic sidestep this by never owning the ride experience (the taxi dispatch does); you own it, which is your differentiator and your exposure simultaneously. Geographic scaling is irrelevant until one neighborhood works without you watching.
THE 4-WEEK PLAN
WEEK 1
Charge strangers real money and log everything that breaks.
WEEK 2
Turn ride data into a unit-economics model and widen supply.
WEEK 3
Make the operation survivable without you watching it.
WEEK 4
Prove repeat demand and package the evidence.
THE COMPETITOR READ
The field is thinner than 'ride app' sounds. Taxi Magic (RideCharge, launched 2008) and SF-based Cabulous (2009) both let riders summon cabs from a phone, but both depend on the existing taxi fleet and dispatch systems — they digitize the request without fixing the no-show problem, which is exactly the pain in your submission. Pre-booked black-car services like Limos.com exist but require advance reservation and hourly minimums, not on-demand pickup. Nobody in SF combines on-demand summoning, live map tracking, automatic card payment, and non-taxi supply. That gap is your opening; it is also visible enough that a funded team could see it, so your advantage is measured in months, not years.
Live since 2008 across multiple US cities, integrated with taxi dispatch systems — proves people will book rides from a phone at scale. But it hands requests to taxi dispatch and cannot guarantee the cab shows up, which is the exact failure you're pricing against.
SF-local app showing available taxis on a live map with driver opt-in — proves the map-based summoning UX resonates in your exact city. Its supply is medallion taxi drivers who can ignore hails, so reliability remains unsolved and no premium is captured.
Established marketplace for booking black cars and limos, which proves riders already pay premium prices for professional drivers. It's reservation-based with hourly minimums — useless for 'I need a car in eight minutes', which is your entire use case.
The incumbent default, and its unreliability at peak hours is the demand engine for your whole idea. Metered pricing around $3.10 flag drop plus $2.25/mile sets the baseline your 1.5x premium is measured against — quote it when you price rides.
THE WEDGE
Your wedge is the supply side, and it's genuinely yours: idle licensed limo drivers cost you nothing to acquire because their standby time is already paid for, and you're the only player whose reliability doesn't depend on taxi medallion holders choosing to show up. Combined with your two stated skills — you built the app, you sign the fleet deals — the winnable position is a small, fanatically reliable Friday/Saturday-night service in three SF neighborhoods that charges a premium the taxi apps structurally cannot. Do not try to be citywide or 24/7 at 5–15 hours a week; be undeniable in a constrained window and let fill rate and repeat riders be the proof. The clock matters: this concept is cloneable the moment it shows traction, so evidence gathered in the next 90 days is worth more than evidence gathered in the next 12 months.
THE SCORE PATH
Your score moves on three pieces of paper: a ride log showing at least 10 completed rides charged to strangers at 1.5x meter with repeat intent recorded (marketDemand, weight 20 — the single biggest lever), a unit-economics model built from those real fares plus one signed fleet-owner commission agreement covering 10+ cars (moneyPotential, weight 18), and measured fill-rate and pickup-time metrics from an operation that ran two nights without you personally dispatching (scalability, weight 12). All three are achievable inside four weeks, your $5K budget, and your stated hours. Effort spent polishing the app instead will not move a single dimension — the prototype already works; what's ungraded is whether strangers pay.
Your idea gets this same teardown.
START WITH THE FREE GRADEThe $29 Deep Dive is offered after your score.