MERITALL FAMOUS IDEAS

Famous ideas — Uber — as pitched in 2009

64/100

Weak Fit

Something is there. Key dimensions hold it back.


THE DAY-ONE PITCH — WHAT THE INSTRUMENT WAS GIVEN

UberCab — press a button, a black car picks you up

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.

CategoryMarketplace
AudienceSan Francisco tech workers who will pay extra for a ride that actually comes
Budget$5K+
Timeline3+ months
Hours5–15
Build statusEarly prototype
SkillsiPhone app development, negotiating with limo fleet owners

Graded from these facts only. No hindsight — the instrument knew nothing after 2009.

MDEMANDMONEYSPEEDSKILLLAUNCHSCALEUNIQUE

THE FULL REPORT — ALL SEVEN DIMENSIONS

Market DemandWEAKEST

6/10

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.

Money Potential

6/10

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.

Speed to First Sale

7/10

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.

Skill Fit

7/10

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.

Ease of Launch

7/10

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.

Scalability

6/10

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.

Uniqueness

6/10

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

  • Prototype is live and you have already used it to order rides — this is demonstrated functionality, not a slide deck, and it puts first revenue within days not months.
  • Founder skills match the two hardest early problems exactly: building the iPhone app and negotiating with fleet owners, both of which are already producing results (prototype + three driver agreements).
  • The idle limo driver supply insight is operationally smart — you are not recruiting new workers, you are monetizing existing downtime, which lowers your cost of supply acquisition.

THE RISKS

  • At 5–15 hours per week, a single driver no-show, payment failure, or app crash during a real customer's ride becomes an unrecoverable reputation event you may not be available to fix in real time.
  • The premium pricing (1.5x taxi) is only defensible while the taxi alternative is reliably bad; any improvement in SF taxi availability or a funded competitor entering the market collapses your differentiation.
  • Three drivers is a fragile supply base — if one leaves, you lose a third of your capacity and likely cannot fulfill rides reliably enough to retain early users.

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

Market Demand

6/10 — can reach 8/10

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.

  • 01Run a paid pilot Thursday and Friday 6pm–10pm in SoMa and the Mission — the exact hours and neighborhoods where taxi availability collapses — and charge the stated 1.5x-meter price to strangers' cards, not friends. SF meter math is roughly $3.10 flag plus $2.25/mile, so quote a concrete number per ride and log whether anyone balks.
  • 02Stand outside two or three tech-company happy hours (South Park, 2nd Street bars) and offer to summon a car on your phone for anyone visibly failing to hail a cab, then charge them. Ten completed paid rides from named strangers converts this dimension's score; ten refusals is equally valuable data on the 1.5x ceiling.
  • 03After each paid ride, ask one question: 'what did you pay and would you pay it again Friday?' Log answers in a spreadsheet with pickup time, wait time, and fare. Fill rate and repeat intent are the two numbers a re-grade will look for.
  • 04Post in Hacker News' monthly threads and the SF-focused startup mailing lists offering pilot access to 50 people — your target user reads HN, and a waitlist of named signups is weak evidence but a completed paid ride from that waitlist is strong evidence.

Money Potential

6/10 — can reach 7/10

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.

  • 01Model it this week in a spreadsheet: assume a $15 average fare, test take rates of 15%, 20%, and 25% (limo dispatch services and affiliate networks commonly take 20%+ from drivers), and estimate rides per driver per idle shift from your three pilot drivers' actual availability. Find the fleet size where you clear $2,000/month — that number tells you whether this is a business or a hobby at your hours.
  • 02Ask each of your three drivers what their fleet owner charges per hour on corporate bookings and how many hours per week they sit idle. Idle hours times realistic fill rate is your addressable supply, and it's a number you can get in three phone calls.
  • 03Use your fleet-negotiation skill to sign two more fleet owners — not individual drivers — with a per-ride commission agreement in writing. One fleet owner with 10 cars changes your supply math more than 10 individual driver handshakes, and a signed commission structure is the evidence this dimension currently lacks.
  • 04Test the premium ceiling directly: on your pilot rides, quote some riders 1.5x and some 1.75x, and record acceptance. If 1.75x holds during peak Friday hours, your margin story improves materially and you should say so in a re-grade.

Scalability

6/10 — can reach 7/10

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.

  • 01Constrain the service surface deliberately: operate only Thursday–Saturday, 6pm–midnight, SoMa/Mission/FiDi, and be personally on call during every operating hour. A service that is reliable 18 hours a week beats one that is flaky 168 hours a week, and it fits your stated availability exactly.
  • 02Write a one-page incident runbook after your first 10 paid rides — what you do on a driver no-show (backup driver's cell number pre-arranged), a payment failure (comp the ride, retry the card next day), an app crash (fall back to SMS dispatch). The runbook is what makes the operation survivable when it's not you answering.
  • 03Recruit one operations partner or first hire from your budget — even $500/month for someone covering dispatch on the nights you can't — because the honest ceiling of this dimension at solo 5–15 hours is a 6 forever. Post in the SF startup job channels and Craigslist gigs for someone who can watch a dispatch screen Friday nights.
  • 04Instrument the app to log every request, acceptance, pickup time, and completion. Fill rate and time-to-pickup are the two metrics that prove the system works without a founder babysitting it, and they're what a re-grade — or an investor — will ask for.

THE 4-WEEK PLAN

WEEK 1

Charge strangers real money and log everything that breaks.

  • Run two pilot nights (Thursday and Friday, 6pm–10pm, SoMa/Mission) with all three drivers on call; charge 1.5x meter to at least five strangers' cards and complete the rides.
  • Log every ride: request time, pickup time, fare charged, what broke, whether the rider said they'd repeat. This log is your product spec and your re-grade evidence.
  • Call each of your three drivers' fleet owners and ask two numbers: idle hours per week per car, and what commission they'd accept per routed ride.

WEEK 2

Turn ride data into a unit-economics model and widen supply.

  • Build the spreadsheet: average fare from week 1's real rides, take rates at 15/20/25%, rides per idle driver-hour, and the fleet size that clears $2,000/month.
  • Sign one fleet owner (not individual drivers) to a written per-ride commission agreement — your negotiation skill applied where it compounds.
  • Run pilot nights again, this time A/B testing 1.5x vs 1.75x pricing on peak Friday requests; record acceptance rates.

WEEK 3

Make the operation survivable without you watching it.

  • Write the one-page incident runbook from weeks 1–2's failures: no-show protocol, payment-failure protocol, app-crash SMS fallback.
  • Hire a part-time dispatch watcher for Friday/Saturday nights from your $5K budget — post on Craigslist SF gigs and in local startup channels; budget roughly $500/month.
  • Add request/acceptance/pickup logging to the app so fill rate and time-to-pickup are measured automatically, not remembered.

WEEK 4

Prove repeat demand and package the evidence.

  • Email or text every week 1–2 rider a one-tap rebooking offer for Friday night; measure repeat rate — repeat riders are the demand proof no competitor research can substitute for.
  • Open a 50-person waitlist and seed it via a Hacker News post and SF tech-worker channels; measure signup-to-first-ride conversion.
  • Compile the four-week evidence pack: paid rides completed, fill rate, average pickup time, repeat rate, signed fleet commission, and the unit-economics model. Resubmit to Merit with those numbers.

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.

Taxi Magic (RideCharge)

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.

Cabulous

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.

Limos.com

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.

SF Yellow Cab / DeSoto phone dispatch

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 GRADE

The $29 Deep Dive is offered after your score.

Next: DoorDash59/100GRADE MY IDEA