Should a Marketplace Require Advance Payment? How BlaBlaCar Cut No-Shows
Short answer: Require advance payment when one customer's no-show destroys scarce capacity for a supplier and the booking value is large enough to cover payment, refund and support costs. Test it on one route or service for four weeks, and keep it only if completed bookings rise after refunds and disputes are counted. Do not use prepayment merely to improve cash flow.
A reservation looks like demand, but it is only a promise. In a marketplace, a weak promise is unusually damaging because the supplier stops offering the capacity while another customer assumes it is unavailable.
The obvious response is to charge everyone upfront. That can reduce casual bookings, but it can also suppress legitimate demand and create refund work. The decision is not cash versus no cash. It is whether transferring commitment improves the whole transaction.
What BlaBlaCar changed
BlaBlaCar spent five years testing six revenue models, according to its company account of the decision. The models included bespoke company platforms, premium placement, subscriptions, advertising, a paid telephone bridge and event pages. Some earned revenue, but none combined fairness, reliability and repeatable economics.
The operational problem remained. Passengers reserved a seat, then cancelled late or failed to appear because payment was due in cash during the journey. Drivers responded by overbooking. Passengers then reserved seats in several cars to protect themselves. Each defensive action made the other side less trustworthy.
BlaBlaCar moved to online payment before travel and took a booking fee. Its account says cancellations fell from more than 35% to less than 3%. A contemporary WIRED profile described a typical example in which a driver asked €20 for a seat and the platform charged the passenger €22 or €23.
Those are founder and company-reported results, not a universal benchmark. The useful lesson is the mechanism: money made the reservation costly to abandon, while the transaction gave the platform a reason to manage reliability.
My view is that marketplaces should charge before service when a no-show consumes non-recoverable capacity. A reminder is not an adequate substitute for commitment when a missed appointment, empty seat or unused delivery slot cannot be resold.
Use the Commitment Transfer Test
The Commitment Transfer Test has four gates. Advance payment should pass all four before you impose it across the marketplace.
| Gate | Question | Evidence to collect | |---|---|---| | Capacity | Does a no-show destroy capacity that cannot be resold? | Lost seats, hours or stock per missed booking | | Behaviour | Does payment change completion rather than merely reduce bookings? | Completed bookings per 100 qualified visitors | | Economics | Does recovered contribution exceed payment and refund costs? | Contribution after fees, refunds, disputes and support | | Fairness | Can a genuine customer cancel or obtain redress predictably? | Refund time, dispute rate and supplier failures |
Failing the capacity gate usually ends the case. If an unlimited digital service loses almost nothing when a user does not attend, prepayment may add friction without recovering value.
Measure completed transactions, not reservations
Compare the old and new systems with the same denominator. Reservations per visitor will probably fall when payment is introduced. That is not failure if a larger proportion becomes completed, useful transactions.
Track qualified visitors, booking starts, paid bookings, supplier acceptance, customer cancellation, supplier cancellation, completed service, refund and dispute. Separate each step. A fall in no-shows can be offset by a rise in supplier cancellations or payment failures.
Use at least one complete purchase and cancellation cycle. Four weeks may be enough for frequent appointments, while seasonal travel may need longer. Do not declare success from the first weekend.
Price the commitment, not a punishment
The amount must be material enough to change behaviour but proportionate to the customer's risk. Options include full payment, a fixed deposit, a percentage deposit or a card authorisation released after completion.
Full payment suits a standard service with a clear price and dependable refund process. A deposit can fit variable final bills. A card authorisation may reduce cash movement but still involves payment-provider rules and failure cases.
Avoid a fee that exists only to trap customers. If the platform keeps money when a supplier fails, trust will deteriorate faster than no-shows improve. State who receives what, when a refund applies and how long it takes.
Payment services, consumer cancellation rights, safeguarding and marketplace liability vary by country and sector. Obtain qualified local legal, tax and payments advice before holding or transferring customer funds.
Worked example: LiftShare North
LiftShare North matches passengers with 40-seat coaches running between Leeds and Manchester. These figures are illustrative.
Under pay-on-arrival, 500 people reserve in a month. The no-show rate is 24%, so 500 x 24% = 120 seats go unused. Each completed passenger pays the operator £18, and the marketplace earns a £2.40 booking margin. Completed bookings are 500 - 120 = 380, producing 380 x £2.40 = £912 of monthly contribution before central overhead.
The company tests full advance payment. Only 450 people complete a reservation, but the no-show and late-cancellation rate falls to 6%. That leaves 450 x 94% = 423 completed journeys. Gross marketplace margin is 423 x £2.40 = £1,015.20.
There are 27 customer cancellations. Twenty qualify for refunds. Payment and refund charges total £72, while extra support costs £96. Net contribution becomes £1,015.20 - £72 - £96 = £847.20.
| Measure | Pay on arrival | Pay in advance | |---|---:|---:| | Reservations | 500 | 450 | | Completed journeys | 380 | 423 | | Completion rate | 76% | 94% | | Contribution after new direct costs | £912 | £847.20 |
Advance payment produces 43 more completed journeys but £64.80 less marketplace contribution in the test month. LiftShare North should not call that an automatic win. It asks whether higher reliability helps operators add routes, improves repeat purchase and reduces defensive overbooking. It also tests a £5 deposit, which may retain the behavioural gain with lower refund cost.
The surprising result is useful: reliability improved, but the first payment design did not yet improve direct economics.
Protect both sides of the booking
Commitment cannot be one-way. A supplier should face a consequence for late cancellation, misrepresentation or failure to deliver. That might be a lower ranking, a charge, temporary removal or compensation funded by the platform or supplier, subject to local law and fair terms.
Use evidence, not accusations. Record timestamps, messages, arrival confirmation and refund reasons with proportionate data collection. Give both sides a route to challenge an error.
Trust also requires plain language. Display the total amount, cancellation deadline and refund route before payment. Hiding a booking fee until the last screen may raise short-term revenue while weakening repeat use.
Run a four-week route test
In week one, calculate the capacity and contribution lost to no-shows on one repeatable service. Write the refund and supplier-failure rules before collecting money.
In week two, introduce one payment design for half the eligible traffic, where lawful and technically safe. Keep the other half on the current process. Compare completed transactions per 100 qualified visitors, not just no-show percentages.
During weeks three and four, price payment fees, refunds, disputes and support. Interview customers who abandoned at payment and suppliers affected by cancellations. On day 28, retain, revise or remove prepayment. Roll it out only if reliability improves without making contribution, access or fairness materially worse.
Related guides
Frequently asked questions
Should I take full payment or only a deposit?
Use full payment when the service and price are fixed, supplier delivery is dependable and refunds can be issued promptly. Use a deposit when the final price varies, the purchase is high value or customers reasonably resist transferring the whole amount early. Test both against completed transactions and net contribution. A £1 deposit may not alter behaviour, while a large non-refundable deposit may deter good customers. State exactly how the deposit is applied and returned. Consumer, payment and unfair-terms rules vary, so have a qualified local adviser review the arrangement.
Will advance payment always reduce no-shows?
Usually it gives customers more reason to attend, but it does not guarantee a better marketplace. Some people will abandon at checkout, use chargebacks or book elsewhere. Others will pay and still fail to appear. Measure completed services per 100 qualified visitors, along with refunds and disputes. The relevant comparison is not simply a lower no-show percentage. If paid booking cuts reservations so sharply that fewer transactions complete, try a smaller deposit, card authorisation or clearer cancellation route before abandoning the idea.
What if suppliers are the ones cancelling?
Do not ask customers to make a stronger commitment than suppliers. Track supplier cancellations separately and create a proportionate consequence, such as compensation, ranking changes or suspension after repeated failures. Refund customers promptly and explain who is responsible. A platform that holds customer money while tolerating unreliable suppliers will destroy trust. Before promising compensation or moving funds, model the cost and check local marketplace, payment and consumer rules. The purpose of prepayment is a dependable exchange, not merely a better commitment from the easier side to charge.
How long should customers have to cancel?
Set the deadline from the supplier's realistic chance of reselling the capacity. A restaurant table may be refillable several hours before service, while a specialist appointment may need two working days. Examine when replacement bookings actually occur, then choose a simple rule customers can understand. Allow exceptions for supplier failure and any rights required by law. Review cancellation reasons after the test. A deadline that recovers capacity is useful. One designed mainly to retain customer money is likely to create disputes and discourage repeat use.
Can reminders solve the problem without taking payment?
They can when simple forgetfulness is the main cause and capacity can still be resold. Test a confirmation message, calendar invitation and reconfirmation deadline before adding payment friction. Compare the cost and improvement with a deposit test. Reminders will not solve speculative booking, defensive multiple booking or weak consequences for late cancellation. If customers confirm and still fail to attend, commitment is probably the missing element. Keep every message proportionate and follow applicable privacy and electronic-marketing rules when sending them.
Should the marketplace keep the booking fee after a refund?
Usually not when the supplier failed or the customer exercised a valid cancellation right. Keeping a fee may cover a real non-refundable processing cost, but it must be lawful, disclosed and proportionate. Calculate what the payment provider actually charges rather than inventing an administration amount. Consider absorbing small costs as the price of trust. Rules differ by jurisdiction, payment method and customer type. Obtain qualified advice, and make the refund outcome visible before purchase rather than burying it in terms customers see only after a problem.
Comments (0)