Short answer. For a coach operator in Algeria in 2027, the first app to build is not the traveller’s: it is the conductor’s phone. It carries the passenger list, sells and records tickets bought en route with a receipt, registers each parcel with a photo and a code for the recipient, and works with no network between two wilayas. Behind it, a single seat stock prevents the same seat being sold twice between the counter, the website and agents, and each trip closes with reconciled cash: tickets, parcels, fuel. With fares regulated, load and cash control make the margin, not price.
This guide is the fourteenth in the series “Seen in France, to build in Algeria”, after tutoring and online ticketing. It is written for private coach operators, staff transport companies and agencies, and for the diaspora founder who has seen FlixBus and wants to build the equivalent with Algerian operators.
What Mahatati has already shown
Selling coach tickets online is no longer a hypothesis in Algeria. SOGRAL, which runs the bus stations, launched the Mahatati app in early 2023: you choose a departure, reserve a seat and pay by CIB or Edahabia, then show the digital ticket at boarding. According to APS, the app sold 2,026,784 tickets in 2025, up 160%, across 60 bus stations, led by Algiers, Jijel, Tlemcen and Annaba. Since 18 May 2026 return tickets have been offered, first on Algiers–Biskra, Algiers–Sétif and Algiers–Tlemcen. Travellers already pay online when they are offered the chance.
What Mahatati does not solve is what happens at the operator: the coach leaving from a stop outside a station, seats sold on board, parcels, fuel. The peaks are known: for Eid al-Adha 2025 SOGRAL announced 2,650 extra departures, and 2,200 for 2026. And margins are squeezed: according to Maghreb Émergent, public passenger transport fares date from 2018 and requests to raise them have been rejected. The private sector reportedly employs about a million people, drivers and conductors included, according to an operators’ federation quoted in the press. When the price does not move, every empty seat and every unrecorded ticket shows straight up in the result.
A trip that holds, in six moments
The system has to hold each of these six moments. The first three happen before departure; the last three, on the road and on arrival.
Departure is described by its line, time, coach, driver and the number of seats allocated to each channel: counter, online, agents, station. The sale draws from a single stock: when a seat is sold at the counter, it disappears from the website in the same second. Boarding uses the passenger list on the conductor’s phone, by QR or by name, and seats left free are offered until the last moment.
On board, a ticket sold between two stops is entered on the phone, with a receipt for the traveller. Parcels entrusted to the coach are registered at drop-off, photographed, and handed over against a code sent to the recipient. Closing, finally, reconciles for each trip the tickets, parcels, fuel and the number of passengers counted on arrival.
«When the price does not move, every empty seat and every unrecorded ticket shows in the result.»
What to build: the conductor’s phone first
The temptation is to start with a beautiful booking app. It is useful, but a large share of the money does not go through it.
The conductor’s phone is the heart of it. At departure it downloads the trip list; en route it sells and records tickets and parcels, prints or sends a receipt, and keeps everything in memory when the network drops in the gorges or on the High Plateaus; it syncs as soon as the network returns. Every sale carries the time, the stop and the conductor. If recording takes longer than a cash sale, it will not be done: the screen has to fit in three taps.
The single seat stock comes next, because it is what allows online sales to open without selling the same seat twice. The dashboard per line shows the owner each trip’s load, revenue, parcels and fuel; many transport companies are family businesses, and the partner living in France can follow the figures without calling, as we describe in running a business in Algeria from France. The offline logic is the same as at a concert gate, covered in online ticketing.
What an untraced on-board ticket costs
An example on assumptions, to redo with your lines: ten coaches, two departures a day, 50 seats, 70% load, 1,000 DA average ticket.
That makes 700 passengers a day. If 30% of tickets are sold on board, 210,000 DA pass through conductors’ hands every day. If 10% of those sales are not recorded, that is 21,000 DA a day, about 630,000 DA a month. We know of no public measurement of these rates, and they vary enormously from one company to another; they are assumptions to show the order of magnitude, not an accusation against conductors.
Checking needs no software: for one week, have passengers counted on arrival on two lines and compare with the tickets in the till. The gap, if there is one, gives the value of the project. And the most effective counter is often social as much as technical: a conductor’s share calculated on recorded sales aligns their interest with the company’s.
«The booking app is what you show; the conductor’s phone is what brings the money in.»
Six leaks, six counters
An operator loses money through small doors, on the road and at the counter. Six come up everywhere.
The seat sold twice is the leak that costs most in reputation: two travellers for one seat on the eve of Eid means an argument at departure and a customer lost. It disappears with a single stock, locked at the sale. The untraced parcel costs in disputes: a photo at drop-off and a pickup code sent to the recipient are enough to know who handed what to whom.
The coach leaving half empty is a silent leak: remaining seats must stay on sale, at the counter and online, until departure, and show on the conductor’s phone for the next stops. The unannounced delay, finally, is fixed with an SMS giving the estimated time, calculated from the coach’s position.
Summer, the diaspora and shuttles
In summer part of the diaspora arrives by plane or ferry, with luggage, often as a family, and still has to reach their wilaya. It is a precise demand: a seat booked and paid from France, in euros, on a shuttle that waits for the ferry or flight, even when late, and drops each family in its town.
The mechanism is the rest of this article, with two additions: tracking the arrival of the ferry or flight, to move the shuttle’s departure, and a written luggage limit at booking, to avoid an argument in front of the coach. For an operator, these are seats paid in advance, in foreign currency, at the time coaches are most in demand; for a founder, it is a simple entry point before a wider network. The principles of booking from France are the same as for hotels.
Which model to build
Six possible models, from the operator equipping its coaches to the network that sells for several.
One operator and one line are the safest starting point: the effect shows within weeks, in cash and load. The line network, FlixBus style, which reports more than 65 million passengers carried in France since 2015, is the most ambitious model: a shared brand and sales for coaches owned by several operators. It needs written rules, punctuality, cleanliness, on-board recording, and a revenue split that each operator can check.
Staff transport is often the most profitable model in the short term: company contracts, a monthly invoice, but routes, attendance and delays to justify; it is a cousin of our transport and delivery platforms. Parcels by coach add revenue to every trip, provided they are tracked. Car-sharing and ride-hailing follow a different logic, which we cover in building the Uber or BlaBlaCar of Algeria.
The calendar for 2027
About fifteen weeks between scoping and a measured pilot, to be ready before Eid al-Fitr.
November: scoping, lines, timetables, coaches and current fares. November and December: the single seat stock and the counter, then the conductor’s phone, tested offline on a real line. January and February: the pilot on one line, counting passengers on arrival and comparing with the till. Early March: Eid al-Fitr, expected around 9 March, the first peak where live seats show their value.
March to May: online sales and parcels, before Eid al-Adha, expected around 16 May, then summer and the arrival of the diaspora. These durations are consistent with those we set out in how long it takes to build an app or platform.
What makes this kind of project fail
- Starting with the traveller app. A beautiful online booking changes nothing if on-board cash stays invisible.
- A conductor’s phone that needs the network. On the road the network drops; sales must continue and sync afterwards.
- Recording slower than cash. Beyond three taps per ticket, the conductor goes back to the paper book.
- Two seat stocks. Separate counter and website means a seat sold twice on the eve of Eid.
- Parcels with no photo or code. The first dispute costs more than the system.
- A project against conductors. Without a share tied to recorded sales, the system will be bypassed.
- Accounts in the developer’s name. Server, domain, data and payment account in the operator’s name from the first week.
Frequently asked questions
- Mahatati already exists: why would an operator have its own system?
- Mahatati sells seats in SOGRAL stations and sold more than two million tickets in 2025. But it does not handle on-board cash, parcels, fuel or the operator’s departures outside stations. An operator’s own system holds those moments and can keep a single seat stock across the different sales channels.
- Can tickets be sold on board without a network?
- Yes, if the conductor’s phone downloads the trip list at departure, records sales and parcels offline with time and stop, and syncs as soon as the network returns. The test is done on a real line before launch.
- Are fares free?
- No: according to the press, public passenger transport fares are regulated and date from 2018. Check the scale in force for your lines. That is exactly why load and cash control are the real margin levers.
- How do you track parcels carried by coach?
- Through registration at drop-off with a photo, sender and recipient, a pickup code sent to the recipient, and handover against that code on arrival. Each parcel is tied to a trip and appears in its cash.
- Do travellers need a mobile app?
- Not at first. A booking website with payment by CIB, Edahabia or euros and a ticket by SMS is enough. The app comes when travellers return often; we cover those choices in building a mobile app in 2027.
- How long does it take to be ready?
- About fifteen weeks between scoping and the end of a measured pilot on one line, with the seat stock, the counter and the conductor’s phone. Online sales and parcels follow over two to three months, before Eid al-Adha and summer.