Moving from two-way radios or old software to the cloud without losing drivers and customers
Changing software is one of the riskiest decisions for an operating taxi service. Customers call the same number and expect the same arrival time. Drivers are used to the radio channel, to «their» dispatcher, to being able to «catch» orders. Any glitch during the transition, and some drivers leave for a competitor while customers remember that «it got worse».
At the same time, staying on two-way radios or ten-year-old software in 2026 is also a risk. Without an app, a bot and automatic dispatch, the service loses on arrival time and convenience, and the 580-FZ requirements for an order log and passenger notification are hard to meet with radios.
This article gives a transition plan that minimizes losses, plus the experience of three services that have been through it: «Alliance» from Belgorod, «Sevtaxi» from Sevastopol and «7-40» from Zaporizhzhia.
Where services move from and what they lose
From radios and a notebook
The classic setup: an operator takes a call, writes down the address, announces it on air, drivers respond, the dispatcher picks one. It works as long as the dispatcher knows the city and the drivers, and there are no more orders than they can keep in their head.
What does not work: there is no order history (nowhere to keep a 580-FZ log), no customer app, arrival time depends on who pressed the push-to-talk button first rather than who is closer, and you cannot measure empty mileage or real driver utilization. Conflicts over «insiders getting the best orders» are a constant background.
From old on-premise software
Software on a server in the office, installed some eight years ago. It works, but: there are no updates or they are paid, the customer app is outdated or not published in the stores, support answers on weekdays only, and nobody customizes it for the city. The server depends on one person who knows how to restart it.
«Alliance» from Belgorod went exactly this way: radios and a notebook, then a first simple program, then foreign software with weak local support that nobody adapted to the city’s specifics. The third switch brought them to UpTaxi.
From expensive software
The features are there, but the rental or support cost is out of proportion to the size of the service. Separate modules are billed separately, and every new order channel means a new invoice. The service pays for capabilities it does not use.
What people fear about switching (and what actually happens)
| Fear | Reality |
|---|---|
| «Drivers won’t install the app» | Some won’t right away. You need a mode for those who don’t want to install anything (in UpTaxi, the driver bot in Max) and both systems running in parallel for the first weeks |
| «Operators won’t learn it» | In Sevtaxi’s experience, dispatchers adapted in about 10 days; the full transition took 1–1.5 months |
| «We’ll lose customer history» | The customer and driver database can be transferred; customer numbers and addresses are the main asset and are moved first |
| «The number will stop working» | The number stays. Only its destination changes: a SIP trunk into the cloud software instead of the office PBX |
| «We’ll need a new server and an IT specialist» | Cloud software needs no server: a computer with 4 GB of memory, a headset and an internet connection of 1 Mbps or more are enough |
| «If something breaks at night, nobody will fix it» | Check the support hours before switching. UpTaxi support is 24/7 and free, and the monitoring system reports incidents on its own |
Week-by-week transition plan
The plan is designed for a service with 50–300 orders a day and 20–100 drivers. For a large service, the stages are the same, the timelines longer.
Week 0. Preparation
Audit current processes. Write down how orders are taken now, how cars are assigned, how prices are calculated, how drivers are settled with, and which reports the owner looks at. Anything not written down will be lost in the transition.
Export data. From the old software: customers (numbers, addresses, history), drivers (details, cars, balances), fares, zones. From the notebook: at least a list of drivers and regular customers.
Choose the software and test it. If the vendor offers a trial, use all of it: set up real fares, the city map, a few drivers, and run test orders. In UpTaxi, the trial is 30 days, a server for your company is set up within a day, and configuration is done remotely. That is enough to go through all the stages below without stopping work on the old system.
Telephony. Arrange a SIP trunk with your telecom provider or prepare a VoIP gateway for analog lines. The number does not change; only the call route does. Make sure you keep multiple simultaneous lines.
Week 1. Setup and operator training
- Load the city map, fares, zones, and customer and driver databases.
- Set up auto-notification: a greeting with the service’s name, «your car is on its way…» phrases, an SMS on arrival.
- Train operators. Two to three hours on the console, then work on test orders. In services’ experience, operators coming from radios miss the air for the first few days, and after a week they don’t want to go back: no need to keep track in your head of where each car is.
- Set up the LLM voice robot: it takes routine calls, asks where from and where to, and creates the order itself. At first you can turn it on only for night hours or for overflow, so operators get used to some orders appearing «on their own».
Week 2. Onboarding drivers
The most important stage. Drivers are the ones who can leave.
Two ways to connect. Drivers with modern Android smartphones get the «UpTaxi Driver» app: an order list or auto-assignment, a taximeter, balance top-ups and vehicle inspection without visiting the office. Drivers who don’t want to or can’t install the app (old phone, little memory, weak internet) get the driver bot in Max. It works without installation, on a poor connection and even without GPS using manual location marking, with automatic dispatch, a feed of open orders, statuses, address editing and chat with the dispatcher. In Russia, Max comes preinstalled on new smartphones, so most drivers already have it. This removes the main objection: «I’m not installing anything.»
Run in parallel. Do not turn off the radios for the first 7–10 days. An order goes both on air and into the software; drivers who have already connected take it in the app or bot, the rest by radio. Gradually the share «in the software» grows, and the air goes quiet on its own.
«Order list» mode. Drivers coming from radios are used to «catching» orders. Start with list mode: the driver sees open orders and takes them, just as on air before. Turn on auto-assignment later, once drivers see that the software dispatches fairly, by road distance, not «by who you know». More on how this affects arrival time in the article on automatic dispatch.
What to tell drivers. Not «now there will be software», but what they get: orders go to the nearest driver, not the loudest; the balance is topped up from the phone; vehicle inspection is done without a trip to the office; they can take orders on the way back and drive empty less. And that nothing needs to be bought or installed if they don’t want to: Max is enough.
Week 3. Launching customer channels
Once operators and drivers are working steadily, add what you didn’t have before:
- Order bot in Telegram and Max: an order in three taps with an address or geolocation, statuses with the car’s make and plate number. Brand the bot with your service’s logo and colors. Through the bot you can send promotions and promo codes to regular customers for free, instead of paid SMS.
- Customer mobile app. The shared «UpTaxi Taxi» app is included in the plan; a branded app for your service costs 20,000 ₽ per OS.
- Dispatch service website. A ready-made website with an order form is included in the package.
- Upfront price. If you previously worked only by taximeter, turn on hybrid mode: the customer learns the price in advance and rides at that price, or by taximeter, whichever they choose. «Alliance» from Belgorod entered the market with exactly this mode after switching and immediately saw its customer base grow: customers like knowing the price in advance.
Tell customers about the new channels: an SMS to your database, stickers in cars, a post in local city communities. The key message: «Same number, and now you can also order in Telegram or Max.»
Weeks 4–6. Switching off the old system and fine-tuning
- Turn off the radios when no orders have gone through them for several days in a row.
- Keep the old software in read-only mode for a month or two, so you can look up history if needed.
- Every week, check in the manager’s account: arrival time by district, the share of unfulfilled orders, the number of drivers on shift by hour, the share of orders via the robot and bots. Adjust fares, zones and the driver response timeout.
- If you lack cars at peak hours, connect the driver exchange: a driver from another service fulfills the order, while the customer and the order stay yours.
In Sevtaxi’s experience, getting fully «on track» takes 1–1.5 months. This is a service in a city with several large chain competitors and pronounced seasonality; UpTaxi became their second system after radios.
A server for your company is set up within a day, configuration and training are remote, and the trial is 30 days. That is enough to complete the transition without stopping work on the old system. Get 30 days free.
Three transition stories
«Alliance» 222-222, Belgorod: four systems over the service’s history
Dmitry Osadchy started with radios and a notebook. Then came a first simple program, then foreign software: decent in features, but with poor local support, and nobody adapted it to the city. The move to UpTaxi was the third change of system.
The main result was not automation itself but a new product for customers: a hybrid mode in which the customer learns the price in advance and rides at that price. This grew the customer base right after launch. The review also notes: «It’s great that the team keeps moving forward»: updates come out, and requests get implemented.
Lesson: switching makes sense not for the sake of «new software» but for capabilities you didn’t have before. Identify that capability before switching and build your customer messaging around it.
«Sevtaxi», Sevastopol: 10 days for operators, a month and a half for everything
A market with several large chain services and seasonality. Radios were the first system, UpTaxi the second. The vendor approached the service itself, and the features covered everything they were looking for. Operators adapted in about 10 days, and the service was fully on track in 1–1.5 months. Evgeny’s verdict: «Everything that was promised is there and works.»
Lesson: transition timelines are realistic if you don’t try to do everything in a week. A month and a half with both systems running in parallel is normal.
«7-40», Zaporizhzhia: five programs, UpTaxi is the fifth
A large service: they started on radios with their own software and have a staff of 10+ operators. Over their history, they have changed five automation programs. What they value in the fifth: open dialogue with the developers and support, and fast response to requests.
Lesson: for a service that has been through several transitions, the deciding factor is not the feature list (many are similar) but how quickly the vendor responds when something needs to be built or fixed. Test this during the trial: message support at night and see what happens.
Full reviews are on the reviews page.
What to check with the vendor before switching
- Support hours. Hours, channels, cost. «Weekdays 9 to 6» support for a taxi service is half of support.
- Data migration. Who transfers customers, drivers and fares, and how. Is there an import format?
- Working without the app. What to do with drivers who won’t install the app. If there is no answer, you will lose part of your fleet.
- Telephony. Are your number and telecom provider supported, is a gateway needed, are call recordings kept?
- Load. Will the system handle your volume and growth? UpTaxi has been tested at a load of 25,000 orders a day for a single taxi service and 900 drivers on shift at the same time.
- Exit cost. If it doesn’t work out, what do you lose? With a percentage model, nothing; with a license purchase, the whole amount. Pricing models are compared in the article «How much does taxi software cost».
- 580-FZ compliance. Order log, notifying the passenger about the car and driver, reports. Radios don’t have this; new software should.
Common transition mistakes
- Turning off the radios on day one. Drivers who haven’t connected yet are left without orders and leave.
- Turning on auto-assignment right away. Drivers coming from radios see it as a loss of control. Start with the list.
- Not migrating the customer database. A regular customer the software doesn’t recognize by number feels like a stranger.
- Not telling customers about the new channels. Nobody will find a bot or an app without an announcement.
- Changing fares at the same time as the software. Two changes at once mean twice the discontent. Switch first, then change fares.
- Not appointing someone in charge. The transition needs one person at the service who makes decisions and talks to the vendor’s support.
Summary
Moving from radios or old software to the cloud requires no server, no IT specialist and no downtime. It requires a 4–6 week plan, both systems running in parallel, a mode for drivers without the app, and a clear answer to «what new will customers get».
UpTaxi provides everything for this in the base package: the driver app and the driver bot in Max with no installation, the LLM voice robot, order bots in Telegram and Max, the customer app, a website, road-based automatic dispatch, the exchange and 24/7 support. A server for your company is set up within a day, and the trial is 30 days.
Get 30 days free · Message on Telegram · Message on Max · Call +7 900 28-28-328
Minimum hardware requirements are on the launch requirements page.