Skip to content
Blog

Central Asia travel trade technology, built for how it sells

Tekravel Editorial
Written with the help of AI

A Saturday in a six-person agency near Chorsu bazaar in Tashkent. A family is at the desk with cash in an envelope, asking for Tashkent to Istanbul for four and a week in Antalya. Their nephew wants a link on Telegram so he can read it at home in Uzbek. The agency's site offers English and Russian, prices in US dollars, expects an international card, and loads so slowly on mobile data that he will give up first. None of this is unusual. Most central asia travel trade technology was designed somewhere else, for a market that pays by card, reads English or Russian, and has fibre in the office.

This piece is for the owner of an agency like that one, in Uzbekistan, Kazakhstan or the neighbouring markets, who is deciding what to put in front of customers and sub-agents. It is not a list of vendors. It is the four places where imported platforms fit the corridor badly, and what to test in each before you commit.

Why the imported defaults fail here

A platform built for Western Europe or the Gulf makes a set of quiet assumptions: that the customer completes the purchase online in one sitting, that the purchase is paid by an international card, that the language list starts with English, and that the page can afford to be heavy because the connection is fast. Each assumption is reasonable where it was made. In Central Asia each one is wrong often enough to cost you bookings.

The trade here also runs on relationships that a retail OTA flow ignores. Much of the volume moves through sub-agents in regional cities who take bookings from you on credit and settle later, and a platform that has no concept of that ends up being used for search while the real business happens in a spreadsheet and a chat window.

Language coverage beyond Russian

Russian is still the working language of CIS travel distribution. Supplier contracts, airline circulars and most B2B conversations between Almaty, Tashkent and Bishkek happen in it, and any platform you choose has to do Russian properly. But your retail customer is not a supplier. A younger customer in Samarkand may search in Uzbek, and in Latin script; a customer in Shymkent may prefer Kazakh. A storefront that offers them Russian or English is not broken, but it is telling them the site was not made for them, and they notice.

The same goes for the outbound market you sell into. Families booking Istanbul read Turkish hotel descriptions; pilgrims and business travellers heading to the Gulf meet Arabic; customers from the Iranian border regions may be more comfortable in Farsi. A language list is a list of who you can serve without an apology.

On this platform the storefront ships in 40 languages, among them Uzbek, Kazakh, Azerbaijani, Russian, Turkish and the right-to-left ones (Arabic, Farsi, Hebrew, Urdu), and each agency chooses which of them its own site offers. Be clear about the gap too: Kyrgyz, Tajik and Turkmen are not in that list today, so an agency in Bishkek or Dushanbe would serve its customers in Russian and should know that before signing rather than after. Whatever platform you look at, ask for its locale list in writing and open the checkout, not only the home page, in each language you need. A translated banner over an English booking form is the most common failure. We wrote about the same trap from the Arabic side in why an Arabic booking site is not a translation; the lesson carries over to Uzbek almost word for word.

Regional carrier content

Your customers fly Uzbekistan Airways, Air Astana, FlyArystan, SCAT, Somon Air, Turkish Airlines and the Gulf carriers. Some of these sell well through the global distribution systems. Others, the low-cost ones especially, put their best fares on their own sites or in channels that a GDS-first platform does not carry, or carries without the cheaper fare families. An imported platform that looks complete on London to Dubai can look thin on Almaty to Tashkent, and you find out from a customer who saw a lower price elsewhere.

So the question to ask is not "do you have flights" but "whose content will my account see on the routes I sell". On this platform, supplier content reaches an agency through supplier groups rather than one-to-one deals: a tenant sees the suppliers its groups grant it. That makes the question concrete. Ask which groups your account would join and run your own ten most-sold routes through search before you decide anything.

Hotels and ground product have the same problem in a different form. The guesthouses in Bukhara and Khiva that your customers actually want are often not in an international bedbank at all; you know them by phone. Flights, hotels, tour packages and attractions are activated per agency here, so a Samarkand–Bukhara–Khiva itinerary you assemble yourself can be sold as a package next to the flights, instead of living in a PDF you send on Telegram.

Cash, bank transfers and the checkout that assumes a card

This is where most imported platforms break, and it is worth slowing down.

In many agencies here, leisure trips are still paid in cash at the desk, by bank transfer, or with local card schemes such as Uzcard and Humo that an international gateway may not accept. Corporate clients pay against an invoice. Sub-agents in regional cities pay you weekly or monthly, often in cash brought to the office. A checkout that only knows "enter your card" turns all of that into manual work: staff search on the site, then re-key the booking, then reconcile by hand.

How the money actually movesWhat an imported platform tends to assumeWhat to check before you sign
Retail customer pays cash at the deskCustomer pays online by international cardCan staff book on the customer's behalf and mark payment taken offline, with an audit trail?
Sub-agent books now, settles weeklyEvery booking is prepaidAre sub-agent credit limits and settlement invoices built in, or a spreadsheet beside the system?
Corporate client pays by bank transferNo invoicing at allCan you register the company as a business customer and verify it manually before it books?
Prices quoted in UZS or KZT, costs in USDOne display currency, the vendor's rateWho sets the conversion rate you show, and can you change it the morning the rate moves?

On the sub-agent and corporate rows, this platform has an answer we can state plainly: sub-agent credit limits and settlement invoicing are first-class concepts rather than spreadsheets, and a storefront can require sign-in before search and register business customers with optional manual verification. That covers the part of the trade that runs on credit. How a retail traveller pays online depends on which payment gateways are available in your country, so ask that question by name for Uzcard, Humo or Kaspi rather than accepting "we support payments".

The currency row deserves its own sentence. Each site picks its own default currency and the list it offers, and prices convert from a rate table the agency itself controls. In a market where the rate your bank gives you and the rate on a vendor's feed can differ, owning that table is the difference between a margin and a loss. The mechanics, and where margin quietly leaks, are in our piece on multi-currency booking sites.

Bandwidth decides more than design does

Your customer is on a phone. Often a mid-range Android phone on mobile data, sometimes outside the city, usually opening your link from inside Telegram. Every large hero image, autoplay video and third-party script on the page is paid for by that customer in seconds of waiting. A lightweight storefront wins here not because it looks modern but because it finishes loading before the customer goes back to the chat.

The practical test is cheap. Take the phone your typical customer owns, switch off Wi-Fi, go somewhere the signal is ordinary, and open a search result page and the passenger form. Time it. Then do the same with the vendor's demo. If a template system lets you change the look from the admin panel without a redeploy, as it does here, pick the plainest template first and add decoration only after you have watched it load on that phone.

What getting Central Asia travel trade technology wrong costs

The cost of a poor fit rarely shows up as one big failure. It shows up as the nephew who closed the link, the sub-agent who went back to sending you bookings on WhatsApp because the portal could not hold his credit line, the corporate client who asked for an invoice your system cannot produce, and the Saturday evening a staff member spends re-keying cash bookings that should have been entered once. Each is small. Together they mean you pay for a platform and still run the business by hand, and in a few months you are migrating again, with customers who have learned your site is not the place to book.

Agencies in other corridors have walked this road; the pattern in South Asia, where the trade also runs on sub-agents, is close to what Tashkent and Almaty look like. The ones that came out well tested against their own routes, their own payment habits and their customers' own phones before they signed. If you want to run those tests on a real site rather than a demo, the signup wizard on tekravel.com provisions a branded site on a platform subdomain from the form itself, without a card, so you can open it on that phone this afternoon.

Tekravel Editorial

Travel technology desk

The Tekravel travel-technology desk writes for the trade: agency owners, consolidators and the developers who integrate them. Every article is checked against the platform it describes before it is published.