Skip to content
Blog

The South Asia travel consolidator market runs on sub-agents

Tekravel Editorial
Written with the help of AI

It is past midnight in Dubai and the phone is still going. An agent in Lahore wants a net fare on Dubai–Lahore for a family of six leaving Thursday, asks whether his credit will stretch to it, and sends a screenshot of a passport rather than the passport details. He is not your customer. He buys from one of your sub-agents, who buys from you, and the ticket will be issued on your plate against your BSP. That chain, and the credit that runs along it, is the South Asia travel consolidator market in one message. Anyone who has tried to put a booking portal in front of it knows how little of it survives first contact with a search box.

This piece is for the consolidator — in the Gulf or in the subcontinent itself — who distributes net fares into India, Pakistan, Bangladesh, Sri Lanka or Nepal and is weighing what technology to put between themselves and the agents who sell for them. The argument is simple: the sub-agent model is not a transitional phase waiting for an app to end it. It is the distribution. Technology that refuses to accept that gets ignored, and technology that accepts it has to make a few concessions a Western product manager would call unreasonable.

How deep the chain actually goes

A consolidator who holds the airline contracts and the BSP settlement sits at the top. Below is a layer of larger agencies in Karachi, Mumbai, Dhaka or Colombo that buy on credit and resell. Below them, agents with a shopfront and a printer. Below those, people with a phone, a Facebook page and a reputation in one neighbourhood, who are the ones who actually meet the traveller. A ticket passes through each of those pairs of hands between the airline's fare and the passenger's WhatsApp, and every pair adds a markup and takes a slice of the risk.

Two things follow. First, the consolidator at the top rarely knows who the passenger is until the PNR is created, and often not even then — the name on the booking is real, the phone number on it belongs to the agent. Second, the price the traveller pays bears only a loose relationship to the net fare. Each layer prices to its own market: a Gulf route sold in Kerala during the school holidays carries a different margin from the same seat sold to a labourer's family in Sylhet. A single published markup table, the kind a Western OTA applies at the top and shows to everyone, flattens exactly the thing that makes the chain work.

The chain is also how the market handles trust. The small agent in the neighbourhood is trusted because he is known; the agency above him is trusted by the consolidator because it has paid on time for years. Nobody along the chain has a contract with the airline, and nobody needs one. That is the point of the consolidator.

It is tempting to read this as inefficiency to be engineered away. Try it and you discover that each layer is doing work the layer above cannot: collecting cash in a town the consolidator has never visited, vouching for a passenger with no card, explaining a fare rule in Sindhi. The depth is not a bug.

Credit is the relationship, not a feature

In a card-driven market, a booking is paid when it is made. In this one, a booking is a promise. The sub-agent issues against a limit you have extended, settles at an interval the two of you agreed, and the size of that limit is the most honest statement of how much you trust them. Raising it is a reward. Cutting it is a sanction that lands faster than any email. Agents move consolidators over credit far more often than over a fare difference, because credit is the thing that lets them sell on a Friday when their own cash is tied up in last week's tickets.

So the ledger is the product. The consolidator who manages limits on a spreadsheet and settles from bank screenshots is not being backward; they are doing by hand what nothing they were offered did properly. On our side, a consolidator account treats sub-agent credit limits and settlement invoicing as first-class concepts rather than a field bolted onto a payment page, and the practical rules for setting and policing those limits are in our piece on sub-agent credit limits. What matters for this market is narrower: the limit has to be visible to the agent before they search, the hold or issue has to be refused at the limit without a phone call, and the settlement statement has to be something an agent in Sylhet can reconcile on a phone screen.

One more thing about credit that the design has to accept. In South Asia the limit is often held not by the person doing the searching but by their boss, or by the agency above them. A system that assumes one login equals one wallet will be worked around within a week by shared passwords. Roles inside the sub-agent's own operation — who can search, who can hold, who can issue against the limit — are not an enterprise nicety here. They are what stops a trainee from spending the owner's week.

A phone on a weak connection is the office

The agent who meets the traveller does not sit at a desktop with fibre. He is on a mid-range Android handset, often on a connection that drops between the search and the results, in a market where data is bought in small bundles and a heavy page is a cost. He switches between your portal, WhatsApp and a bank app, and he screenshots everything because screenshots survive the connection dropping.

What that demands is dull and non-negotiable. The search result has to arrive in pieces and survive a refresh. The fare rules have to be readable at phone width without pinching. A booking that was half-made when the signal went has to be there when it comes back, not lost. And the language has to be his: Urdu reads right to left, Hindi and Bengali have their own scripts, and a storefront that ships Urdu as a proper right-to-left layout rather than an afterthought is one he will actually use at ten at night. The storefront we provision ships 40 languages with right-to-left support for Urdu among them, with the tenant choosing which to offer — which is a start, not the finish, because the agent's own reseller brand is what the traveller recognises.

What fails when you import a Western OTA design into the South Asia travel consolidator market

Most portals sold into this market were designed for a direct-to-consumer customer with a card, then had a "B2B mode" switched on. The assumptions beneath them do not switch off. The table below is the list we keep hearing from consolidators who tried one; it is a table you can argue with, and the rows are in the order they usually hurt.

Imported assumptionWhat the chain actually doesWhat has to give
Search is public; sign-in happens at paymentNet fares are confidential; an open search box leaks them to the next agent down the streetSign-in before search, with business customers verified by a human before they see a fare
One markup table, set at the topEvery layer prices its own marketMarkups the sub-agent controls for their own customers, separate from yours
Payment at booking, by cardIssue against credit, settle later, often in cash collected two layers downA limit that blocks issuance, and statements the agent can reconcile
One login, one walletThe limit belongs to the owner; the searching is done by staffRoles inside the sub-agent's own office
The passenger is the customerThe agent is the customer; the passenger may never see the brandReseller branding, with the consolidator invisible to the traveller
Desktop first, mobile adaptedPhone first, on a connection that dropsResults that arrive in pieces and bookings that survive a refresh
English, with translation as a toggleUrdu, Hindi, Bengali, Tamil, Sinhala — and the fare rules in themRight-to-left and Indic scripts as layouts, not strings

Every row traces back to the same mistake: treating the sub-agent as a slightly unusual consumer rather than as the distribution channel. The consolidators who got technology to stick did it by picking the thing that treated the agent as the customer — the same shift we describe for the Gulf trade in what agency software in the GCC has to get right, with the credit dimension turned up.

What getting this wrong costs you

The first cost is quiet. The portal goes live, the sub-agents are trained, and within weeks they are back on WhatsApp asking you for fares by message, because the portal showed them a price they could not resell and would not let them issue against the limit they thought they had. You now pay for a system and for the manual loop it was meant to replace.

The second cost is louder. A sub-agent whose limit the system did not enforce issues more than they can settle, and the debt sits with you against your BSP, with no mechanism for recovering it except the relationship itself. That is the ADM-shaped hole in the accounts that consolidators in this market recognise on sight. How net fares are handed down without creating it is the subject of our guide to distributing net fares to sub-agents.

The third cost is the one that rarely gets counted: the agent who leaves. Credit, speed and respect are what hold a sub-agent to a consolidator, and a portal that gets credit wrong, is slow on a phone and talks to them in the wrong language removes all three at once. The next consolidator's WhatsApp number is already in their phone.

What the technology has to accept

None of this means the chain cannot be put online. It means the online version has to be built around what the chain already is: credit as the relationship, the agent as the customer, the phone as the office and the markup as something every layer owns. A consolidator account on this platform is opened to sell exactly that way — white-label sites for the agents below you, your supplier content shared to the network you choose, limits and settlement in the ledger rather than on paper — and growing that network is a question of which agencies you extend credit to, not which features you turn on.

The test, in the end, is the midnight message from Lahore. If the agent can see his limit, search a confidential fare, hold it for the family of six and issue it in Urdu on a phone that drops the connection twice — and if you can see all of it in your ledger in the morning — the technology has accepted the market. If he is still asking you by WhatsApp, it has not.

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.