Airline consolidator in Dubai: the sub-agent portal question

It is eleven at night in Dubai and a sub-agent in Karachi sends a screenshot of a fare with two words: still available? Somebody on your desk checks, types a number back, and hopes it is still true when the customer says yes. Tomorrow the same exchange happens with Nairobi, Riyadh and Dhaka. For many an airline consolidator in Dubai this is not a failure of the business. It is the business, and it has worked for years.
It stops working at a particular size, and the size is not about ticket volume. It is about how many people at your end have to remember what was promised to whom.
Why Dubai makes this harder than it looks
A wholesaler in most cities sells to agents who share a currency, a language and a banking week. A Dubai consolidator rarely has that luxury. The sub-agent base spreads across the GCC, South Asia and Africa, and each corner has its own working hours, its own way of paying and its own preferred language for a quote. So settlement, currency and language are not extras bolted onto the product. They are the product the sub-agent experiences.
That is also why a generic booking site does not solve it. Selling flights to the public is one problem. Handing a net fare to a business that will add its own margin, on credit, in a different currency, is another.
What a sub-agent portal has to do for a Dubai wholesaler
Four jobs, in the order sub-agents tend to complain about them:
- Show each agent their own price. A sub-agent should see a fare built for them, not a published fare they must ask you to improve.
- Hold a credit limit and enforce it. The portal, not a person remembering a conversation, decides whether the next booking goes through.
- Settle in a way both sides can read. What the agent owes you is a document with lines on it, not a figure in a message.
- Speak the agent's language and currency. Arabic, Urdu and Hindi at minimum for this base, with prices shown in the agent's own money.
On this platform the pieces exist as features rather than workarounds. Storefronts can require sign-in before search, register business customers and verify them manually if you want to approve each one. Sub-agent credit limits and settlement invoicing are built in, not spreadsheets. The storefront ships in 40 languages, right-to-left ones included, and each site picks its default currency and the list it offers, with prices converted from a rate table kept per site.
What it does not do is decide your commercial terms. How much credit an agent deserves, and what their price should be, remain your calls. The portal only makes sure they are applied every time.
WhatsApp and a spreadsheet against a portal, operation by operation
| Operation | WhatsApp and spreadsheet | Sub-agent portal |
|---|---|---|
| Quote | A number typed from a desk, valid until the fare moves | The agent searches and sees their own price at that moment |
| Hold | A message saying it is held, and a reminder to watch the time | A booking record with its own deadline, visible to both sides |
| Issue | Someone re-keys the details, from a chat, into the system | The sale is recorded once with customer, supplier, cost and markup attached |
| Refund | A thread you scroll to find who was promised what | One record, so the amount follows the booking |
| Statement | Built by hand at month end, then disputed | A settlement invoice per sub-agent, and the credit limit already applied |
Be fair to the left column, because plenty of good wholesalers still live there: it is fast for a small circle of agents who trust you, and it asks for no training. The problem is not any single row. It is that every row depends on one person's memory, and memory does not scale with the number of agents, the number of currencies or the number of nights you are asked to answer.
Where manual issuance exposes you to an ADM
An Agency Debit Memo is the airline telling you that a ticket was issued in a way that broke its rules, and asking for the difference. Manual issuance is a steady source of them for ordinary reasons: a fare quoted from a screenshot that had already changed, a name keyed with a typo, a duplicate booking made because the first confirmation was lost in a long thread.
The memo usually lands on the agency whose IATA number issued the ticket, which is you. Then you try to recover it from a sub-agent, and the whole dispute rests on a chat history that neither side can fully reconstruct. A portal does not stop an airline from raising a memo. It leaves a record of what was quoted and issued, which is what you need to argue one.
What it costs to stay manual
The visible cost is staff time at eleven at night. The invisible one is the sub-agent who moves.
There is a second, plainer risk. In a network run on chat, one person usually knows which agent was promised which price and which one is quietly over their limit. When that person is on leave, ill or hired away by a competitor, the knowledge leaves the building the same afternoon. A portal is dull for exactly this reason: it keeps the rules where anyone on the desk can read them, so a good week does not depend on one memory.
Think about the agent in Nairobi who messages you and two other wholesalers. One of them has given her a login. She searches at her own pace, sees her price, sees what she owes and does not wait for anyone to type. She does not leave you because you are more expensive. She leaves because the other desk answers faster than you can, at any hour, without being asked. By the time you notice, her volume is not slowing down. It is simply elsewhere.
Where to start without rebuilding everything
You do not need to move every agent in one week. Move the ones who write to you most often, keep the rest on chat for a while, and let the portal's records replace the spreadsheet one agent at a time.
If you want to see how a consolidator account is opened, start from the consolidator page; it covers selling white-label sites to your sub-agents and sharing your supplier content with a network. If your own developers will connect to inventory directly, the flight and hotel APIs begin with a conversation rather than a self-serve key.