Wholesale SMS aggregator business: routes, margin, MNP, suppliers

A wholesale SMS aggregator buys termination from carriers and other aggregators, sells it to resellers, enterprises and other aggregators, and keeps the spread. It is one of the few businesses in telecom that can still be started by a small team with contracts, a server and discipline, and it is also one where the margin is thin enough that a wrong route or an unchecked supplier eats a month’s profit in a week. This guide covers how the wholesale SMS aggregator business actually works: what you buy, how you tier and route it, where the margin lives, why ported numbers matter, how to choose suppliers, and the honest line between the aggregator you can build on a self-hosted platform and the wholesale trader who needs something bigger.

What a wholesale SMS aggregator actually sells

The product is termination: the ability to get a message onto a phone in a given country, on a given operator, with a sender ID that survives and a delivery receipt that is true. Every part of that sentence is a variable that changes the price. A route that terminates directly on the operator with the sender ID intact is worth more than one that goes through two intermediaries and rewrites the sender to a random number. A receipt that reflects what the handset saw is worth more than one the supplier generated because it was cheaper than asking the operator.

Clients buy from an aggregator for reach and price: one contract and one bind instead of thirty carrier relationships, at a price lower than they could get with their own volume. Your job is to hold those thirty relationships, or the subset that matters, and to sell access to them with a margin on top. Everything else in this guide is how to do that without losing money.

Routes: direct, secondary and grey

A route is a supplier connection that reaches a set of destinations at a price list. Three kinds exist and you should know which each of yours is.

Direct routes are binds to the destination operator or to its official partner, on a contract with a price per operator, a deposit and usually a minimum monthly commitment. They deliver reliably, keep the sender ID, and produce receipts you can trust, because the operator has no reason to lie about its own delivery. They are also the routes clients pay the most for.

Secondary routes are other aggregators. You bind to them the way your clients bind to you, and they carry your traffic onward over their own direct and secondary routes. They give you reach into destinations you send too little to for a direct deal, at a higher unit price and with quality you must measure, because you cannot see what happens after their platform.

Grey routes are the cheap ones: traffic pushed through SIM boxes, consumer bundles, or interconnects that were never meant for application traffic. The price is a fraction of a direct route’s; the delivery is unreliable, the sender ID is lost, operators block them without notice, and in many countries they are illegal. An aggregator who sells grey routes as “economy” is one complaint from losing a client and one investigation from losing the business. The discipline is to know exactly which routes you have, and to label them truthfully to the clients who buy them.

Quality tiers: sell the route, not the destination

The first structural decision is to stop selling “SMS to country X” at one price. Sell tiers. A premium tier of direct routes with sender ID preserved and true receipts, for OTPs, banking and anything time-critical. A standard tier for transactional and campaign traffic where a slightly lower delivery rate is acceptable at a lower price. Possibly an economy tier for bulk marketing, if you can source it from routes you are prepared to defend.

Tiers do three things. They let a client choose a price they understand. They let you place each supplier route in the tier its measured quality earns, rather than in the tier its price suggests. And they turn “your delivery was bad” into “which tier did you buy”, which is the beginning of a rational conversation instead of a refund.

The measurement is not optional. Delivery rate, receipt latency, and sender ID survival per route per destination, tracked continuously, decide which tier a route belongs in this month. A supplier whose route quietly degrades from direct to grey (it happens; they lose a contract and re-route without telling you) shows up first in your receipt statistics, and only later in your client’s complaint if you are watching.

Least-cost routing without routing into a loss

With several routes per destination, the platform picks one per message. The simple rule, cheapest first with failover to the next on error, is called least-cost routing (LCR), and for most aggregators it is the right rule for years. Two refinements make it safe.

First, LCR runs within a tier, not across tiers. The cheapest route to a country is never the right route for a client who paid for premium. The routing table is per tier, and within the tier it is ordered by price with quality as a tiebreaker.

Second, margin floors. Your sell price to a client is fixed on a rate card; your cost changes when a supplier updates their price list, which they do weekly and sometimes with hours of notice. If cost rises above sell price for a destination and nobody notices, every message to that destination is sold at a loss. The platform must therefore know the cost of the chosen route at send time, record it against the message, and be able to show you, per client and per route, where the margin went negative. That is the “frozen at send” idea behind the Sales & Profit view on the revenue page: the cost is stamped on the message with the price, so the margin you see is the one you earned, not an estimate from last month’s rate card.

The aggregator who loses money rarely loses it on a bad month. They lose it on one destination, on one route, for three weeks, because nothing on the screen turned red.

Wholesale trading at scale, where you buy the same destination from more than five suppliers and change the mix during the day based on live scoring, is a different tool and a different business; the enterprise platforms guide draws that line honestly. If you have three contracts and a preferred route, you need LCR with failover and margin floors, and a self-hosted platform gives you that.

Where the margin lives

The spread per message is small, often measured in ten-thousandths of a dollar on a destination that costs a few thousandths, so the business is volume times discipline. Three places decide the number.

Buying. A direct contract on your top three destinations, negotiated on committed volume, usually lowers cost more than any routing trick. Know your traffic mix by destination before you sign anything; commit only on destinations where your clients’ volume is steady.

Selling. Rate cards per client, per destination, in tiers, with prepaid balances so you are never funding a client’s traffic before they pay. Big accounts negotiate; the rate card for everyone else is where the aggregate margin is made. Review it monthly against cost, not annually.

Leakage. Failed messages you paid for and could not bill, receipts that arrived too late to charge, ported numbers sent to the wrong operator and billed at the wrong price, and destinations quietly below cost. Leakage is invisible in a traffic report and obvious in a margin report, which is why the margin report is the one to look at every morning.

MNP and HLR: the ported-number problem

Number portability breaks prefix-based routing. A number that starts with operator A’s prefix may have been ported to operator B years ago; route it by prefix and it goes to A, who either rejects it (a failure you paid for) or forwards it to B at an interconnect price that is not the one on your rate card. In markets with high porting rates the error is a measurable share of all traffic.

Two tools fix it. An MNP database maps numbers to their current operator, updated from the national porting registry or a commercial feed, and the platform consults it before choosing a route. An HLR lookup asks the network live whether a number exists, which operator serves it and whether it is reachable; it costs a fraction of a cent per query and is worth it for cleaning a marketing list before paying to send to numbers that are dead. Smppcube handles MNP data in its archive layer (MongoDB, alongside MySQL as the system of record, as the platform page shows) at the level a reseller or aggregator needs; wholesale-scale HLR through a signalling interconnect is the enterprise tier’s territory.

The commercial point is that MNP is a margin feature as much as a delivery feature. Correct operator resolution means the right route, the right cost and the right receipt, and it removes one of the four leakage sources above.

Choosing and testing suppliers

Every supplier, direct or secondary, gets the same treatment before a client’s traffic touches them:

1. A written price list per operator, with notice terms for changes, and the tier they claim for each destination.
2. Test traffic to real handsets in the destination country, from several sender IDs, at different times of day: did it arrive, how fast, with which sender, and did the receipt match what the handset saw? Probe testing can be bought as a service if you have no handsets there.
3. A DLR honesty check: compare the supplier’s delivered rate with the handset results. A supplier whose receipts say 98 percent delivered while handsets saw 80 percent is generating receipts.
4. Commercial terms: prepaid or postpaid, deposit, minimum volume, settlement currency and how disputes are handled.
5. Continuous monitoring after go-live, because the route you tested in week one is not guaranteed to be the route you have in week ten.

Keep at least two suppliers for every destination you sell, in the same tier where you can. A destination with one route is a destination you cannot promise.

The platform underneath

Everything above assumes the platform can do five things: hold many supplier binds over SMPP and HTTP at once, route per destination and per tier with failover, stamp cost on every message and show margin per client and per route, bill clients on rate cards with prepaid balances, and let your own clients bind over SMPP under their own brand. Those are the core of what Smppcube builds for SMS aggregators, on a one-time licence with no per-message platform fee, which matters in a business where the platform fee would otherwise be a percentage of the very spread you are trying to keep.

The honest counterpoint

Not every aggregator should be wholesale. If your clients are your own contracted resellers and enterprises, you buy on contract and sell retail, and the “wholesale” parts of this guide (supplier policing, reconciliation, fraud exposure from traffic you do not control) are problems you have chosen not to have. That is a good business, and a simpler one. Go wholesale, taking traffic from other aggregators and trading routes, only when your buying power on specific destinations is real enough that others want to buy through you, and when you have the margin view and the monitoring to police it. The tools for the first business are on the platform page; the tools for the second start there and grow with the contracts.