The Role of SMS Aggregators in Modern A2P Messaging

There is no global mobile network. There is no single authority that can deliver a text message to any handset on earth, and nobody is building one.

Yet a bank in Toronto can send a verification code to a customer travelling in Nairobi, and it arrives in a few seconds, and the bank’s engineering team knows almost nothing about how. That works because a layer of intermediaries assembled the connectivity that no single operator has. SMS aggregators are that layer, and their function has changed substantially in the last decade, mostly in ways their own marketing has not caught up with.

The old description of the role was arbitrage: buy termination cheaply, sell it dearer, keep the difference. That still happens. It is no longer the interesting part, and treating it as the whole story leads enterprises to buy the wrong things and operators to misjudge who they are dealing with.

The problem the layer exists to solve

Start with the arithmetic that makes aggregation inevitable.

There are several hundred mobile networks operating across more than two hundred countries and territories. An enterprise that wants global messaging reach would need a commercial agreement, a technical integration and a reconciliation process with each of them. No operator wants several thousand such relationships with individual businesses either, most of which would generate trivial revenue.

So the market did what markets do with N-to-N problems. It inserted a layer. Operators sign with a manageable number of counterparties, enterprises integrate once, and the aggregator absorbs the combinatorial mess in between. Nothing about this was designed centrally. It accreted, contract by contract, which explains both its resilience and its opacity.

Four roles aggregators actually perform

Translation, technical and commercial

The obvious role is protocol conversion: an HTTP request becomes an SMPP submission, which becomes a MAP operation somewhere downstream, and a delivery receipt travels back through the same chain in reverse.

The less obvious translation is commercial. Operators sell termination in wholesale terms with volume commitments and corridor-level pricing. Enterprises want a per-message rate and a monthly invoice. Someone has to hold the mismatch between those two models, carry the working capital, and absorb the risk when a customer’s volume forecast turns out to be fiction. That is why an aggregator’s SMS platform is as much a rating and reconciliation system as a message router. The routing is the easy half.

The compliance layer

This is the role that has grown fastest, and it is the one most enterprises underestimate until they enter a new market.

A message sent into the United States now needs brand and campaign registration under the 10DLC framework the major carriers administer through The Campaign Registry, or a verified toll-free number, or a short code. India routes commercial traffic through registered headers and templates on a distributed-ledger system. Singapore requires registered sender IDs and labels unregistered senders to recipients as likely scams. In the UK, brands register the sender IDs they use through the Mobile Ecosystem Forum’s SenderID Protection Registry so operators can filter impersonations. Australia’s regulator brought in an industry code obliging telcos to identify, trace and block scam SMS, and has been moving toward a sender ID register of its own.

Each of these regimes has its own registration workflow, evidence requirements and rejection reasons. An enterprise operating in fifteen markets either builds a team to manage that or relies on its messaging provider to do it. The aggregators worth their margin have become, functionally, compliance operators who also move messages.

Traffic quality gatekeeping

Aggregators sit at a chokepoint. Every message crosses their platform before it reaches a network, which makes them the earliest place where bad traffic can be stopped.

The incentive to use that position has strengthened. Operators now measure quality per sender and per connection. A provider that consistently delivers spam, phishing content or artificially inflated traffic gets throttled, then priced worse, then disconnected. Reputation at the interconnect has become an asset with a book value, and providers protect it by filtering their own customers.

That is a genuine shift. Fifteen years ago the aggregator’s interest was volume. Now a large part of the business is refusing volume.

Artificially inflated traffic is the clearest test of it. When automated clients hammer a signup flow to manufacture verification sends, the enterprise pays for messages nobody reads and somebody downstream collects. The awkward part is that the aggregator is in the middle of that transaction, billing for traffic it may also be positioned to detect. Enterprises noticed, started demanding credits, and providers responded by building pumping detection into their verification products and in some cases offering contractual protection against inflated volume. Whether a provider will put that in a contract is a reasonable proxy for whether it takes the gatekeeping role seriously.

Price discovery in a market with no exchange

Wholesale messaging has no public order book. Rates move by corridor, by month, by counterparty, and the only participants who see the whole picture are the ones trading across many routes at once. Aggregators perform price discovery for a market that never built an exchange, which is useful to everyone and also explains why enterprises so often cannot tell whether their rate is reasonable.

What changed: firewalls, registries and the end of easy arbitrage

The grey route economy defined the previous era. Commercial traffic disguised as ordinary person-to-person messaging avoided A2P termination fees, and the savings funded a whole segment of the market.

Operators eventually did the obvious thing and deployed inspection at the border. Modern SMS firewall deployments identify traffic that has been dressed as P2P, block it or reclassify and bill it correctly, and enforce sender identity against registries. Home routing closed the gap where terminating messages could be handled outside the home network. The arbitrage did not disappear, but its risk profile inverted: the cheap route is now the one that gets blocked mid-campaign, with the enterprise discovering it through a drop in conversion.

That has consequences for the aggregator’s role. Value has migrated from “who can source the cheapest termination” toward “who can guarantee that a message arrives intact, from the sender ID it was submitted with, in a market whose registration regime is satisfied.” Those are different competencies, and not every provider made the transition.

The accountability question nobody has settled

When a customer receives a text that appears to come from their bank and loses money to it, who is responsible?

The enterprise never sent it. The operator delivered it. The aggregator submitted it, possibly on behalf of a reseller, possibly on behalf of a reseller’s customer. Somewhere in that chain, an entity accepted traffic it should have refused, and reconstructing which one requires cooperation from parties with no obligation to cooperate.

The chain’s length is what makes this hard. An aggregator may know its direct customer well and know nothing about that customer’s customers, and resale can run two or three layers deep. Know-your-sender discipline at the first hop is the only practical control, and it is uneven across the market. A provider that cannot tell you who originated a given message is admitting something about its own onboarding.

Regulators have started pushing responsibility toward the parties who can actually see the traffic, which in practice means operators and the aggregators connected to them. Sender ID registries are one expression of that: they create a checkable record of who is entitled to use a name, which converts an unanswerable question into an auditable one. Expect more of this rather than less, and expect it to keep raising the compliance floor for anyone in the delivery chain.

Pressure on the model

Three forces are reshaping the layer at once.

Direct network APIs are the most consequential. Under GSMA’s Open Gateway initiative and the CAMARA project, operators are exposing standardised interfaces for functions including number verification and SIM swap checks. Ericsson and a group of major operators went further and formed a joint venture, Aduna, to sell network APIs across markets. Where silent number verification is available, an enterprise can authenticate a user without sending a message at all, which removes the aggregator from a use case that has been one of its highest-volume revenue lines.

Consolidation is the second. The market has spent a decade absorbing itself: Infobip acquired OpenMarket, Sinch acquired both Inteliquent and SAP’s messaging business, Ericsson bought Vonage. Scale matters more when the value is compliance coverage and direct interconnects rather than clever routing, and that logic favours fewer, larger players.

Channel expansion is the third. RCS gives brands richer content and verified sender presentation, and aggregators are extending platforms to carry it. The commercial model there is still forming, and the security model around verified senders will matter more than the media capabilities.

None of this eliminates the layer. Verification is only one messaging use case, and network API coverage will be partial for years. But an aggregator whose value proposition is still fundamentally “we have routes” is describing a business with a shrinking moat.

What good looks like now

For enterprises evaluating the layer, the useful signals have shifted along with the role.

Ask what the provider does when a market’s registration regime changes, and whether they will name the last time it happened and how they handled it. Ask what they refuse to send, because a provider that accepts anything is telling you about the company it keeps at the interconnect. Ask where their receipts originate and which corridors are direct rather than resold. Look at whether they can pin a route for verification traffic while allowing dynamic routing elsewhere, since that single control separates providers who treat routing as a shared decision from those who treat it as their own margin lever. Anyone assessing where A2P messaging is heading should weight compliance capability and interconnect quality well above headline rate.

For operators, the corresponding question is which aggregators improve the quality of traffic entering the network and which ones are simply a pipe. That is now measurable, and it belongs in commercial conversations rather than only in the fraud team’s reports.

Key takeaways

• Aggregation exists because there is no global network and no operator wants thousands of direct enterprise relationships. That structural reason is not going away.
• The role has moved from route arbitrage to compliance operations, traffic quality gatekeeping and interconnect reputation management.
• Operator firewalls and sender ID registries inverted grey route economics. The cheapest route is now the one most likely to be blocked mid-campaign.
• Registration regimes differ by market and are proliferating. Judge providers on whether they operate them rather than pass them back to you.
• A provider that refuses traffic is protecting the interconnect reputation your delivery depends on. Treat that as a feature.
• Network APIs for silent verification will erode the highest-volume A2P use case first. Plan for verification and marketing traffic to diverge.
• Enterprises should evaluate compliance coverage, receipt provenance and routing control ahead of per-message price.
• Operators should treat aggregator traffic quality as a commercial metric, not only a security one.