In a January steering meeting, a Tier-2 MNO in Southeast Asia (~28M subs) faced a binary: expand IMS and run its own MaaP stack, or anchor on Jibe. The sticking point was not feature parity, but the unit economics of MaaP interconnect, A2P routing control, and who owns the enterprise contract. The rest followed from that, from OCS logic to price floors that would not cannibalise SMS.
Inside the operator room: where the RBM P&L is decided
The case was a Tier-2 MNO, Southeast Asia, ~28M subscribers. The operator had an existing enterprise messaging book, a defensible domestic SMS termination position, and a board mandate to grow non-access wholesale revenue without creating a new arbitrage channel. The RBM launch was therefore not treated as a product overlay. It was treated as a margin architecture decision.
The first path was control. Expand IMS capacity, add SBC, TAS and MRF headroom, procure a MaaP layer, and stand up agent review, anti-abuse, analytics and reporting. The internal estimate was 12–15 months before a commercially useful launch. That route carried capex, procurement friction and integration risk. It also preserved routing control, richer event data, tighter enterprise SLAs and lower long-run COGS if utilisation reached plan.
The second path was speed. Anchor on the Jibe RBM hub, accept hosted economics, and launch in 4–6 months with a lighter integration load. The commercial trade was visible on the spreadsheet before any agency pitch entered the room: per-conversation or MAU opex, hub fees, less granular routing discretion and a narrower claim over agent-level data. Reach improved quickly, but the operator ceded part of the yield stack.
The debate then moved from platform to channel. Aggregators would still be needed for long-tail enterprises, particularly retailers, delivery platforms and regional banks that already buy messaging through CPaaS contracts. The top-50 domestic accounts were different. They had existing A2P commitments, known compliance profiles and enough volume to justify direct contracting. The operator’s wholesale team wanted the option to route those accounts outside aggregator take-rates, while still using hubs for reach and operational elasticity.
That split made brand verification an operational issue, not a launch checklist item. Every path needed KYC, agent vetting, template review, abuse handling and lifecycle governance. The hosted route reduced some integration work, but it did not remove accountability when a verified agent sent a phishing flow or a regulated sender disputed billing. In practice, brand operations had to scale with the sales plan.
OCS alignment was the point at which the room became quiet. RBM does not rate like bulk SMS if the offer is based on conversation windows, bundled fallback and differentiated rich media payloads. Finance wanted a single billable event that could survive disputes. Wholesale wanted the freedom to include SMS fallback without creating SMPP bill shock. Product wanted vertical price tiers. Without a common charging model, each margin gain in the route table could be erased in reconciliation.
The board did not ask for message volumes as the primary KPI. It asked for gross margin per conversation versus SMS, attach rate to the existing A2P book, verified-brand growth, and payback at 50–60% load. That was the correct framing. RBM economics are set before go-to-market starts. Architecture decides the cost floor. Interconnect decides the take-rate. Rating decides whether revenue survives the first billing cycle.
Interconnect and routing define unit economics
The broader pattern is now visible across several operator discussions. Aggregator hubs expand coverage and compress time-to-market, but they introduce an economic layer that rarely disappears later. A 15–35% commercial share, plus hub fees and settlement costs, can be rational in a launch phase. It is less comfortable once the same route carries high-volume banking, travel and authentication traffic that the operator could have contracted directly.
Direct enterprise onboarding through an MNO-hosted MaaP changes the equation. It avoids an aggregator share, improves access to event data, and gives the operator more control over agent policy and quality. It also shifts cost into KYC, vetting, template workflow, sales engineering, misuse response and account management. The result is not automatically cheaper. It is cheaper only where account concentration, compliance maturity and traffic predictability justify the operating model.
Peer MNO interop is not a shortcut for RBM agent reach. It helps P2P messaging coherence, but cross-operator enterprise agent routing still lands on agreed hubs or bilateral MaaP peering. Cross-border traffic adds further settlement complexity. A domestic operator can control its own subscriber base and enterprise contracts, but it cannot assume that a rich message from a regional airline or bank will traverse foreign networks at domestic economics.
Routing policy is therefore a margin lever. Capability discovery must be accurate enough to avoid unnecessary fallback. Fallback must be priced inside the bundle or explicitly excluded. NEF/SCEF gating matters where enterprise messaging begins to touch richer network functions. Operators also need controls that suppress OTT bypass and unmanaged traffic, particularly where agencies or aggregators try to optimise around formal capability-discovery rules.
Fraud and abuse economics should be priced as part of the route, not treated as a security overhead. Spam filters, phishing detection, agent attestation and emergency suspension workflows all carry cost. The liability position must be explicit in aggregator MSAs. If a post-event investigation forces refunds, remediation credits or regulatory reporting, the apparent margin on the campaign was never real.
A Tier-2 head of wholesale recently observed, “RCS is sold in the contract, but won in the OCS.”
Pricing constructs that prevent SMS cannibalisation
SMS cannibalisation is not prevented by positioning RBM as premium. It is prevented by pricing constructs that keep floors intact. Conversation-based pricing is the most workable model, provided the operator defines the window, the event trigger, the included fallback and the exception cases before contracts reach procurement. A 24-hour customer-care conversation is not the same product as a 72-hour travel disruption flow with media, buttons and fallback.
Vertical pricing matters because willingness to pay and regulatory exposure differ sharply. Banking, insurance and public-sector traffic often carries higher assurance requirements and lower tolerance for failure. Retail campaigns are more elastic and more exposed to agency substitution. Travel and logistics sit between the two, with high value during disruption events and lower value in routine notification flows. A flat rate leaves too much money on the table in regulated use cases and too much risk in promotional traffic.
Bundled fallback is usually cleaner than separate SMPP billing. The enterprise buys an outcome within a conversation window. The operator rates RCS delivery, SMS fallback and exception handling under a single commercial event. This reduces disputes and makes gross margin legible. It also forces the operator to set a hard floor, because every fallback SMS has a real network and interconnect cost.
Verified-brand fees should be treated as recurring access economics, not waived as launch garnish. A per-agent monthly fee pays for review, attestation, policy updates, reporting and misuse response. Waivers can be justified for regulated public-sector senders or emergency communications, but only with capped volumes and defined use cases. If brand verification becomes free for high-volume commercial senders, the operator gives away one of the few parts of RBM where it has defensible control.
SLA tiers provide another pricing surface. Throughput, template approval turnaround, analytics granularity and support response can be priced without disturbing base message economics. Rich media payload bands should also be monetised. A text-only notification, a carousel with images, and an authenticated customer-service flow do not impose the same cost or create the same value for the enterprise.
Wholesale to CPaaS needs discipline. Set MTR equivalents or commercial minimums where the market uses that construct. Enforce MFN terms carefully, because broad clauses can lock the operator into rates that do not reflect route quality or data access. Anti-bypass clauses should be tied to capability-discovery audits and traffic source transparency. Otherwise, the operator will price RBM as a managed channel while accepting traffic behaviour closer to unmanaged messaging.
Geography closes the model. Taxes, consent rules, data residency and consumer protection frameworks alter the real cost of a conversation. A regional enterprise footprint does not justify one regional price if local obligations differ. Price per market, then negotiate enterprise discounts against verified volumes and route commitments. That keeps the economics visible and prevents cross-subsidy from masking weak markets.
Go-to-market: aggregator-led, hybrid, and direct enterprise
The practical go-to-market answer is usually hybrid. Aggregators provide breadth, low acquisition cost and early volume. They also limit margin and reduce control over agent behaviour, analytics and enterprise relationship depth. That is acceptable for the long tail. It is less acceptable for a domestic bank, airline, marketplace or public-sector sender that already sits inside the operator’s enterprise account plan.
Direct enterprise sales should not be built as a separate RBM island. It should sit beside existing A2P, DCB and identity conversations, because the buyer often funds customer engagement from the same budget. The sales motion needs vertical playbooks, but the commercial backbone is technical. Account teams must understand MaaP constraints, OCS rating, fallback policy and data reporting well enough to avoid promises that operations cannot honour.
Hybrid guardrails are commercial instruments. MFN rate cards reduce channel conflict. Co-marketing funds can support aggregator demand without giving up named strategic accounts. Top-50 reservations should be explicit, not handled informally after a reseller has already opened the door. Where an enterprise buys through a global CPaaS contract, the operator can still negotiate domestic route commitments, verified-brand treatment and reporting obligations.
The MVNE and MVNO layer adds another route to volume. An MVNE can expose RBM through platform APIs to tenants, with platform fees and shared brand-verification operations. An MVNO with enterprise distribution may package RBM into sector propositions. Multi-IMSI stacks need brand policy alignment, because the sender relationship, serving network and subscriber identity may not map cleanly to one domestic compliance model.
Operations determine whether the commercial model holds. Brand review cannot run only during office hours if banks and travel providers send time-sensitive flows. Misuse response needs authority to suspend agents quickly. Quarterly business reviews should report conversion, read, fallback and complaint metrics in the same structure used for pricing tiers. If analytics are sold as premium, the operator must make them reliable enough for an enterprise procurement team to renew against them.
RCS Business Messaging becomes a margin business only when interconnect, rating and pricing are locked before launch. Set the MaaP, OCS and channel contracts first; the campaigns will follow the economics baked into the route table.