MVNO revenue share agreements often settle on a percentage before either party has agreed what constitutes economically useful growth. That ordering creates predictable disputes: the MNO sees low-value traffic and support load, while the MVNO sees a host charging for activations that fail to convert. The contract should instead place IMSI activation quality, traffic mix, and remedy mechanics ahead of the headline share, so commercial returns follow verified operating performance.
The headline revenue share is a poor proxy for value
A single revenue-share percentage compresses several different economic outcomes into one number. It treats a durable, regularly recharging subscriber as equivalent to an activation that disappears after a promotional allowance. It also obscures the difference between revenue that carries a healthy wholesale margin and revenue whose termination, roaming, support, or fraud costs absorb most of the host return. The percentage may be easy to negotiate, but it is a weak measure of value.
Gross activation counts are particularly unreliable. Promotional SIM distribution can produce a large provisioned base before demand is established. Incomplete KYC journeys, porting failures, device incompatibility, and short-lived multi-SIM use further separate issued subscriptions from paying customers. If settlement rewards each nominal activation, the MVNO can meet an acquisition objective while the host receives little recurring traffic and inherits additional operational load.
The MNO cost base starts before recurring revenue is proven. IMSI allocation, provisioning workflows, customer-care interfaces, BSS/OSS configuration, number-range administration, roaming exposure, and fraud controls all have operating consequences. Some costs are marginal. Others arise from reserved capacity, specialist operations, reconciliation, or changes that must be maintained for the life of the tenant relationship. None is captured adequately by counting provisioned records.
An MVNO may rationally tolerate weak early retention when low acquisition costs, investor-funded promotion, or a broader platform objective justify it. The host has no corresponding reason to subsidise that strategy through an undifferentiated share. Its commercial test is contribution margin after wholesale network, interconnect, roaming, support, bad-debt, and settlement costs, not merely the service revenue booked by the tenant.
The issue was material in the commercial review of a Tier-2 MNO, Southeast Asia, ~18M subscribers. The relevant distinction was not between a nominal 20% and 40% host share. It was between activations that developed a repeat usage pattern and those that generated provisioning and support activity without sustainable wholesale income. Once that distinction is accepted, the negotiation moves from a headline percentage to the activity that earns it and the point at which usage becomes economically credible.
Define performance through observable lifecycle events
Activation quality should be translated into a limited set of auditable lifecycle events. A valid activation requires more than an HLR or AuC provisioning event. A stronger definition can combine successful provisioning, first authenticated network attach, a qualifying chargeable event, and continued service status after an agreed period. Each event should have a timestamp, a system owner, and a rule for handling reversals.
The qualifying period must follow the proposition rather than a standard template. A prepaid travel MVNO may have concentrated usage over several days and long intervals between journeys. A domestic value brand may need a recharge or rated-usage test across a full billing cycle. An IoT/eSIM offer may remain provisioned for months before transmitting small but contractually valid data volumes. Applying the same 30-day retention threshold to all three would create false failures as well as false successes.
Traffic mix also matters where unit economics differ materially. On-net and off-net voice do not produce the same interconnect position. Domestic data, roaming data, international termination, A2P exposure, and DCB activity carry different cost, risk, and settlement profiles. A revenue threshold reached through low-margin or anomalous traffic should not automatically earn the same commercial treatment as a balanced, recurring subscriber base.
The agreement should set a source-of-truth hierarchy before invoices are disputed. OCS and mediation records can govern rated usage. HLR or HSS records can establish subscriber state. MNP data can determine whether a port completed or failed. Finance-controlled settlement files should govern invoiced revenue. Measurement windows must then specify time zone, rating finality, treatment of late CDRs, reversal handling, and the point at which fraud, non-payment, or regulatory disconnection excludes an activation.
A monthly operating pack should expose the same denominator to both parties: provisioned IMSIs, attached IMSIs, chargeable IMSIs, retained IMSIs, and exclusions, with movements reconciled between periods. This is less about adding reporting volume than preventing one party from presenting activations while the other presents active revenue-generating subscriptions.
An MVNE servicing 12+ tenants in EMEA faces this issue at portfolio scale. Tenant propositions, rating cycles, and retention patterns vary, yet host settlement still requires a controlled data model. The commercially useful approach is not to force every tenant into one behavioural profile. It is to standardise evidence, ownership, finality, and exception handling while allowing proposition-specific thresholds.
Commitments should protect capacity without becoming punitive floors
Minimum commitments are defensible when they correspond to identifiable host exposure. Reserved number ranges, IMSI blocks, dedicated service operations, bespoke BSS/OSS integration, or non-standard roaming and interconnect configurations create costs that cannot always be recovered from uncertain usage. A guarantee can compensate for that commitment. It becomes harder to justify when it merely transfers all launch-demand risk to an early-stage tenant.
For a greenfield launch, a stepped structure is usually more credible than an immediate volume floor. The first stage can recover agreed implementation work. A launch-period minimum can then cover operational readiness and reserved resources. Higher thresholds should apply only after jointly verified activation, retention, or revenue milestones show that the proposition has entered normal operation. This preserves commercial discipline without assuming that a forecast is equivalent to realised demand.
The form of the commitment matters. A subscriber-volume minimum can encourage dormant IMSIs if the active-base definition is weak. A minimum wholesale-spend obligation protects host revenue but may reward an inefficient traffic mix. A minimum contribution-margin obligation aligns more closely with economics, although it requires greater cost transparency and tighter reconciliation. The parties should choose deliberately because each form assigns demand, pricing, and traffic risk differently.
Credits against future usage can soften an initial minimum without eliminating the obligation. Their expiry, eligible service categories, and application order must be explicit. Otherwise, accumulated credits become a second settlement dispute. Exclusions should cover host-network outages, material provisioning failures, regulatory restrictions, and agreed force-majeure events. They should not convert ordinary tenant underperformance, weak acquisition, or poor retention into an automatic waiver.
A Greenfield MVNO, post-2023, multi-IMSI stack illustrates why an active-base test requires more than subscriber status. A multi-IMSI architecture can support legitimate coverage, roaming, or service-continuity objectives, but it can also multiply records that do not represent distinct revenue-generating customers. The commitment should therefore follow meaningful usage, recharge, or margin conditions rather than a raw count of allocated identities.
Remedies determine whether a guarantee has commercial force
A performance guarantee has little value unless a missed metric triggers a proportionate, time-bound response. The first distinction is control. Provisioning latency, MNP completion, API availability, CDR delivery, and network attachment failures can support host-side remedies because the MNO or its delivery chain controls the relevant systems. Retail conversion, campaign performance, and most retention outcomes remain primarily within the MVNO’s commercial control.
Where activation-quality thresholds are missed because of host-side defects, appropriate remedies may include service credits, deferred minimums, remediation support, or termination rights after repeated material failure. The remedy should follow the economic effect. A short reporting delay does not warrant the same response as a persistent provisioning defect that prevents subscribers from attaching or completing a port.
Where the MVNO misses retention, revenue, or traffic-quality thresholds, the response can include a revised forecast, additional security, reduced commercial concessions, a funded cure plan, or termination after a defined period. Automatic penalties are not always the strongest mechanism. A remedy that restores forecast credibility or limits further host exposure may have more commercial value than a charge the tenant cannot sustain.
The cure process must state the reporting date, disputed-data procedure, corrective-action owner, evidence required, and escalation path. A clause saying that the parties will work together does not resolve a billing dispute. Nor does it establish whether the metric failed because of source-data timing, rating logic, a host defect, fraud, or market performance. The agreement should permit provisional settlement where necessary, followed by a controlled true-up once the defined evidence reaches finality.
Remedy caps require similar balance. A low service-credit cap can negate the only practical response to persistent operational failure, particularly when the defect also causes the MVNO to miss a minimum commitment. Uncapped credits for minor or transient misses can make the host model uneconomic. Materiality thresholds, recurrence tests, aggregate caps, and termination triggers should therefore work as a system rather than as isolated clauses.
Quarterly governance should examine more than recorded exceptions. Acquisition-channel mix, recharge behaviour, roaming profile, fraud loss, MNP fallout, and material tariff changes can invalidate assumptions that were reasonable at signature. Governance has commercial force when it can reset forecasts, launch a cure plan, amend thresholds, or escalate a structural issue before the next settlement cycle.
The more variable the MVNO proposition, the less useful a flat revenue-share percentage becomes. Agreements that connect settlement to verified activation quality, margin-relevant traffic, and defined cure paths give both parties a basis to invest, forecast, and intervene before a wholesale relationship becomes a dispute. The headline share still matters, but only after the contract establishes which subscribers qualify, which traffic creates value, and what happens when operating evidence falls short.