A wholesale rate card can appear commercially viable while leaving the host MNO, MVNE and MVNO exposed to different forms of cost volatility. The central discipline is a transparent OCS cost allocation: a pricing model that separates network consumption, control-plane expenditure, traffic imbalance and contractual risk before margin is divided. Without that discipline, minimum commitments become substitutes for underwriting and settlement disputes arrive after traffic has already scaled.
Headline wholesale rates conceal the real cost base
Access pricing is only one layer of the wholesale cost base. A price expressed per SIM or as a bundled allowance must also absorb subscriber activation, suspension and termination activity, rating, customer-care interfaces, fraud controls, settlement work and regulatory administration. Those demands vary by tenant design. Two tenants with similar active subscriber counts can create materially different workloads if one maintains a stable tariff set while the other runs frequent promotions, recurring allowance changes and complex recharge rules.
The economics remain OCS-driven even where the host retains the core BSS/OSS estate. Real-time balance control creates processing demand across charging events, counters and policy interactions. Product teams must configure allowances and precedence rules. Operations teams must investigate rejected events, stale balances, duplicate records and disputed depletion. These costs do not move neatly with active subscriber numbers. A low-usage subscriber may create little network consumption but still generate recurring balance queries, care contacts and adjustment activity.
The distinction between fixed and variable cost is therefore central. Tenant onboarding, integration testing, product configuration, reporting design and settlement controls are relatively fixed. Domestic data, voice termination, SMS delivery and roaming consumption vary with traffic. Exception handling sits between the two: part standing capability, part event-driven workload. Pricing everything through a single access fee forces the host or MVNE to estimate a blended cost before the subscriber mix and operating pattern are known.
The case for simplicity is not without merit. An all-in price can shorten negotiations, support faster retail modelling and reduce invoice lines during an early launch phase. At low volume, estimation error may remain tolerable. The model weakens once a tenant introduces several bundles, promotional mechanics or a high-usage cohort. What looked like commercial simplicity becomes cross-subsidy between products, subscribers or tenants, with no agreed method for identifying who created the additional cost.
A portfolio review by an MVNE servicing 12+ tenants in EMEA illustrates the governance problem. The relevant comparison was not merely revenue per active subscriber. It was the relationship between tenant-specific product complexity, rated event volumes, care adjustments and network usage. Where those drivers were not separated, margin pressure appeared to be an inadequate headline access fee. In practice, the deficit could originate in configuration intensity, traffic mix or repeated exceptions. Raising the base rate would recover some value, but it would not correct the allocation error.
Traffic asymmetry must be priced as a portfolio exposure
Average usage is a weak basis for underwriting an MVNO portfolio because direction, concentration and volatility determine the economic exposure. A balanced consumer-oriented base may contain enough low- and medium-usage subscribers to offset a smaller heavy-usage segment. A specialist proposition can invert that distribution. It may attract persistent high-data users, inbound-heavy voice traffic, A2P exposure or roaming-intensive corporate cohorts from launch.
Nominally profitable average ARPU can therefore conceal loss-making segments. A small share of subscribers may consume a disproportionate amount of radio capacity, backhaul, IP transit, roaming capacity or interconnect resources. The retail bundle can remain attractive while the marginal wholesale contribution turns negative. Portfolio averages delay recognition because the cost concentration is netted against subscribers whose allowances remain largely unused.
The contract should instead define traffic corridors and usage bands for domestic data, roaming data, outbound voice, inbound voice, SMS and MNP events. These categories carry different cost bases and risk profiles. Direction matters for voice. Destination and sponsor-network arrangements matter for roaming. Porting events create operational and regulatory work that subscriber averages do not capture. Bands need not reproduce every underlying network cost, but they should identify where the commercial assumptions stop applying.
Bundled voice and OTT bypass can further weaken the expected contribution. An app-led proposition may suppress chargeable voice while the host remains responsible for access, QoS exposure and support escalations. Conversely, an inbound-heavy proposition may appear inexpensive to the subscriber but create termination, numbering or fraud-control obligations for the operator. A generic bundle rarely distinguishes these effects until volumes are commercially significant.
The appropriate response is a scheduled review mechanism based on cohort-level consumption and defined reopener thresholds. Parties can agree bands for concentration, roaming share, voice direction or sustained data consumption, then specify the observation period and pricing response. This treats asymmetry as a portfolio condition to manage, not as misconduct to allege after the economics have deteriorated. It also gives the MVNO an opportunity to revise tariffs or product controls before a wholesale adjustment becomes unavoidable.
Minimum commitments should underwrite fixed capacity, not mask uncertainty
Minimum revenue commitments are defensible when they correspond to identifiable host exposure. They can recover implementation and integration costs, assure a base contribution to shared operating costs and protect the wholesale provider from speculative distribution plans. Each purpose has a different time profile. Integration cost is concentrated around launch. Shared operating cost continues throughout the term. Distribution risk declines only when acquisition performance and subscriber quality become observable.
A minimum tied solely to projected subscriber numbers can misstate that exposure. An estate may contain large numbers of inactive SIMs, delayed MNP migrations, promotional free periods or low-recharge cohorts. The volume forecast can be met while cash contribution remains below plan. The opposite can also occur: a smaller base of higher-value subscribers may produce adequate contribution despite missing the contracted volume. Subscriber count is useful for capacity planning, but it is not a reliable substitute for revenue quality or cost-to-serve.
A staged structure is more closely aligned with the underlying risk. Implementation fees can be paid at launch or amortised over an agreed initial term. A ramping minimum can then cover migration and early commercial operation, when forecasts remain least certain. At steady state, the commitment can reference active revenue-generating subscribers, defined usage assumptions and an agreed contribution floor. The calculation should exclude statuses that create no retail receipt unless the host continues to incur a specified cost for maintaining them.
Hard minimums do impose discipline on a greenfield business. They test whether distribution forecasts have financial backing and discourage indefinite reservation of host capacity. Yet excessive front-loading can force the MVNO to protect short-term cash through tariff increases, reduced allowances or aggressive acquisition. Those responses may worsen churn, attract low-quality traffic or weaken the proposition before the base matures. The host then secures an early contractual claim at the expense of a smaller long-term wholesale contribution.
Remedies should therefore precede penalties. Cure periods, forecast refreshes, traffic re-pricing and controlled product changes can preserve more value than immediate termination rights. A missed minimum may reflect delayed migration, a distribution problem or a traffic mix that invalidates the original assumptions. Each requires a different response. Termination remains necessary for persistent underperformance or credit deterioration, but it should follow an operating review that identifies whether the gap is temporary, structural or attributable to the pricing model itself.
A margin waterfall turns negotiations into operating governance
A durable wholesale agreement makes each party’s margin claim visible. The waterfall begins with retail receipts, less indirect tax, payment costs and bad debt. Host network and OCS charges follow, then MVNE platform and operational charges, regulated or third-party pass-throughs, and finally the MVNO’s distribution and service margin. The sequence matters because it shows which costs are deducted before commercial value is divided and which party carries volatility when actual performance diverges from forecast.
Transparency does not require the MNO to disclose every underlying network cost. Operators can protect commercially sensitive inputs while agreeing charging logic, auditable source data and treatment rules. The essential distinction is between pass-through items and negotiated margin. Roaming charges, regulated fees or third-party termination may move according to external schedules. Platform fees and access margins are commercial terms. Combining them into one number makes subsequent changes difficult to test and invites disagreement over whether an increase reflects cost recovery or margin expansion.
The model depends on governance data that both parties can reconcile. Subscriber status must show activation, suspension, dormancy and termination. Rated usage must connect to allowances and product rules. Roaming records, MNP events, payment failures, credits, disputes and invoice adjustments require consistent identifiers and accounting periods. Without those controls, a mathematically sound waterfall still fails because the parties are applying it to different populations, dates or event states.
Monthly settlement should connect commercial reporting directly to operational systems. OCS rating records need to reconcile with allowance consumption and adjustments. Roaming files should flow into the agreed true-up process. Interconnect invoices must be mapped to the traffic categories used in the rate model. Management reports remain useful for P&L oversight, but they should not become an alternative ledger assembled manually after the operational records have diverged.
The waterfall also creates a disciplined basis for change. A new retail bundle can be tested against expected network consumption, charging activity, operational workload and third-party exposure before launch. A pricing reopener can target the affected layer rather than reopening the entire agreement. This reduces the tendency to recover every adverse movement through a higher access fee and gives each party a clearer view of the action available to it.
As digital MVNO portfolios become more segmented, wholesale pricing will need to move from flat access discounts toward governed allocation of cost and risk. The commercially durable agreements will be those that make OCS exposure, traffic asymmetry and minimum commitments measurable before they become settlement disputes. A visible waterfall gives the MNO confidence that wholesale value is protected, gives the MVNE a defined operational margin, and gives the MVNO a basis for responsible tariff design rather than artificial retail pricing.