A wholesale rate card can make an MVNO look profitable in an afternoon. The operating model usually tells a different story. To understand how to validate MVNO unit economics, founders and operators need to test the full cost to acquire, connect, serve and retain a customer – then stress-test the assumptions that are most likely to move.
This is not a spreadsheet exercise to satisfy an investor deck. It is the commercial proof that your proposition can survive real customer behaviour, wholesale billing, support demand and working-capital pressure. A strong MVNO business can have a modest headline margin. A weak one can show an attractive gross-margin percentage while quietly losing money on every active subscriber.
Start with the customer, not the wholesale rate
Unit economics should be calculated at the level at which a customer actually buys. That may be a monthly SIM-only plan, an eSIM package, a connected-device tariff, a fleet subscription or a bundled offer with hardware and services. If your target customer has several products, model each cohort separately. Averaging them too early can hide the product that is damaging the business.
For a consumer MVNO, the initial unit is normally one active subscriber. For IoT, EV charging, security cameras or mobility deployments, it may be a connected asset, site or contracted account. The principle remains the same: define one economically meaningful unit, the revenue attached to it and the direct costs needed to keep it active.
Begin with net revenue, not the advertised price. Remove VAT and account for discounts, introductory offers, credit notes, failed collections and expected bad debt. If a £15 monthly plan is routinely sold for £12 in the first three months, £15 is not your starting revenue assumption.
Then model wholesale costs against expected usage rather than included allowance. A plan with unlimited data or calls is not a fixed-cost product simply because the retail price is fixed. Segment expected usage into low, typical and high cases, and establish what happens when customers exceed fair-use thresholds. The cost curve matters more than the marketing wording.
Build the contribution margin before fixed overhead
The first test is contribution margin: the cash generated by an active customer before central overheads and major launch costs.
Monthly contribution per active customer = net monthly revenue – direct recurring costs
Direct recurring costs typically include mobile network wholesale charges, data and voice usage, messaging, eSIM or SIM lifecycle costs, payment processing, customer care, billing and CRM charges, fraud losses, commissions paid over time, and product-specific platform fees. For an IoT MVNO, include connectivity management, device-management licences, provisioning activity and any data-routing or security services that vary by connection.
Do not classify a cost as overhead merely because it is inconvenient to allocate. If a customer requires a support interaction, a replacement SIM, a number port, a failed payment chase or a fraud review, that is part of serving that customer. The exact allocation can be debated; pretending the cost does not exist cannot.
The right level of detail depends on the proposition. A digital-first eSIM brand may have low fulfilment costs but meaningful payment, fraud and app-support costs. A retail-led MVNO may carry higher commissions, physical SIM distribution and call-centre demand. An enterprise connectivity offer may have excellent retention but expensive onboarding and account management. There is no universal ‘good’ contribution margin without understanding the operating model.
Treat support as a variable, not an aspiration
Business cases often assume customers will self-serve. They may, eventually. During launch, porting issues, activation failures, handsets, coverage queries and billing confusion create contacts at precisely the point when the team is least prepared to absorb them.
Model contacts per active customer per month, cost per contact and the change in both figures during launch. Include a higher case for service incidents and supplier defects. If a proposition only works when support demand is unrealistically low, it does not yet work.
Calculate acquisition cost with honest payback
Customer acquisition cost is more than paid media. It includes all expenditure required to bring a customer to an active, paying state: marketing, agency fees, retailer or affiliate commission, sales salaries, promotional subsidies, credit checking, SIM or eSIM activation, onboarding and any hardware contribution.
A useful calculation is:
Fully loaded CAC = total acquisition and activation spend / customers who become revenue-generating active connections
The denominator matters. Counting orders, downloads or shipped SIMs will flatter the result if a meaningful share never activates, fails credit checks or cancels within the cooling-off period. Track each step from lead to order, activation, first payment and 90-day retained customer.
Payback should be based on contribution margin, not revenue. A £60 acquisition cost on a £15 plan is not recovered in four months if the customer contributes only £5 per month after wholesale usage, support, payment fees and other direct costs. It takes twelve months, assuming the customer remains active and pays on time.
That assumption leads to the next issue: retention. Monthly churn must be modelled by customer cohort, channel and product. A low-cost acquisition channel with poor early retention can be more expensive than a higher-CAC channel producing long-lived customers. For business and IoT contracts, distinguish contractual term from actual economic retention. A customer may be under contract while consuming support resource, disputing invoices or reducing its connection base.
Validate MVNO unit economics through cohorts and scenarios
The practical answer to how to validate MVNO unit economics is to build a cohort model, then try to break it. Each monthly acquisition cohort should carry its own acquisition cost, activation rate, revenue, usage profile, churn, bad debt and service costs over its expected life.
This exposes timing that annual averages conceal. New cohorts often lose money at first because acquisition costs are paid upfront while revenue arrives monthly. If your wholesale supplier bills before customer cash is collected, or requires deposits and minimum commitments, the gap becomes a working-capital issue rather than a theoretical accounting adjustment.
Run at least three cases: base, downside and severe downside. The downside should not be a token 5 per cent reduction in revenue. Use operational events that actually occur in mobile businesses: higher data consumption, an increase in churn after an introductory period, delayed activation, higher than expected customer contacts, wholesale price changes, commission leakage, bad debt or a slower sales ramp.
The severe case answers a board-level question: how much cash, time and management capacity are needed if the launch does not perform to plan? If the business becomes unviable after a small movement in usage or retention, the proposition needs redesign before it needs more marketing.
Separate margin from cash
An MVNO can report a positive contribution margin and still run short of cash. Upfront costs can include supplier deposits, platform implementation, numbering and regulatory work, SIM inventory, integrations, launch marketing and working capital. A minimum wholesale commitment can make low subscriber volumes particularly expensive.
Map cash receipts and payments month by month for at least 24 months. Include payment terms, VAT timing, refunds and chargebacks. Then identify the maximum cumulative cash requirement, not only the eventual break-even month. That figure is often the true funding requirement.
Test supplier terms as hard as customer assumptions
The network agreement is part of the unit-economics model, not a legal document to consider after the business case has been approved. Minimum monthly commitments, data-rating rules, roaming treatment, connection charges, number-porting costs, price escalators, termination rights and settlement periods can materially change profitability.
Ask for representative invoices or detailed billing examples. A simple per-GB rate is rarely the complete story. Clarify whether charges are based on rounded usage sessions, pooled versus individual allowances, domestic versus roaming traffic, out-of-bundle events and inactive connection fees. For a multi-network or international proposition, test each route separately.
Commercial negotiations should also reflect the shape of the launch. A start-up may need lower minimums and a staged commitment more than it needs the lowest theoretical unit rate at scale. A larger enterprise may place greater value on service credits, operational support and predictable pricing. The best deal is the one that matches the risk profile of the actual plan.
Use early evidence to replace assumptions
Before committing to a full launch, seek evidence that reduces the most expensive unknowns. A controlled pilot can reveal real activation rates, usage distribution, support contacts, payment behaviour and retention signals. It will not prove lifetime value in a few weeks, but it can quickly disprove a fragile assumption.
Instrument the journey from acquisition source through to first top-up or bill payment, network activation, usage, contact reason and cancellation. Finance, product, operations and commercial teams should work from the same definitions. If marketing reports customers acquired, operations reports SIMs activated and finance reports paying accounts, without reconciliation, the unit-economics model will drift from reality.
At Christian Borrman, this is where senior operator experience adds value: translating a proposition into the supplier, technical and operational decisions that determine whether its economics hold outside a presentation.
The most credible MVNO plans are not those with the prettiest lifetime-value chart. They are the plans that show exactly which assumptions matter, what evidence supports them, where the downside sits and what the team will change when actual customer behaviour says the model was wrong.
