An MVNO in a box is a packaged service, not a piece of software: a multi-tenant SaaS BSS, a host network the vendor has already connected, SIM supply and the operational work around them. You launch faster and own less. Your own BSS takes longer to stand up but leaves the product, the subscriber data and the margin logic under your control. The question is which of those you need, not which is cheaper.
A retailer’s product lead sends the same brief to three suppliers and gets back three “turnkey” proposals. One quotes a launch in three weeks and a price per active subscriber. One quotes one to three months on an enabler’s network in two countries. The third quotes four to six months after the host agreement is signed, for software the retailer would own. All three are honest; they are selling different things under the same word. Juniper Research titled its January 2026 forecast “MVNO in a box platforms to drive MVNO market” and estimates 333 million MVNO subscribers globally in 2026, rising to 438 million by 2030, so the box is now the default entry point for many brands. Working out what is inside it, and what is not, is the decision that matters.
What “MVNO in a box” means
An MVNO in a box is a packaged service, not a piece of software. In the form the SaaS vendors sell it, the package has four parts. First, a multi-tenant BSS delivered as SaaS: product catalogue, charging, billing, CRM and a self-care front end, configured through the vendor’s portal and API. Second, a host network that is already connected, either through the vendor’s own wholesale agreement or, in some cases, with the vendor acting as carrier of record so the brand never signs with an operator at all. Third, SIM and eSIM logistics: profile ordering, activation and inventory. Fourth, care tooling, typically an agent console and the flows for porting, suspension and top-ups.
The commercial model follows the shape of the service. Telness Tech states a low set-up fee followed by pay-per-user or revenue share; Amdocs prices BRAND/ON as a set-up fee plus a per-subscriber charge; Alepo sells BSS Now on pay-as-you-grow terms. Time to launch is quoted in weeks. Telness Tech says “3 weeks”, Alepo “30 days”, Gigs “weeks”, Amdocs “under 3 months”, and Transatel gives “a few weeks” or “1–3 months” for its MVNO-in-a-Box depending on the page. Those are the vendors’ own claims, and they assume the host is already connected in your country.
The same term is also used by enablers with their own network elements, such as Plintron, Transatel and Effortel. There the brand is a tenant on the enabler’s platform and the enabler holds the host relationship and often its own core or HLR/HSS. The comparison of MVNE and MVNO BSS platforms describes each vendor in turn.
What is inside the box, and what usually is not
The table sets the SaaS box against the two alternatives a brand usually compares it with: a Light or Medium MVNO on an enabler’s network, and a BSS the brand owns and operates as a Medium or Full MVNO. Entries describe the usual arrangement. Individual contracts vary, and the useful question to any vendor is which column it is actually in for each row.
| Component | SaaS box | Enabler on own network | Own BSS (Medium or Full) |
|---|---|---|---|
| Catalogue and pricing rules | Inside: configured in the vendor’s portal within its plan templates | Inside: the enabler’s catalogue, often with a wider template set | Yours: any rule the charging engine can express |
| Charging and billing engine | Inside, shared and vendor-operated | Inside, shared and enabler-operated | Yours, dedicated; on-prem, private cloud or managed |
| Host interfaces | Inside: pre-connected hosts in the vendor’s countries | Inside: the enabler’s own core elements and host agreements | Yours to integrate; a host agreement is the prerequisite |
| Numbering and interconnect | Vendor’s or host’s ranges; vendor or host is carrier of record | Enabler’s ranges and interconnect | Host’s ranges (Medium) or your own (Full) |
| KYC and SIM registration (RICA and equivalents) | Inside for the vendor’s markets; the regulatory duty stays with the brand | Inside for the enabler’s markets | Your integration to the national provider; your records |
| Subscriber data ownership and export | Records in the vendor’s tenant; access by API and export | Records in the enabler’s platform; export on contract terms | Records in your perimeter and your data model |
| Custom billing logic (sponsored billing, balance-based rewards) | Usually not: limited to what templates support | Sometimes, as a chargeable change request | Yes: this is the reason to own the engine |
| Multi-tenancy (hosting other brands) | No: you are one tenant among many | No: the enabler is the multi-tenant party | Yes, if the platform is built for MVNE or MVNA use |
| On-prem or private-cloud option | Usually not: vendor cloud | Usually not: enabler infrastructure | Yes |
| Exit and migration | Export formats and exit terms in the contract; leaving is a migration project | The same, and the host relationship may not transfer | You hold the assets; a later Full stage extends the same platform |
Two rows carry most of the weight. Custom billing logic decides whether the box can express your product at all. Subscriber data ownership decides whether your risk engine, loyalty platform or data-residency obligation can be met by an API and a nightly export. If both are comfortable, the rest of the table is about price and convenience.
Where the box is the right answer
Plainly: for a consumer brand adding mobile as a feature, in its first twelve to twenty-four months, with a small team and standard plans, the box is usually the right call. The vendors’ reference lists show the pattern. Gigs names Revolut, Cash App, Klarna, Nubank, Santander, Starling and Motorola; Telness Tech names Lebara, Truecaller and New York Mobile. These are brands that wanted a SIM inside an app, a few bundles and a porting flow, and did not want a telecom department.
The arguments are concrete. Time to first customer is weeks rather than months, because the host is already connected. Fixed cost is low and variable cost tracks subscribers, which suits a product whose demand is unproven. Amdocs states that “10 people or fewer” can run a full MVNO or MVNE on its platform; whatever the exact number, the operating team is small. And the box is a legitimate way to test whether your customers want mobile from you before you commit to a platform, a host negotiation and a regulatory posture of your own.
A Light MVNO on an enabler’s network is the neighbouring answer. It is better when you want a Light or Medium position in a country where the enabler already operates, and you accept the enabler running the BSS and sitting in the commercial chain. Retailers and communities in Europe have launched this way for years, and an enabler that also acts as MVNA can offer wholesale terms a single brand would not get alone.
Where it stops
The box stops where the product stops being a plan. Six situations recur.
Pricing logic as the product. A bank that wants data allowances to move with account balance, a retailer that wants a loyalty tier to change the price of a bundle at the moment of purchase, or an employer that pays part of an employee’s usage (the sponsored-billing model Brilliantel runs in South Africa) needs a charging engine it can program, not a template it can choose from. Templates can be extended by the vendor, at the vendor’s pace and on the vendor’s roadmap.
Event-level data in your systems. An API returns what the vendor exposes; a nightly export returns yesterday. A risk model, a churn model or a real-time offer engine that needs the usage event as it happens needs the mediation and charging layer to be yours, or at least to feed your data platform directly.
B2B and B2G billing. Corporate accounts with hierarchies, cost centres, pooled allowances, split invoices, purchase orders and public-sector payment terms are a different billing problem from a consumer plan with a card on file. Most boxes are built for the latter.
Regulated in-country data. Where a regulator, a central bank or a procurement rule requires subscriber records to stay inside the country or inside your own perimeter, a vendor-cloud tenant in another jurisdiction is not an option, whatever its features.
Hosting other brands. If the plan is to become an MVNE or MVNA, or an MNO wholesale team wants to onboard partners, you need to be the multi-tenant party. In a box you are the tenant.
Host renegotiation. When the vendor holds the wholesale agreement or is carrier of record, the wholesale rate, the network and sometimes the number ranges are the vendor’s to renegotiate. A brand that reaches a few hundred thousand subscribers and wants to take its own volume to the host, or to add a second host, needs the host relationship in its own name. The operating-models page sets out who owns what at each stage.
Total cost over three years
Avante does not publish prices, so this comparison stays qualitative and the only figures are market estimates, labelled as such. MVNO Index puts a thin MVNO in the EU or UK at roughly EUR 105,000 to 475,000 in total set-up, and Spenza estimates USD 100,000 to 400,000 for a light MVNO against USD 2 million to 10 million or more for a full one. Those are single-source market estimates, not Avante numbers, and they describe launch cost rather than three-year cost. The cost-structure guide goes through the components one by one.
The shape of the two cost curves is what matters. A box starts low and grows with the subscriber count: a set-up fee, then a per-subscriber charge or a revenue share, plus wholesale airtime bought through the vendor’s agreement. There is little to depreciate and little to staff. Your own BSS starts higher: licence or subscription, integration to the host, a project team for four to six months, and either an operations team or a managed-service fee. It then grows slowly, because software cost does not scale one-for-one with subscribers, and because wholesale is bought at your own volume.
Where the two curves cross depends on four variables: subscriber count, average revenue per user (a revenue share bites harder on a high-ARPU base), the number of change requests your product needs, and your wholesale rate. Below a certain base, and with a plan that fits the templates, the box is cheaper for the whole three years. Above it, and with a product that needs custom logic, the per-subscriber line plus change requests plus the margin taken by the intermediary tends to pass the fixed cost of owning the stack. Ask any box vendor to model year three at your target base and ARPU, with the change requests you already know you will need. Ask any BSS vendor for the managed-service and integration cost over the same period. Compare those two numbers, not the launch prices.
Moving from the box to your own BSS
Brands do outgrow the box, and the move is a migration, not an upgrade. The subscriber base, billing history, SIM and eSIM inventory, number ranges, active plans and balances all move from the vendor’s tenant into your platform, while porting windows and service continuity are managed with the host. The work is on the scale of the original launch, which is why the exit terms deserve reading before signature: export formats, what happens to number ranges if the vendor was carrier of record, and the notice period.
Done well, it is one migration window. Subscribers keep their numbers, active plans and history, and the brand switches charging and care to the new platform in a defined cut-over rather than re-enrolling customers. This is the path Avante describes for Light MVNOs moving to Medium, and the launch page sets out the stages, including the host and legal work that sits on the critical path.
Two things shorten the project. Negotiating your own host agreement early, in parallel with the box period, so it is ready when the base justifies the move. And choosing a target platform on which Medium and Full share one data model, so the migration happens once and a later Full stage adds network elements without a second move.
Where Avante fits
Avante supplies the BSS an MVNO, MVNE or MVNA owns and runs; it does not run a network or resell airtime, so a host or enabler is always a partner. The MVNx Suite covers product catalogue, online charging, convergent billing, CRM, provisioning, mediation, interconnect and partner management, delivered on-premises, in private cloud or as a managed service, with an OPEX model available. A Medium MVNO launch is planned at four to six months after host agreement and legal readiness; Full is eight to ten months and extends the same platform, with no second subscriber migration. The platform is multi-tenant for MVNE and MVNA use. Brilliantel (sponsored billing, B2B and B2G, OPEX model) and VENTAmobile (multi-IMSI, eSIM, M2M) are published cases. The platform page lists modules and host interfaces.
Frequently asked questions
What is an MVNO in a box?
A packaged mobile service for brands: a multi-tenant SaaS BSS (catalogue, charging, billing, CRM, self-care), a host network the vendor has already connected or for which it acts as carrier of record, SIM and eSIM logistics, and care tooling. It is priced as a set-up fee plus a per-subscriber charge or revenue share, and vendors state launch times from three weeks to under three months. The brand configures within the vendor’s tools rather than owning the platform.
Is MVNO as a service the same as using an MVNE?
They overlap. MVNO as a service, or the SaaS box, is a managed platform plus a wholesale relationship, sometimes with the vendor as carrier of record. An MVNE provides network elements and a multi-tenant platform in specific countries and takes brands on as tenants. In both cases the brand configures within someone else’s stack; the difference is mainly who holds the host agreement and the core elements.
How much does an MVNO SaaS platform cost?
Vendors do not publish figures. The stated structure is a set-up fee followed by a per-subscriber charge or revenue share, plus wholesale airtime. Market estimates from MVNO Index and Spenza put a thin or light MVNO launch in the low to mid hundreds of thousands of euros or dollars; those are single-source estimates for launch, not for three years of operation. Model year three at your target base and ARPU before comparing with owning a BSS.
Can we launch on a turnkey MVNO platform and move to our own BSS later?
Yes. Treat it as a migration project: subscriber base, billing history, SIM and eSIM inventory, number ranges, active plans and balances all move, with porting and continuity managed with the host. Read export formats and exit terms before signing, start your own host negotiation early, and choose a target platform on which Medium and Full share one data model, so the migration happens once.
Who owns the subscriber data on an MVNO in a box?
The records sit in the vendor’s tenant, in the vendor’s cloud, and you reach them through an API and scheduled exports. Contractually the data is usually yours; operationally it is in someone else’s system and often another jurisdiction. Where a regulator or your risk team needs event-level data inside your perimeter, or in-country, that arrangement does not meet the requirement, whatever the contract says.
When does owning the BSS pay off?
When pricing logic is the product (sponsored billing, balance-linked rewards, loyalty rules on purchase events), when event-level data has to be in your systems, when you bill corporates or public bodies, when data must stay in-country, when you intend to host other brands, or when you want the host relationship in your own name. Below those thresholds, a box or an enabler is usually cheaper and faster for the first year or two.