A connectivity launch rarely fails because the SIMs do not work. It fails because nobody has clear ownership of the customer journey, the supplier relationship, the device estate, the data bill or the exception process. A managed connectivity operating model turns those moving parts into an accountable business function rather than a collection of contracts, portals and urgent support calls.
For an MVNO, a private 5G network, an IoT estate or a connected vehicle programme, the operating model is the point where strategy meets the realities of daily delivery. It determines who can activate a device, who approves a tariff change, what happens when a camera goes offline, how usage anomalies are detected, and who owns the commercial consequences when a supplier misses an SLA.
The technology matters. But technology without an operating model merely shifts operational risk into a different platform.
What a managed connectivity operating model must do
At its simplest, an operating model defines how people, processes, technology, suppliers and commercial controls work together to deliver a connectivity service. The word managed is significant. It means the organisation has made deliberate choices about governance, monitoring, incident response, lifecycle management and decision rights.
A good model gives leadership a clear line of sight from a business outcome to the operational activity that protects it. For example, a logistics company may care about proof of delivery, vehicle location and driver safety. Its connectivity team needs to translate that requirement into coverage rules, device provisioning, roaming policy, data thresholds, replacement stock, support escalation and reporting. None of those decisions can sit in isolation.
This is also where many connectivity programmes become unnecessarily expensive. Finance sees a growing mobile bill. Operations sees devices in the field. IT sees a security concern. Procurement sees a supplier contract. If each team acts separately, the organisation optimises local problems while losing control of the service as a whole.
The operating model provides the shared mechanism for making trade-offs. It should make clear who decides, who delivers, who is consulted and what evidence is needed before a material change is made.
It is not a network diagram
A network design may show radios, gateways, APNs, cloud platforms and device connections. It does not explain how a failed SIM is replaced at 7am, how a customer is informed of an outage, or how a disputed charge is resolved.
Likewise, an MNO or managed service provider may offer a service-management portal. That portal is a tool, not an operating model. It cannot resolve ambiguity between your product team, customer operations, field engineers and commercial lead.
The distinction matters most when a programme starts to scale. A hundred connected assets can often be managed by knowledgeable individuals using spreadsheets and goodwill. Ten thousand assets across multiple countries, suppliers and device types require repeatable controls.
Start with the service, not the supplier
The first practical question is not which carrier, eSIM platform or MDM tool to select. It is what service the organisation is actually committing to provide.
Define the service in terms that matter to the end user and the business. For a private network at a port, that may mean reliable coverage in defined operational areas, prioritised traffic for critical equipment and a stated restoration process. For an EV charging operator, it may mean monitored connectivity, remote recovery capability and no avoidable loss of transaction data. For an MVNO, it includes onboarding, number porting, billing accuracy, customer care and fault ownership.
Once this is clear, map the service lifecycle. There are usually four connected stages: design and onboarding, live operations, change and improvement, then retirement or replacement. Each needs named ownership and measurable outcomes.
The onboarding stage should establish whether a device, SIM profile or site is suitable before it enters service. Live operations need monitoring and triage rules. Change management must protect customers while allowing the offer to evolve. Retirement needs secure deactivation, asset recovery where relevant, and a reliable record of what has been removed.
Skipping the final stage is a common source of waste. Dormant SIMs, forgotten devices and legacy tariffs can remain on the bill for years because nobody owns decommissioning.
Put decision rights where the operational knowledge sits
Senior leaders do not need to approve every activation or troubleshoot every coverage incident. They do need confidence that decisions are being made at the right level and that risks are escalated early.
In practice, a managed connectivity operating model should establish clear authority across commercial, technical and operational decisions. Product ownership should define the proposition and customer promise. Connectivity operations should manage provisioning, monitoring, exceptions and supplier escalation. Technical architecture should set security, integration and device standards. Finance and commercial teams should govern commitments, billing and margin exposure.
Those boundaries must be explicit, particularly where a third party operates part of the service. An outsourced NOC can monitor alerts, but it cannot decide whether a major incident warrants customer communications unless that authority has been agreed. An MNO can provide network information, but it cannot own your end-customer experience.
A useful test is to take a realistic failure scenario and walk it through. A fleet tracker has stopped reporting across a region. Who detects it? Who confirms whether the issue is device, coverage, platform or configuration? Who engages the network operator? Who communicates with customers? Who authorises a workaround that may create additional cost? If the answers are vague, the model is not ready.
Design controls around the failure modes that matter
Not every connectivity service needs a 24/7 operations centre. A pilot for agricultural sensors may tolerate delayed intervention. A security camera estate or autonomous vehicle programme may not. The right model depends on operational criticality, revenue exposure, regulatory requirements, geography and the realistic capability of the organisation.
However, most scaled programmes need controls in five areas:
- Provisioning and lifecycle management, including activation, suspension, replacement and termination.
- Usage and cost control, with thresholds, anomaly detection and a clear process for resolving billing disputes.
- Incident management, covering severity definitions, escalation paths, customer communications and root-cause review.
- Supplier governance, including SLA reporting, commercial performance, roadmap alignment and remedies when commitments are missed.
- Data, security and compliance, particularly where devices collect video, location or other sensitive operational information.
The objective is not bureaucracy. It is to prevent routine issues becoming executive escalations. A short, well-used exception process is worth more than a detailed governance pack that nobody reads.
Build the data foundation early
Connectivity data is often fragmented across MNO portals, eSIM platforms, device management tools, billing systems, CRM records and field-service applications. If those records do not reconcile, the organisation cannot confidently answer basic questions: which customer owns this device, which tariff is applied, what service is live, or whether the asset is still in use.
Create a minimum common data model before scale makes this difficult. Every active connection should have a unique identifier and a trusted relationship to the customer, device, product, tariff, location where relevant, status and cost centre. The system of record must be clear. So must the process for correcting poor data.
This is especially important for eSIM and multi-network strategies. The ability to switch profile or network can improve coverage and resilience, but it adds operational complexity. Profile inventory, entitlement, device compatibility, activation status and commercial liability all need to be visible. Flexibility without control creates a different class of problem.
Measure service performance and business performance together
Network availability alone is an incomplete measure. A service can meet a carrier availability target while still failing the customer because activation takes too long, an integration is broken or support cannot identify the affected assets.
Use a balanced set of measures. Operational indicators might include activation success, time to restore, incident recurrence, device reporting rate and dormant connection volumes. Commercial indicators should include cost per active asset, out-of-bundle exposure, revenue leakage, churn or service credits. Customer indicators may cover first-contact resolution, complaint themes and the time required to deliver a requested change.
The measures should drive action, not merely produce monthly reporting. If a specific device type causes most failed activations, the response may be a certification change. If roaming costs are creating margin loss, the answer might be tariff redesign, a usage policy or a different commercial agreement. Data becomes valuable when it changes a decision.
Treat suppliers as part of the operating system
MNOs, platform providers, hardware vendors and service partners each have their own priorities, constraints and support structures. Their standard operating procedures may be reasonable, but they are not automatically aligned to your product promise.
A strong supplier model begins before contract signature. Due diligence should test not only coverage, pricing and technical capability, but also escalation access, change notice periods, portal limitations, support hours, data availability and the real process for resolving a serious service issue.
Once live, governance needs the right cadence. Weekly operational reviews may be necessary during launch or recovery. A stable programme may need monthly performance reviews and quarterly strategic sessions. The important point is that meetings result in owned actions, dates and commercial accountability.
Do not mistake a supplier account manager for an operational model. The relationship has to survive staff changes, competing priorities and a difficult incident.
Scale the model before growth forces it
The best time to define roles, data ownership and exceptions is before customer volume makes every change risky. That does not mean building a large operations team on day one. It means creating a model that can mature in stages.
At launch, a small team may combine product, technical and operational responsibilities. As the estate grows, specialist ownership becomes necessary. Automation should be introduced where it removes repeatable work and improves control, such as provisioning workflows, usage alerts, ticket routing and reconciliation. It should not be used to hide an unclear process.
A managed connectivity operating model is ultimately a commercial protection mechanism. It protects customer experience, margin, delivery capacity and management attention. Get it right, and connectivity stops being a recurring source of surprises and becomes a service the business can confidently build upon.
