Summary

  • Ofcom’s current innovation licences support research and public demonstrations, but they do not authorise full commercial operation or promise long-term access. Moving from a promising test to a live service can therefore require a new licence and another round of technical and policy review.
  • The regulator’s consultation is closed, but no decision has been published. Responses differ over paid trials, term length and incumbent participation; the useful reform test is whether applicants can see the route, timetable and exit conditions before they invest.

A successful test still faces a second gate

A wireless trial can prove that equipment works in a laboratory or at one site. It cannot by itself prove that a service can operate at scale, coexist with other users, attract customers or justify the cost of a durable spectrum right. Ofcom’s innovation-licensing review is about the gap between those stages.

Ofcom says it issues more than 750 Innovation and Trial licences each year. Its two products divide early work by purpose. An Innovation & Research licence supports equipment development or research at a specific site. A Demonstration & Trial licence lets an applicant test a radio concept, application or service at a site or across an area, and can involve third parties or the public. Both are temporary, non-operational and on a non-interference, no-protection basis. The consultation lists a £50 annual fee for a single-location research licence and £50 per location per month for a demonstration licence.

Those terms make sense for experiments. They also draw a sharp boundary at the moment a developer needs to test with real users, accept payment, extend a deployment or rely on access long enough to recover investment. Ofcom says commercial use and long-term access are outside the current test regime; the operator then needs a different authorisation. A laboratory success may consequently precede a pause in service while the next application is assessed.

The queue itself is measurable. Ofcom reports that roughly 70% of current innovation and trial applications are processed within 42 days and about 96% within 60 days; complex applications can take longer. Those figures do not say how long it takes to reach commercial operation, how many projects stop between licences or how much uncertainty affects financing. They do show why a reform should preserve a published processing benchmark and add visibility to the hand-off after a test. Faster first permits alone would not close the transition gap.

The consultation, published in July and closed on 30 September 2026, did not propose rules. It asked for evidence about commercial trials, renewable or longer licences, low-risk applications, bespoke requests and security of tenure. The public page continued to label the matter “Pending Statement (Closed)” on 17 October and said Ofcom plans to consult on specific proposals in early 2027. The 17 response files posted on 16 October are arguments from individual respondents, not a vote or a final policy.

Those arguments expose the design tension. Amazon Leo says predictable timing matters at least as much as raw speed and asks Ofcom to keep reporting against its processing benchmarks. Apple favours a visible route from trial to limited early deployment, with published criteria, evidence requirements and indicative timelines. The Dynamic Spectrum Alliance similarly supports temporary transitional permissions with defined technical conditions. Their common point is that a developer needs to know what evidence and decision come next before committing to equipment, staff and customers.

BT puts more weight on the boundary between testing and a commercial spectrum right. It says paid services can raise competition and spectrum-pricing questions that should be handled through a separate licence, potentially awarded competitively. It accepts incidental use by existing paying customers where the new radio capability is not separately charged, and suggests a strict time limit for any commercial activity under a test permit. That distinction asks regulators to define what is being tested: a device and network feature, or a service already being sold to customers.

VodafoneThree brings a different vantage point because it is both an applicant for test licences and an incumbent holding commercial spectrum that other applicants may seek to use. It says the current process works broadly well, but sees avoidable paperwork in the split between research and demonstration permits and proposes one flexible trial licence. For paid beta tests, it proposes limiting the number of users and making clear that service is experimental and can be withdrawn. It also calls for a 12-month initial term with simple extensions where the technical parameters and prior coordination remain unchanged.

A completed test, it argues, must not create a presumption that Ofcom will grant a commercial licence.

That last safeguard is not a technical footnote. A temporary authorisation can generate customer expectations, supplier commitments and investment that become politically difficult to unwind. If a successful demonstration quietly becomes evidence of entitlement, a trial process can pre-empt the later allocation decision without the open competition, interference assessment or pricing that a permanent right may require. Conversely, if the authority offers no defined next step, investors may not fund the very pilots that create evidence about commercial value and coexistence.

Ofcom already points toward a risk-based route. It asks whether low-power or remote-site applications could receive lighter assessment, whether common interference thresholds could reduce case-by-case coordination, and whether low-risk zones or sandbox arrangements could support longer or more operational trials. For bespoke requests, it asks whether applicants should produce more of the coexistence analysis themselves.

Apple and the DSA support objective criteria, digital processing and automated coordination; BT suggests pre-cleared locations; VodafoneThree asks that coordination reach only spectrum managers likely to be affected and that relevant incumbents remain informed.

The burden can move when a process speeds up. Asking applicants to model coexistence may reduce Ofcom engineering work, but smaller firms and research groups may lack access to propagation data, incumbent parameters or specialist staff. A published method, reusable templates and access to the data needed to apply them are therefore part of a fair fast lane. Ofcom should still validate the assumptions and decide whether the evidence protects other users. An applicant’s analysis can inform the decision; it cannot transfer the regulator’s statutory responsibility.

Duration presents the same balancing act. Longer permits can prevent avoidable gaps in multi-year research and make investment easier to plan. But a long, renewable trial permission can tie up a frequency or create a de facto commercial right before Ofcom has made the broader allocation choice. Respondents offer different numbers: VodafoneThree favours a 12-month starting point for trials, while BT says about five years may suit some commercial uses. Apple rejects a universal minimum and ties term to investment, technology maturity and deployment.

The proposals concern different kinds of permission, so their figures are not directly comparable.

A workable framework would publish a ladder rather than promise a result: research, controlled demonstration, limited transitional operation and the separate route to durable commercial authorisation. Each step could specify who qualifies, evidence and coordination duties, a target decision clock, fees, the authorised user group and the grounds for extension, modification or withdrawal. A trial result would remain relevant evidence, not priority over other applicants. Operators and investors could then price the risk of a test without treating it as an award.

Ofcom’s early-2027 proposals will show whether it treats the problem as a licensing queue or as a sequence of decisions. The key measures will be time from complete application to a usable authorisation; publication of reasons when targets are missed; clarity on how paid beta activity is bounded; consistency in identifying affected incumbents; and how often projects must repeat analysis after parameters stay unchanged. For commercial transitions, Ofcom should also track whether applicants receive a clear next-step decision, not just whether their temporary licence arrived on time.

Sources