Summary

  • Hull’s full-fibre rollout did not settle how rival networks could reach the remaining streets economically. The next review turns on access to KCOM’s ducts and poles, including what happens when a section is damaged or has no spare capacity.
  • Ofcom proposed that KCOM make necessary network adjustments and recover most costs across infrastructure users up to a £2,725-per-kilometre limit. That is a December 2025 proposal, not the final rule.
  • KCOM and competing operators offered different views of forecasts, repair work and cost sharing. Their submissions are evidence from affected businesses; the authority to set the binding regime belongs to Ofcom, whose statement remained pending on 10 October 2026.

An access right can look complete on paper and fail at the first unusable duct. A rival operator may have a route, customers and capital ready, yet discover that a conduit is blocked, a pole is unsafe or capacity has run out. If the entrant must start again with a new route, the physical-infrastructure remedy may exist in name while the economics still favour the incumbent’s own network.

That is why Hull’s next telecoms review is less about whether fibre reaches the city than about the terms under which existing infrastructure can be reused. KCOM finished its full-fibre rollout in 2019. Ofcom’s market analysis nevertheless describes a vertically integrated incumbent that owns infrastructure, sells directly to households and businesses, and supplies some wholesale access. The question is whether a competitor can make use of the lower-cost route through KCOM’s ducts and poles without inheriting an unpredictable sequence of approvals, delays and bills.

Coverage is not the same as a usable route

The market has changed since Ofcom’s previous review. Alternative network builders expanded in Hull, and CityFibre acquired Connexin’s fibre assets. Ofcom estimated that 70–79% of local premises had access to KCOM and at least one other network in January 2025. It also estimated that KCOM still carried 71–80% of fixed-broadband connections in the data it reviewed. Those measures describe coverage and connections at different points in the evidence base; neither proves that every household can switch easily or that every rival can reach a commercially useful route. [1]

The distinction matters because a broadband supply chain has several layers. Ducts and poles carry the network; the network operator offers wholesale access; retail providers package service, support and price for customers. A rival can compete at one layer and still depend on another. Physical-infrastructure access, or PIA, is intended to let a provider place and maintain its own network in an incumbent’s ducts, chambers and overhead poles. It can avoid duplicated civil works, but only if the route is actually usable and the terms are predictable.

Ofcom said that, until recently, alternative providers had not had access to KCOM’s physical infrastructure and built most of their expansion using their own infrastructure. It expected PIA could matter more in the 2026–31 period, subject to availability. Yet the consultation reports that KCOM’s product had no take-up as of 28 November 2025, while KCOM was processing an onboarding request from one altnet. That is a dated snapshot rather than a current uptake count, but it shows why a published offer and operational adoption must be measured separately. [2]

The remedy begins where the route fails

Ofcom’s December 2025 consultation proposed a specific, area-wide PIA obligation for both broadband and leased-line markets. It would cover access to KCOM’s underground ducts and chambers and its overhead poles, without geographic or use restrictions. The proposal also recognized a practical limit: faulty or full infrastructure can make a section unusable. Ofcom therefore proposed requiring KCOM to adjust its network in certain cases so another provider could use it. [3]

This is not a minor engineering detail. A request to use a duct is an investment decision made before the entrant knows every blockage on the route. If surveys require a complete plan before the operator can inspect the path, a late discovery can invalidate the plan. If a repair needs several rounds of approval, a build schedule becomes contingent on the asset owner’s response. If each unforeseen obstruction is charged to the requester, the entrant’s expected cost becomes a distribution of possible bills rather than a price it can underwrite.

The consultation’s cost proposal tried to allocate that uncertainty. KCOM’s existing offer used a £700-per-kilometre spine-duct limit and split adjustment costs 50:50 between KCOM and the requesting provider up to that limit, according to Ofcom’s analysis. Ofcom proposed instead that KCOM cover the full adjustment cost up to £2,725 per kilometre, recovering it across users of the infrastructure. Costs above the limit would be charged to the access seeker. It chose £2,725 as the midpoint between KCOM’s £700 limit and the £4,750 per kilometre applied to Openreach elsewhere in the UK.

Ofcom said the evidence on Hull’s adjustment frequency and costs was limited, and it invited further evidence. [3]

That distinction changes who can price a route. A shared recovery rule makes the entrant less exposed to one unexpectedly difficult section, but spreads that cost across other users. A requester-pays rule keeps the cost closer to the build that triggered it, but can make entry unattractive before a provider knows how many adjustments a route needs. Ofcom’s proposal drew a boundary: share typical adjustment costs up to the limit, then give the access seeker a direct reason to reconsider a particularly expensive route. Dropwire-related work and certain pole repairs would receive different treatment.

The final instrument could change that balance.

The parties offer evidence, not a substitute decision

CityFibre’s critique of KCOM’s 2025 offer argued that advance approvals, survey restrictions, uncertain timescales, forecasting demands, limited duct occupation and the cost-sharing cap would make the product unattractive at scale. The critique also argued that the requester could discover extra blockages only after committing to a route. These are CityFibre’s claims about the commercial offer; they should be read as an interested network builder’s evidence, not as Ofcom’s finding. [4]

KCOM made the opposite case on important parts of the design. It said its trial generated information for a product suited to Hull, supported proportionate manual processes in a small market, and argued that access seekers should provide forecasts so KCOM could plan resources. It also defended a 50:50 cost-sharing approach as a way to encourage requesters to choose efficient routes. Those are KCOM’s stated positions, not neutral measurements of what each arrangement will cost. [5]

Neither side’s participation settles the question of who may bind customers, operators or the wider market. That is not a reason to exclude their evidence. KCOM controls the infrastructure and knows its operating constraints; network entrants know how approval, survey and repair rules affect investment. Those perspectives help expose failure modes. The regulator’s decision must still explain how it weighs them against the interests the regime is meant to serve. A consultation is a channel for evidence and objection, not a mandate conferred on the most active participants.

The test is operational

The 2026–31 framework is designed to begin on 1 November 2026. Ofcom’s consultation page still showed “Pending Statement” on 10 October, while its plan of work said it intended to publish the decision in October. The proposed £2,725 figure must therefore not be described as an adopted tariff. [6]

When the statement appears, the headline limit will matter less than the whole route through the remedy: Can an operator survey before making a final capital commitment? Can it identify spare capacity? Are response and repair times stated clearly enough to schedule construction? Can a provider perform safe work itself where that is more efficient? Are the adjustment rules and charges consistent when KCOM builds for itself? Are confidential route plans protected from use by the retail business that competes with the applicant?

The evidence should then shift from proposal language to operations: accepted PIA orders, elapsed time from survey to service, the share of orders requiring adjustments, the split between KCOM-performed and self-performed works, cost per route, and wholesale providers actually offering service to premises passed. Ofcom reported that a broader choice of networks was emerging. The new regime still has to show that access to infrastructure translates into deployable routes and retail options.

Sources

[1] Ofcom, Volume 2: Market Analysis · Ofcom consultation page

[2] Ofcom, Volume 2: Market Analysis

[3] Ofcom, Volume 3: Remedies

[4] CityFibre, “Critique of the KCOM PIA Offer”

[5] KCOM, consultation response

[6] Ofcom consultation page · Ofcom, Plan of Work 2026/27