UK telecoms: owning the network prints money, renting or digging one doesn't

UK telecoms tells one story three ways: the profit sits with whoever owns a network, and both ways of not owning one are expensive. EE cleared £1.43bn of pre-tax profit on £7.05bn running Britain’s biggest mobile network. Tesco Mobile, which books over a billion pounds a year selling airtime it buys wholesale, kept £1.2M of it. TalkTalk, a reseller at real scale, lost £465M and cut a quarter of its staff. And the challenger fibre builders digging new networks from scratch are the bleeding edge in the literal sense — the four heaviest-burning groups lost roughly £133M between them on £27M of combined revenue. Across the 406 UK telecoms companies that publish a full profit-and-loss, the quiet winners turn out to be B2B specialists most consumers have never heard of. Figures are approximate — verify against a company’s own accounts before relying on any single number.

Read the register first

Three things change how you read every number below.

The headline total isn’t the UK market. These companies book £117.5bn of combined turnover, but this register holds the corporate spine of telecoms, not its shop floor — and the spine files in layers. BT Group (£20.37bn) and its main operating company British Telecommunications plc (£20.36bn) both appear; so do TalkTalk and its holding company; so do Vodafone Group and its internal services company. Strip the visible duplicates and the total is nearer £93bn — and most of that is Vodafone’s worldwide group consolidation, earned largely outside Britain. Treat the register as a map of who’s here and how they’re doing, not a market-sizing.

Vodafone’s register profit line is not the group’s result. The register carries £18.99bn of pre-tax profit against Vodafone Group, but that is the parent company’s standalone figure — largely intra-group dividends flowing up from subsidiaries, not trading. The consolidated group actually reported a €1.5bn pre-tax loss in its latest year, after €4.5bn of impairments in Germany and Romania, on revenue of €37.4bn (+2%, in euros). Its operating reality is a giant holding steady — not a 50% margin and not a €1.5bn crisis. We exclude it, and the other group-layer entities, from every competitive read.

Margin is not margin here. A virtual mobile operator that books the full airtime bill as its own revenue (Tesco Mobile: 0.1% margin) cannot be compared with brand-and-licensing entities showing 40–73% margins on the same shelf, or with an infrastructure owner monetising scarce capacity (Inmarsat: ~35% at the operating line, before intra-group financing income swells its pre-tax figure to 62%). The spread between those numbers is accounting structure and business model, not management skill — and that spread is most of this report’s story.

The giants: one spine, three economics

The top of the register splits into network owners, resellers and pass-through vehicles — and the profit column tracks that split almost perfectly.

…and around fifty more companies above £100M of turnover.

*Group layers, noted not double-counted: British Telecommunications plc (£20.36bn, £2.06bn PBT) sits inside BT Group; TalkTalk Holdings (−£607.0M) is TalkTalk one layer up, its bigger loss carrying the group’s debt burden; Vodafone Group Services (£2.73bn) is Vodafone’s internal services company. Vodafone’s consolidated loss follows €4.5bn of impairments in Germany and Romania; the £18.99bn profit the register carries against the parent company is its standalone result — largely intra-group dividends — not a trading number, which is why the table shows the consolidated line instead.

†Inmarsat’s latest filing is denominated in dollars and covers a fifteen-month period (a year-end change to align with its parent, Viasat); the sterling figures here are converted. Roughly 44% of that pre-tax profit is financing income — largely intra-group interest on receivables from other Viasat companies — not satellite trading. Operating profit was ~£476M, a ~35% margin.

The reads worth taking. EE is the cleanest profit machine on the register — £1.43bn of pre-tax profit at a ~20% margin, revenue and headcount both roughly flat: a mature network being harvested. Inmarsat looks even more extreme — £849M of pre-tax profit on £1.37bn — but read the footnote: the filing covers fifteen months, and nearly half of that profit is intra-group financing income rather than satellite trading. The trading result — roughly £476M of operating profit, a ~35% margin — is still the economics of owning scarce orbital capacity, a margin no terrestrial operator can copy. Colt’s +32% is bought growth — it acquired Lumen’s European business — and the €249M loss (≈£211M; Colt files in euros) carries the integration. Telecom Plus is on this shelf as a telecoms company but most of its £1.84bn is bundled gas and electricity, so its −10% tracks falling energy prices more than customer loss. TalkTalk is the register’s distress story: revenue −7%, headcount −24%, a £465M pre-tax loss at the operating group and £607M at the holding company above it — what a decade of renting your network at value-brand prices looks like when the debt comes due. And Ciena is a reminder to read names before margins: it’s an equipment vendor’s trading entity, not a service provider, and we leave it out of the competitive reads.

Where the money is: the B2B niches nobody’s heard of

Between the giants and the long tail sits the register’s most instructive band: profitable mid-market operators at £5–100M of turnover. Almost none of them sell connectivity to households. They sell around the network — software, billing, messaging, installation, and connectivity for places the big networks don’t reach.

…plus a tail of business-phone, radio and reseller specialists (4Com, Radiocoms, Convergence, Core Communication) in the 10–13% band.

*Handle the outliers with care: the Lebara entities’ ~40% margins (a second Lebara company shows 42.5%) look like brand-and-licensing group structure rather than what selling SIMs earns — read them alongside Tesco Mobile’s 0.1%, which is what booking the full airtime bill looks like. Talkmobile, Vodafone’s budget sub-brand, shows a 72.7% margin with no visible staff — an intra-group arrangement, not a retail business, so we exclude it. Hearst Networks UK (13.3%) is a television-channels business shelved here by classification accident and is excluded too.

The pattern is consistent: the best genuine margin in mid-market telecoms belongs to whoever sells software or scarce access, not connectivity. Cerillion, a listed billing-software house, converts nearly half its revenue to profit. Tampnet earns 34.6% connecting North Sea platforms — 14 staff, £14M of profit — because it owns the only network where its customers are. Fonix takes 19.8% clipping payments that ride on other people’s phone bills. Meanwhile the actual connectivity businesses in this band cluster at 11–17%, respectable but unremarkable — and the ones below that line don’t make this table at all.

Growth, read with care

Ranked by raw growth, this register is a warning label. Most of the fastest “growers” are loss-making fibre challengers ramping revenue from a tiny base while burning capital, and much of the rest is acquired or pass-through.

The altnet story is the one to watch. Toob (−£52.8M on £14.0M), the Full Fibre group (−£52.5M on £5.1M at its consolidating holdco Iris Infra, with Full Fibre the operating layer inside it — one group, counted once), 4th Utility and Cohiba are all losing several multiples of their revenue digging challenger networks — that is what building infrastructure looks like in year three, and the staff columns say the phase is turning. Toob is still hiring into its losses (+43%); Full Fibre, 4th Utility and Cohiba are cutting roughly a third of their staff while the red ink flows — the consolidation and retrenchment stage of Britain’s fibre land-grab, arriving on schedule.

The remaining rows are different stories. The One Touch Switching Company — the body the industry was required to set up so customers can switch broadband provider in one step — is the only entry that is both fast-growing and profitable, funded by the providers themselves; a nice irony that the best new business model in telecoms is the one the regulator ordered into existence. And Aurora Networks International is a Ciena-style catch: it’s the renamed UK trading arm of CommScope, the US network-equipment maker, so its +331% to £202.5M is equipment trading, not carrier volumes — we leave it out of the competitive read for the same reason. Reliance Jio’s UK arm is genuine pass-through at a 0.4% margin: £88M of revenue on its way somewhere else.

The shape of the market

The register is bottom-light and top-heavy in a way few industries are. Below £1M sits a graveyard (36% profitable); the £1–5M band barely exists, because small resellers rarely publish a full profit-and-loss; and then a substantial mid-market appears — 138 companies at £5–25M and 101 at £25–100M — with roughly two-thirds profitable in every band all the way up. Losses here don’t concentrate by size; they concentrate by model (the fibre builders lose money at every scale they appear).

Concentration is the steepest we’ve mapped: the top five companies carry 74.9% of the register’s turnover (road freight: 38%; nurseries: 13.5%), and the top twenty carry 87.6%. That’s what a national infrastructure industry looks like — two incumbent groups and a handful of global entities over a wide, shallow field of specialists.

Ownership and vintage: the deregulation generation, then the fibre wave

The founding profile carries two clear signatures. The 1990s and 2000s cohorts (89 and 139 companies) are the deregulation-and-dotcom generation — the resellers, wholesale carriers and B2B specialists that grew up after the market opened. Then the 2010-onward cohorts (147 companies) hold the altnet wave: nearly every loss-making fibre builder above was incorporated after 2010. About 18% of the register carries a Holdings/Topco/Bidco-style name — private-equity structure — and it clusters exactly where you’d expect: in the fibre build-out, where patient capital is funding today’s losses for tomorrow’s network.

Of the 406 companies, 225 are corporate-owned and 150 individual-owned — a corporate register by nature, where even the “independents” are often the local arm of a global carrier.

What the map shows

1. Network ownership is the profit line. EE cleared £1.43bn on the mobile network it owns; Inmarsat earns a ~35% operating margin on satellite capacity nobody else has. Owning scarce infrastructure is the only reliably fat margin in telecoms.
2. Reselling keeps pennies. Tesco Mobile kept £1.2M of £1.11bn; TalkTalk — renting its network at value-brand prices, at scale, with debt — lost £465M and cut a quarter of its staff.
3. The fibre land-grab has hit its bleed phase. The four heaviest-burning altnet groups lost ~£133M on £27M of combined revenue, and three of the four are now cutting staff — retrenchment and consolidation, not collapse, but the build-at-any-cost era is over.
4. The quiet winners sell around the network, not through it. Billing software (Cerillion, 47.9%), offshore connectivity (Tampnet, 34.6%), carrier billing (Fonix, 19.8%) — the best mid-market margins belong to software and scarce access, while connectivity itself earns 11–17% at best.
5. Margins here are mostly accounting, so compare with care. The same shelf shows 0.1% and 73% “margins” on similar activities; the spread is revenue model — principal airtime vs brand fees vs intra-group arrangements — not operational skill.
6. It’s the most concentrated market we’ve mapped — the top five carry 74.9% of turnover — but the £117.5bn headline is group consolidation (much of it earned worldwide), not the UK market.

Methodology and caveats

This covers only the UK telecoms companies that publish a full profit-and-loss — 406 of a register of around 1,143, so the long tail of small resellers and dormant vehicles doesn’t appear. Several groups file at more than one layer (BT, TalkTalk, Vodafone); tables use one entity per group and note the others, and the combined-turnover headline still contains group-level consolidation of overseas revenue, so it should not be read as a UK market size. Vodafone and Colt file in euros (their table rows carry € figures) and Inmarsat in dollars over a fifteen-month period; Vodafone’s row shows the consolidated group result rather than the parent company’s standalone profit, which is largely intra-group dividends. Business-type labels are directional; mis-shelved companies (a TV-channels business, two equipment vendors’ trading arms, a defence contractor) are excluded from competitive reads where identified. Large losses at the fibre builders are build-out economics and may include financing and impairment charges rather than pure trading losses. Figures are approximate — this is analysis, not financial advice; verify any specific figure against the company’s own accounts.