UK software publishing: the margin follows the licence, not the labour

About 197 UK software-publishing companies publish a full profit-and-loss, booking £12.2bn of combined turnover — and the shelf is a lesson in where software profit actually sits. Take-Two Interactive’s UK company puts £668M through 20 employees — over £33M of revenue per head — and keeps about 4p in the pound; its sibling Rockstar Games UK, whose 1,744 developers build some of the most valuable entertainment on earth, keeps 5p. The licence lives in New York, so the margin does too. Sage (UK), whose products are owned at home, turns £414M into £118M of pre-tax profit — 29p in the pound — and the vertical specialists below it run at 25–45%. Meanwhile the single biggest company on the shelf, Capita at £2.42bn, isn’t a software publisher at all: it’s Britain’s best-known outsourcing group, selling divisions as it shrinks — its 2024 pre-tax line only turned positive on the gains from those disposals. Figures are approximate — verify against a company’s own accounts before relying on any single number.

Read the shelf first

This is the publisher’s corner of UK software — companies that sell a product licence many times over, as distinct from the developers and SaaS builders we map separately. Three things change how you read every number below.

The biggest company isn’t a software business. Capita — £2.42bn of turnover, 38,508 staff, revenue down 14% and headcount down 20% as it sells divisions — is an outsourcing conglomerate that happens to sit on this shelf. It is a fifth of the headline turnover on its own, and its profit line is the disposals, not the trade: a £9.9M operating loss in 2024 became a £116.6M reported pre-tax profit only through £184.6M of gains on sold businesses, and its 2025 accounts — filed after our data cut — report a £170.9M pre-tax loss. Alongside it sit other companies that aren’t publishers either: Skyscanner (£540.6M — a flight-search marketplace), Tide (£177.2M — a business-banking platform), Squareup Europe (payments), Vir Health (the Numan men’s-health brand) and Yu Life (group life insurance). Strip them out and something over a quarter of the £12.2bn leaves the room. We set them all aside in the competitive reads below.

The UK arms of global vendors report whatever the group decides. SAP (UK) (£1.32bn at a 3.6% margin), VMware UK (£396M, a £39.3M loss), Dassault Systèmes UK, GitHub’s UK company, Claris (Apple’s FileMaker arm — £24.8M through 11 staff at a 48.6% “margin”) — these are sales-and-support or licensing entities whose revenue and margin are transfer-pricing settings, not market outcomes. Never benchmark them against a business that owns its own code. The Take-Two pair is the cleanest demonstration: the same group runs a 20-person licensing conduit at a 4% margin and a 1,744-person studio at 5% — neither number describes how good the games are, only where the group keeps the profit. (What the studio actually banks is another story: after the UK video-games tax credit, Rockstar’s after-tax profit is £87M — the relief, not the recharge, supplies most of that margin.) (The studios themselves are mapped in our video-games report.)

Losses at the top aren’t always distress. VMware UK and Aveva Solutions (£374.2M, −£32.5M — the Cambridge industrial-software house now owned by Schneider Electric) both lose money at the entity level; for arms of global groups, restructuring charges and group recharges can swing the pre-tax line as much as trading does.

The giants

† Capita’s £116.6M is not a trading profit: the group made a £9.9M operating loss in 2024, and £184.6M of gains on business disposals lifted the pre-tax line; its 2025 accounts, published in July 2026 after our data cut, report a £170.9M pre-tax loss. ‡ Take-Two’s accounts are US-dollar denominated ($892.8M of turnover), converted here. Rows use each company’s latest accounts at our data cut, and year-ends differ — Skyscanner runs to December 2025, Take-Two and Rockstar to March 2025, Anaplan to January 2025, SAP and Capita to December 2024, Sage to September 2024 and Finastra to May 2024; Sage and Finastra, like Capita, have filed newer accounts since.

Two patterns run down the table. First, the profitable software lines cluster where the product is owned or run from Britain — Sage at 29%, Finastra at 23%, Anaplan at 24% — while the pure sales arms sit at low single digits. (Finastra’s 23% includes dividend income from its German subsidiary — its trading margin is nearer 18%.) Second, look at the last two columns: almost every enterprise vendor is growing revenue while cutting people. SAP +6% / −6%, Sage +5% / −5%, Finastra +10% / −12%, Anaplan +13% / −18%. That is the industry-wide post-2023 playbook — defend the margin, shrink the payroll — and it shows up here with unusual consistency. Skyscanner’s +39% revenue on −39% headcount is the extreme; read it as group reorganisation of where people sit as much as a trading result — and its £47.6M profit line is mostly dividends and licence fees from group companies, another reason not to read it as one. Towers Watson Software’s +60% is a group vehicle’s growth — intercompany flows can drive it as easily as new sales.

The shape of the market

Below the giants this is a genuinely healthy mid-market. The £5–25M band holds the most companies (71, two-thirds profitable) and profitability climbs steadily with scale — 71% at £25–100M, 76% at £100M–1bn. The graveyard is the sub-£1M tier, where only 38% make money: part young product companies burning towards scale, part products that never found a market.

The engine room: vertical specialists at 25–45%

The best economics on the shelf belong to mid-sized companies that own the standard software for one unglamorous vertical. Road Tech Computer Systems — haulage compliance software from Hertfordshire — makes a 40% margin with 85 staff. iPipeline (life-insurance quote software) runs at 43%, Glass’s Information Services (the vehicle-valuation “Glass’s Guide”) at 34% with 29 people, Stark Software International (energy-data analytics) at 31%, Unilink Software (custodial and prison systems) at 28%, Brighton’s Virtusales (software for book publishers) at 26%. Once a vertical’s workflow runs on your product, it rarely leaves — and the margins read like annuities.

…and another two dozen profitable mid-market publishers behind them. The raw ranking would also include Squareup Europe (£84.0M at a 44.5% “margin” through 19 staff — a payments company, not a publisher), Blackbaud Global (a 93.4% “margin” with no employees — a group vehicle whose profit line is dominated by non-trading income) and the UK arms of GitHub, Wolters Kluwer, Linedata and Objectway — group companies whose margins are set at head office. We’ve set those aside: none of them tells you anything about running a software business in Britain.

Growth, read with care

Most of the fastest “software publishing” growth isn’t software publishing. The top of the growth table is venture-funded platforms buying scale at a loss: Vir Health (Numan, +132% to £69.3M, still loss-making), Tide (+56% to £177.2M, −£13.2M), Yu Life (+99% on a £27M loss — an insurance business in any case), Gravitee (API-management software, +67%, a £12.3M loss that reads as venture spend) and Timeline Holdings (retirement-planning software and funds, +110%, −£4.1M). The classic pattern: revenue bought ahead of profit.

The genuine signal — profitable growth backed by hiring — is thin: tiny Activecore (+184% revenue, +260% staff, profitable) and, at ten times the size, Azerion UK (+61%, profitable, modest hiring — though as part of a Dutch ad-tech group, some of that growth can be group reallocation). Among the established names, Mercury xRM is the quiet compounder: profitable, growing, and adding people.

Market structure: a top-heavy shelf where the top isn’t software

On paper this is a concentrated market — the top 5 hold 44.6% of the £12.2bn, the top 10 hold 60%. In practice the head of the curve is mostly companies whose economics answer elsewhere: an outsourcing conglomerate, a flight-search engine, two Take-Two group entities and the sales arms of SAP and VMware. Strip those out and British software publishing proper is a much flatter market — one national champion in Sage, a band of £100–400M product companies, and a long healthy tail of vertical specialists. Note also who’s missing: Microsoft’s, Oracle’s and Adobe’s UK companies sit on other shelves entirely (Oracle’s appears in our wider software map), so no single view captures “software in Britain”.

Ownership and vintage

Of the 197 companies, 95 are corporate-owned against 74 owned by individuals, and about 11% carry a Holdings/Topco/Bidco-style name — the structural fingerprint of a buyout or a planned exit (Netsparker Holdings, Gravitee Topco, Timeline Holdings among them). Private equity has learned what the margin table shows: a vertical software product with captive customers is about as annuity-like as a trading business gets. The vintage profile says this is a young shelf that keeps renewing — the 2000s cohort is the largest (55 companies), but 58 of the 197 were incorporated since 2016, the venture and buyout wave that produced most of the loss-making growers above.

What the map shows

1. The margin follows the licence, not the labour. Take-Two’s 20-person UK conduit keeps 4p in the pound and its 1,744-person Rockstar studio keeps 5p — while Sage, whose products are owned in Britain, keeps 29p. Where the IP lives decides whose accounts show the profit.
2. The biggest name on the shelf isn’t a software company. Capita’s £2.42bn is a fifth of the headline turnover; add the marketplaces, banks and insurers shelved here and over a quarter of the “market” isn’t software publishing at all.
3. The engine room is vertical specialists at 25–45% — haulage compliance, life-insurance quotes, vehicle valuations, prison systems. Own the standard product for one vertical and the P&L reads like an annuity.
4. The enterprise playbook is revenue up, headcount down. SAP +6%/−6%, Sage +5%/−5%, Finastra +10%/−12%, Anaplan +13%/−18% — margin defence through payroll is near-universal at the top of this table.
5. Fast growth is mostly bought at a loss — Numan, Tide, Yu Life and the venture-funded platforms dominate the growth table; profitable, hiring-backed growth is as rare here as in any market we’ve mapped.
6. This is one slice of UK software — the developers and SaaS builders and the games studios each have their own map, and the American giants’ UK companies sit on other shelves again.

Methodology and caveats

This covers only the 197 UK software-publishing companies that publish a full profit-and-loss; the long tail files abridged accounts with no revenue figures and doesn’t appear, and several of the world’s biggest software vendors report their UK business through companies mapped in our other reports. Many of the largest entities here are UK arms of global groups whose revenue and margin reflect transfer pricing and group recharges rather than market performance — their figures are reported but never benchmarked against independent businesses, and some report in dollars or euros, converted at the relevant year’s rate. Each row uses the latest accounts available when the data was assembled in early July 2026 — year-ends span May 2024 to December 2025, so rows are not all the same trading year, and a few giants (Capita, Sage, Finastra) filed newer accounts between our data cut and publication. Companies that plainly aren’t software publishers (an outsourcing group, marketplaces, payments, banking and insurance platforms) are shown where scale demands it but excluded from the competitive reads. Extreme proportional outliers are excluded from the charts. Figures are approximate and business labels are directional — verify any specific figure against a company’s own accounts. This is analysis, not financial advice.