The Trillion Dollar Concierge
Comparative lessons for Britain's new scale-up strategy.
Introduction
Britain wants to grow its first trillion dollar company. At London Tech Week, Business Secretary Peter Kyle unveiled a "concierge service" in aid of that ambition: a single point of contact inside Whitehall, offering escalating tiers of support depending on a firm's size and promise.
Around that desk sit several supporting measures. There is a Global Talent Taskforce, a Visa Fees Reimbursement Scheme worth up to £5,000 per international hire, and a fast-track referral from the Office for Investment to speed up the sponsor licence that any scaling firm needs to recruit from abroad. One further measure is still taking shape: a contractor is being tendered to build the Scale-Up Catalyst Pilot, which will develop a pipeline of pre-scale companies.
Taken on its own terms, this is a good idea. A dedicated desk that clears administrative friction for fast-growing firms is cheap, sensible and well-targeted. The international evidence suggests that competent facilitation of this kind can earn its keep, though it is a good deal more particular about the conditions than the announcement assumes.
But what is the concierge actually for? Growing a trillion dollar firm is a very different objective from smoothing the path of a scaling business, and it calls for very different instruments. The concierge can do the second job well; the first is another matter entirely. Most of the weakness in this announcement flows from yoking a sound policy to a giant ambition it cannot reach.
This suggests two questions, which need to be held apart. The first is narrow: if a country runs a concierge service, how does it run a good one? The comparative evidence on that turns out to be surprisingly rich, and it points in a clear direction. The second question is broader: what actually determines whether a country produces a firm of extraordinary scale, and can a concierge touch it at all? Here the evidence points elsewhere, towards the depth of a nation's capital markets and the underlying cost of doing business on its soil.
This essay takes each question in turn, and ends by returning the concierge to its proper — and useful — place.
What is a concierge service?
At root, a concierge service is simply a dedicated point of contact who deals with a problem directly rather than routing it through a standard process. Governments have borrowed the term for a similar function at firm level: someone sits inside the state, and their job is to clear whatever is blocking a company's growth instead of leaving it in the Whitehall queue.
That is the easy part of the definition, and it is also roughly where the agreement ends. The same word has been bolted onto at least four mechanisms that have almost nothing in common with each other.
The cleanest way to tell them apart is to ask four questions of any scheme. Does the state put its own capital into the firms? Does it pick the winners itself, or leave that to others? Can it coerce, making participation something a firm cannot afford to refuse? And is the institution built to last, or is it a pilot with a launch date and no expiry? Those four axes (capital, selection, coercion, permanence) sort the international "concierge" menu into recognisable types.
At one end sits pure facilitation: the state is used to reduce friction from interactions with itself, handling paperwork, permits and introductions, without committing capital or selecting winners. Ireland, Italy and India have all implemented versions of this. A step along is the equity stake, where the state commits real money in exchange for capital and access, as Abu Dhabi's Hub71 does.
A third type, certification by proxy, keeps the state out of capital markets but very much in the business of selection: outside nominators put promising firms forward and a government committee blesses them, as in Japan's J-Startup. At the far end sits leverage: facilitation welded to coercive power, where a firm that declines to participate is shut out of something it needs. Saudi Arabia's headquarters programme is the purest case, and I will return to it later.
These are not mere academic distinctions; they reflect the instruments each economy actually has at its disposal. A state with little to offer beyond institutional know-how competes on facilitation, because that is the only lever available to it. A state with a sovereign wealth fund can buy a stake. A state that is the dominant buyer in its own market can coerce.
The equity stake is the model most easily mistaken for a free lunch. A state that takes a position acquires more than a share of the upside. It acquires influence too: board seats, informational rights, and a hand in decisions a private investor would otherwise leave to management. The firm's incentives then bend away from value maximisation and towards the state's second-order political objectives, whether that is employment, national prestige, or strategic aims that cut against profit (cf. China's "golden shares"). An equity stake smuggles selection and direction into what looks like passive capital, and at the limit shades into coercion. Which is worth remembering before deciding that Britain, lacking the firepower to take stakes at scale, loses much by being confined to facilitation.
Britain has a foot only in the first of these four models. It is not entirely without state investment vehicles — the British Business Bank and the National Wealth Fund both deploy public capital — but neither operates at the scale that would let Britain take meaningful stakes in its scaling firms the way a Gulf sovereign fund can, and nothing in the British state resembles the directed procurement power on which coercion depends. It sits in the facilitation category by default. That matters considerably when asking what kind of concierge Britain can credibly operate, as distinct from the kind it might aspire to.
What facilitation can do
Facilitation is the oldest of the four models, and the one with the most evidence behind it. Its exemplar is Ireland's Industrial Development Authority (IDA), which has run without interruption since 1949 and now works with more than 1,800 client companies employing over 300,000 people.¹ Its 2025 figures were the best in its history: standing proof that a facilitation desk, given enough time, compounds.
It is also a warning against reading too much into the brochure. A great deal of Ireland's headline success of course reflects the tax-driven location of multinational subsidiaries rather than anything the agency itself conjured. Apple alone is reckoned to move a remarkable share of measured Irish output.² The effect of an agency cannot simply be read off its own annual report, and a serious assessment should not pretend otherwise.
This is the selection problem, and once you've seen it you see it everywhere. Schemes of this kind tend to attract firms that were already heading in the right direction, and then claim the credit for where those firms end up. The clearest case is Start-Up Chile, the world's first government-run public accelerator, which from 2010 handed equity-free grants and a visa to founders of any nationality. The principal academic evaluation of the programme found that its basic offer of funding and co-working space had no measurable effect on venture performance on its own; what carried a firm was whatever it already possessed when it arrived.³ The accelerator was, in effect, taking credit for the weather.
So is there cleanly identified evidence that facilitation works — and, just as important, evidence that bears on a country like Britain? The most-cited finding does not say quite what an enthusiast would want it to. Harding and Javorcik, looking across more than a hundred countries, found that the sheer professionalism of an investment promotion agency — how competently it answers an enquiry and how usable its materials are — is associated with higher foreign investment, independent of any financial sweetener attached.⁴ One widely cited estimate puts the payoff at around $189 of additional inflows for every dollar an agency spends.5 It is the strongest single piece of evidence that facilitation pays.
But it comes with an important caveat. The same study finds the effect concentrated in developing countries, where red tape and information asymmetries are most severe, and finds that it does not show up in industrialised economies at all.6 Britain, plainly, is an industrialised economy. Read carelessly, the paper that seems to vindicate the concierge instead warns that promotion of the generic kind buys a rich country very little.
The usual explanation for this — that developing economies simply have more friction to clear — is not the only one. Where institutions are weak, a facilitation desk can be the official channel through which a firm navigates, and pays for, access to a state that is otherwise capricious. It does the work that functioning rule of law does for free elsewhere. In the worst cases, it is the polite face of the rent extraction a firm must engage in to operate at all.
The line between facilitation and coercion can blur more than the four-part test implies. Where the unaided alternative is prohibitively costly, being denied access to the desk is itself a penalty – coercion by omission rather than imposition. Much of the measured return to promotion in poorer countries is the value of an institutional shortcut that a functioning rule of law economy does not need. The headline figure therefore overstates what a British concierge could hope to capture, and by a wider margin than the rich country comparison alone suggests.
For Britain, the question is whether it has particular pockets of friction severe enough to justify a concierge — and it arguably does. The sponsor-licence regime, visa costs and delays, and the weeks a scaling firm loses to immigration paperwork are exactly the kind of high-tax compliance frictions whose removal is rewarded, even in an advanced economy. Here facilitation can lift outcomes at the margin, in the specific places where friction is significant, provided one measures honestly enough to tell the marginal effect from the weather.
One Irish design choice deserves singling out. The Irish keep the agency that courts foreign investors, the IDA, separate from the one that backs homegrown firms, Enterprise Ireland. World Bank evidence finds that agencies with this kind of narrow remit pull in more investment in their target sectors than agencies asked to do everything at once.6 Britain's single desk, by contrast, is meant to serve newly arriving investors and domestic scale-ups simultaneously: precisely the conflation that finding warns against.
The trouble with lists
Of the four models, the one structurally closest to what Britain has announced is not facilitation at all. It is certification: an official and curated list of recognised high-potential firms.
Japan's J-Startup, launched in 2018, is the clearest case. Tellingly, the Japanese government does not pick the firms itself. Outside recommenders — including venture capitalists, accelerators and corporate innovation leads — put names forward, and a committee certifies the ones that pass. Certification then unlocks preferential access to subsidy and procurement, regulatory fast-tracks, help with visas and funded slots at international showcases. By early 2025, 239 firms had been certified across five rounds.7
Does it work? Seven years in, the honest answer is that there is not yet enough data on exits, valuations and overseas revenue to say whether certification produces firms that break out abroad, or whether it mostly works as a credibility stamp: handy for raising the next round, and neutral as to whether the firm ever wins in a world market.
There is, though, a problem with certification that can be reasoned about in advance. A certified list is a scarce, rivalrous privilege. The firms on it enjoy preferential access to public money and public contracts; the firms off it do not. Those on the list have every incentive to organise to stay on it and to keep newcomers off. The far larger, far more diffuse population of excluded firms has almost no capacity to organise in response. The conflict of interest can reach inside the nomination process itself, since a venture fund or corporate scout asked to put names forward is rarely indifferent as to which names succeed.
This is simply George Stigler's canonical account of regulatory capture, transposed: a scheme "designed and operated primarily for the benefit" of its concentrated insiders, because concentrated insiders always out-organise a dispersed majority.8 The same knowledge problem that defeats a utility regulator trying to divine an efficient price defeats an official trying to divine a future champion. Worse, once a list exists, the firms on it acquire a stake in the official's continued ignorance.
Pure facilitation — the kind that handles thousands of all-comers and turns nobody away — is largely immune to this, for the simple reason that there is no scarce list to fight over. The important word is "pure," and it is exactly where Britain's scheme runs into difficulty.
What has Britain actually built?
Held up against the four-part test, the announcement comes out mixed.
The visa reimbursement and the fast-track referral from the Office for Investment are pure facilitation. They lower an administrative cost and shorten a wait, with no equity taken and no winners named. So far, so Irish.
Two features pull the package the other way, towards the list. The first is scale. Officials expect the concierge to support dozens of firms, not the thousands a serious facilitation desk handles in a year. A cohort that size sits much closer to a J-Startup round of thirty or fifty than to anything the IDA does. The second is the showcase: the launch named individual firms rather than describing the kind of firm that might apply, with Oxford Quantum Circuits, Quantexa, Wayve, Elliptic, TreQ, and Yonder all held up by name. A facilitation desk has no need of a showcase; naming firms is what certification programmes do.
It would be easy to overstate this. Picking six firms for a launch is a comms decision, and the size of a cohort is a function of budget as much as of design. Neither proves that Britain has built a closed, rivalrous list of the capture-prone kind, and on the present evidence it plainly has not. But facilitation is immune to capture only when it is pure: only when it turns nobody away and there is nothing scarce to compete over. A desk that serves a few dozen named firms is not that. The features that distinguish Britain's scheme from the Irish model — the small cohort and the public roll of honour — are precisely the features that reintroduce scarcity and rivalry.
One element works plainly in the policy's favour. The scheme targets a defined sector and stage. Eligibility runs through the OECD's standard definition of a high-growth firm: ten or more employees growing at around twenty per cent a year for three years. The government estimates there are around 44,600 such firms in the UK, under one per cent of the total.⁹ The same World Bank evidence finds that sectoral targeting of this kind has a measurable and positive effect on investment into the targeted sectors.¹⁰ It is the soundest part of the design, and it should be kept.
There is one model on the menu that Britain cannot copy. Saudi Arabia's headquarters programme is effective because, from 2024, most multinationals without a regional head office in the Kingdom are barred from bidding for government contracts at all.¹¹ That kind of leverage is available only to a state that dominates demand in its own economy. Britain is not that buyer, and no policy can turn it into one. A facilitation programme can make itself more attractive; an equity programme can offer better terms. But neither can force a firm's hand.
All this sits within a broader case for free markets and limited government — one that bears directly on what follows. The instinct behind certification and equity stakes is the same instinct behind industrial policy in general: the belief that officials, given the right framework, can spot winners earlier or better than markets do. The empirical record for that belief is poor, and economists have been saying so since Stigler's generation. Facilitation escapes the objection only because it does not ask government to be right about which firm will succeed; it only asks government to get out of the way faster for whichever firms succeed on their own. That, and not the showcase, is the part of Kyle's scheme worth defending.
Making the concierge work
The concierge is worth running, though it is a smaller thing than its billing suggests. Everything above has argued that the desk is, at best, a marginal instrument. It cannot reach the trillion dollar ambition under which it was launched, and even its narrower promise holds only where Britain's administrative friction is truly severe.
It is fair to ask whether such a thing should exist at all. A concierge is, after all, a second-best. The first-best is not to build the maze: a tax and regulatory environment simple enough that no firm needs a guide through it. A guide, moreover, has an interest of its own. Its usefulness depends on the maze remaining intricate, and a desk that lives off the complexity it navigates will rarely be the body that argues for the complexity's removal.
This is the rentier objection, and the honest answer is that the maze will not be cleared overnight. But while it stands, clearing its worst chokepoints cheaply is still worth doing, provided the desk is disciplined enough not to become a lobby for keeping it in place. It is also why the larger share of this essay's recommendations, set out at the end, are aimed at the maze itself rather than at the guide.
Three things are likely to go wrong, each grounded in the case studies above or in Britain's own record, and each has a straightforward fix.
Passive sourcing
Britain is not starting from scratch. The Department for Business and Trade already spends around £80 million a year and fields some 634 staff across an overseas network of more than ninety posts in nine regions, each led by an HM Trade Commissioner with a formal remit to develop investor relationships. On paper, that is not so different from the active outreach that makes the IDA, or Invest India, effective. In practice, Parliament's own spending watchdog found that the Department does not even survey the investors who look at Britain and decide against it. There is no systematic way of learning why opportunities are lost. The Office for Investment, responsible for the very largest projects, told the Committee it had worked on only around fifty in total.¹²
The temptation is to call this a resourcing failure and reach for more staff and a better feedback loop. That would misread the problem. The information a concierge would need in order to source actively — the knowledge of which little-known firm is three years from breaking out, and where to find it — is not sitting in a database waiting for a larger team to retrieve it. It is dispersed across thousands of private balance sheets and order books, revealed slowly and only through the test of market competition. A desk that waits for firms to present themselves will clear blockages for the firms already known to officials, which is to say the firms with the lobbying capacity to be known.
The countries that have solved this did not do so by hiring more civil servants. Singapore's Economic Development Board, established in 1961 and still running, embeds its officers directly inside industry clusters: electronics, biomedical, financial services. It rotates them through private sector secondments so that knowledge flows both ways. The Netherlands Foreign Investment Agency takes a narrower version of the same approach, using private sector specialists on short-term contracts to cover sectors the state has no natural visibility of.
You cannot centralise knowledge that is inherently dispersed, but you can put people — or institutions — close enough to reduce the guesswork. J-Startup delegates the sourcing problem to outside recommenders because it knows accelerators and venture funds already have information a government committee never will. The nomination should travel up, rather than the search travelling down.
Britain's overseas network is large enough to do something similar. The fix is not structural so much as directional. We could open the referral pipeline formally to private intermediaries: the venture funds, accountancy networks, law firms and angel syndicates already embedded where the unheralded scale-ups are being built. The civil servant's job then becomes processing referrals and clearing blockages, not finding firms.
Institutional impermanence
Britain has a habit of launching business support schemes with a stated lifespan, and then closing them well before that lifespan is up. Help to Grow: Digital was meant to run for three years; it shut to new applicants after two, with fewer than a thousand vouchers ever redeemed. The Industrial Strategy Council was abolished in 2021 with nothing put in its place, taking its own success metrics down with it.
The obvious answer is to legislate a minimum life — five years, say — before any concierge-branded scheme is allowed to relaunch. That is a sound instinct, but it secures less than it seems. A statute does not bind the government that passes it and binds its successor even less. Schemes die for a deeper reason: the diffuse constituency that benefits from continuity is always outvoted by the concentrated constituency that benefits from the next announcement, the relaunch, the rebrand and the fresh ribbon to cut.
That asymmetry holds so reliably because a launch is politically visible in a way that maintenance is not. A new scheme yields a photograph, a figure, and a minister's name attached to something fresh and shiny. An incumbent scheme yields none of these. Its benefits are spread thinly across many firms over many years, and none of them will necessarily lobby to keep it. Help to Grow and the Industrial Strategy Council failed partly because nobody's standing was bound up with their survival, whereas a fresh relaunch always has a sponsor.
The achievable goal might be to invert the default — to make closure something that has to be actively chosen and defended, rather than something that happens through neglect. That requires a constituency whose own interests are bound up with the desk's survival.
The referral pipeline can supply that constituency. If the desk sources its firms through private intermediaries — the venture funds, accountancy networks, law firms, and angel syndicates mentioned previously — then those intermediaries acquire a direct stake in its continuation. Unlike the firms the desk ultimately helps, who are diffuse and move on, the intermediaries are concentrated and organised. This of course also introduces a rentier interest that good institutional design is meant to avoid, and that is why the constituency must be paired with a pre-registered performance metric, as set out in the next section.
Metrics
Third, and connected to the second: what counts as success? It might mean more firms crossing the OECD's scale-up threshold; it might mean fewer scale-ups decamping to the United States. It might mean the trillion dollar firm itself, the ambition under which the whole thing was launched. These can change independently of one another, as Start-Up Chile demonstrated. A government that has not fixed its metric in advance is free to claim victory against whichever number happens to have risen, while its critics are equally free to claim failure against whichever number has not.
Israel's Office of the Chief Scientist, now the Israel Innovation Authority, handled this differently. Its evaluations of R&D support programmes used matched-firm methodology from early on. They found comparable firms that did not receive support and tracked both groups over time, so that the counterfactual was explicit and the credit-claiming problem had to be confronted rather than papered over.¹³ The results were not always flattering, but they were trusted. That meant the programmes that survived scrutiny carried legitimacy, and the ones that did not could be shut down without a political fight.
The concierge's primary success metric — a single ranked measure rather than a menu of possibilities — should be lodged with the National Audit Office (NAO), alongside an explicit counterfactual methodology and a publication timetable no minister can amend. A government serious enough about growth to announce a trillion dollar ambition should be serious enough to say, in public and in advance, exactly what it would accept as evidence of failure.
Demolishing the maze
Suppose Britain ran the best concierge in the world. Would it then grow its trillion dollar firm?
Almost certainly not. A trillion dollar company is a draw from the extreme tail of a steeply skewed distribution of firm value, and a concierge cannot reach it. The first reason is the knowledge problem already met in the discussion of certification. Which firm becomes a giant is not visible in advance, not because the data is missing but because the information is created only by the test of competition itself. The second follows from the instrument: a concierge works firm by firm, on firms that have already presented themselves — which means, by construction, that it can only ever act on the identified middle of the distribution, never on the unidentified tail.
The third reason is about inputs. What produces giants is a set of distributional conditions — deep capital, cheap inputs and aligned incentives — that act on every firm at once and lift the whole curve. The concierge operates on the wrong unit, at the wrong end, with the wrong kind of lever. Shifting a distribution is something policy can do; hitting a chosen point in its tail is not, and no quantity of concierge service changes that arithmetic. The most a government can do is widen the distribution that occasionally throws up a giant, and then wait.
Britain is short of deep, patient, late-stage growth capital. It also lacks a public market worth listing on once a firm has drawn that capital down. The British Business Bank's own figures show that, tracking a matched cohort from their first funding round, UK firms raise 2.6 times less than their American peers by the fifth round.¹⁴ Nobody can expedite their way to a capital market that barely exists.
For proof that the binding constraint is late-stage capital rather than a front desk, look across the Channel, to the most generously backed champion Europe has produced, and to how far even that backing falls short. Through Bpifrance, the French state made itself a sovereign co-investor in Mistral; ASML, Europe's most valuable company, took a strategic stake as lead investor in its Series C.15 This is the equity-stake model from the opening taxonomy, seen at the frontier: the state and a strategic corporate taking direct positions rather than merely clearing friction. No British scaling firm enjoys anything close to that depth of patient, prestigious and partly public capital.
And yet, on the most generous reading of the figures — counting every equity round, its debt, and the further raise mooted in 2026 — Mistral has taken in comfortably under ten billion dollars in total. Its valuation sits in the low tens of billions, around $23 billion if its latest round closes on the reported terms. Over the same span OpenAI has raised on the order of $186 billion and Anthropic around $161 billion.16 Europe's best-funded AI lab, with the deepest state backing on the continent behind it, sits more than twentyfold below the American frontier on the one axis that determines whether a firm can reach giant scale. If that is the ceiling with far deeper state support than Britain is offering, then what limits a British trillion dollar firm is not the absence of a concierge but the depth of the capital available to feed one.
The capital that would close Britain's gap has drained out of the domestic system, largely by Britain's own hand. UK pension funds now hold around 4.4 per cent of their assets in domestic equities, against an international norm closer to ten.17 Defined-contribution schemes that put some forty per cent into British shares a decade ago now hold a small fraction of that. In 1997 British pension funds were roughly three-quarters in equities; today they are barely a third. Britain is among the very few major economies whose own pension system has effectively decided not to back its own market. The exit shows up at the listing end too: in 2024 some 88 companies left the London Stock Exchange by delisting or moving their primary listing, against a bare handful of arrivals, with New York the usual destination.18
A more pointed diagnosis locates Britain's problem in the plumbing that prices and directs capital, rather than in the quantity of capital itself: a pension sector so fragmented that no fund has the scale to take large illiquid positions, the near-collapse of investment research on small and mid-cap firms after the unbundling of research from trading commissions, and transaction costs that deter domestic holding.19 There is force in this, but it does not so much refute the capital diagnosis as complete it. Patient capital does not pool in a market where firms cannot be cheaply researched, cheaply traded or held at scale; thin coverage and a fragmented investor base are part of why the deep late-stage pools never form. "Britain lacks growth capital" and "Britain's capital markets are structurally broken" are the same problem seen from opposite ends, and the reforms that follow are aimed at both.
That diagnosis has also summoned a tempting and wrong-headed remedy, worth explaining before the better ones. If domestic institutions will not hold domestic shares, the argument runs, they should be made to: through a "British ISA," a mandated minimum allocation, or a domestic investment quota tied to the tax relief pensions already enjoy.
This policy is a mistake for many reasons. It misdiagnoses the disease, because the binding problem is not that British firms cannot find buyers for their shares, but that the machinery which prices and channels capital has decayed. Forcing money into that machinery does not repair it. It harms the savers it claims to help, concentrating retirement risk in a single national market for the sake of an industrial policy goal those savers never chose.
A "UK equity" defined by listing would also route most of the money to multinationals with little domestic footprint, while a definition robust enough to capture actual domestic activity would exclude much of the FTSE and force savings into an artificially narrow, illiquid pool. It operates at a scale too small to matter, since the sums involved are dwarfed by institutional flows, and most savers do not use even their existing allowances.²⁰ The lesson of Mistral is that Britain needs deeper capital — not captive capital – and the two are not the same thing.
A government serious about the tail would therefore tilt its political capital heavily away from the desk and towards the maze. The reforms that widen the distribution are well-rehearsed and unfashionably dull, and they are supply-side in the oldest sense: a matter of removing obstacles the state has placed in its own path rather than bolting a new scheme on top. The case for each is not that novel, but it bears directly on the single binding constraint Mistral shows: how much capital a British firm can actually draw on, and how well the market around it works.
Start with the tax on owning British shares. The 0.5 per cent stamp duty on UK equities is charged where New York charges nothing. It is, baldly, a tax on the domestic ownership of domestic firms, levied at the exact point where the goal is to deepen the home market. It is hard to think of a worse thing to tax. Then fix the tax treatment of employee equity, so that founders and early staff in Britain are not penalised against their Californian counterparts for being paid in the upside they help to create: the mechanism by which talent compounds into ownership, and ownership into the next company's capital.
We should also restore scale and risk appetite to the pension system, ideally through consolidation into larger funds able to hold illiquid growth assets. As argued above, that is a far sounder route than the crude mandation that would force savers into concentrated domestic bets they never chose. The same logic recommends reviving the research coverage that lets investors evaluate smaller listed firms in the first place, without which any capital directed their way is flying blind. And as for the deep-tech and life-sciences firms most likely to throw up a giant, address the physical cost base directly: the planning regime that keeps laboratory and data-centre space scarce, and the industrial energy prices that sit among the highest in the developed world.
None of these is a scheme to be launched or a ribbon to be cut, which is why they are harder to do, and likelier to last.
Conclusion
The pattern across the cases is consistent. Facilitation works, modestly, conditionally, and only when the institution running it is left alone long enough to accumulate a track record. Ireland's IDA did not produce a trillion dollar firm in seventy-six years. It produced 300,000 jobs and a template others have copied, which is a great deal more than nothing and a great deal less than a champion.
No government has ever certified its way to a superstar, either. The firms that sit above a trillion dollars in market value got there through ordinary market competition, fed by deep pools of capital — not through a selection panel, a visa rebate or a named cohort at a tech-week launch. The lone exception only proves the rule. Saudi Aramco reached the trillion dollar mark not because a committee picked it but because it owns a natural resource monopoly. Even then, its valuation was realised the orthodox way, through the largest public offering in history on deep domestic and international markets. Aramco is a double confirmation of the thesis: scale comes from owning something extraordinary and from capital markets able to price it, never from the act of selection. It is, needless to say, not a model anyone proposes to import.
The comparative evidence, then, does not leave Britain empty-handed. It points in one direction. Keep the concierge, which is cheap and has a long record, and run it with discipline: private sourcing, a guaranteed lifespan and a publicly-fixed metric. Resist the drift towards certification, which carries the thinnest evidence and the heaviest capture risk. And do the politically costly work that actually governs whether a giant emerges: abolish stamp duty on shares, fix employee-equity taxation, consolidate the pension system, revive the research coverage that lets domestic capital find its targets, and clear the planning and energy costs that keep Britain's deep-tech firms undersized and contemplating a New York listing.
The firms that do reach the trillion dollar mark — and Britain has plausible candidates in its ecosystem already — will get there because the underlying conditions were finally made right. A concierge can help build those conditions at the edges. The question raised by the Business Secretary's announcement is whether it is being designed to do that quietly and well, or to do something else loudly and badly.
Martha Evans - 29/06/2026
Notes
1 IDA Ireland's employment and investment figures are the agency's own: see its 2025 end-of-year statement, 'Strength and resilience of FDI propels growth, innovation, and competitiveness'.
2 On the distortion of headline Irish output by a handful of multinationals, see John FitzGerald, 'National Accounts for a Global Economy: the Case of Ireland', Quarterly Economic Commentary: Special Article (ESRI, Summer 2018); and the Central Statistics Office (Ireland)'s explainer on modified gross national income (GNI*).
3 Juanita Gonzalez-Uribe and Michael Leatherbee, 'The Effects of Business Accelerators on Venture Performance: Evidence from Start-Up Chile', The Review of Financial Studies 31, no. 4 (2018), pp. 1566–1603.
4 Torfinn Harding and Beata Javorcik, 'Roll Out the Red Carpet and They Will Come: Investment Promotion and FDI Inflows', The Economic Journal 121, no. 557 (2011), pp. 1445–1476.
5 Jacques Morisset and Kelly Andrews-Johnson, 'The Effectiveness of Promotion Agencies at Attracting Foreign Direct Investment', FIAS Occasional Paper 16 (World Bank, 2004).
6 Harding and Javorcik, 'Roll Out the Red Carpet and They Will Come' (n 4).
7 J-Startup is administered by Japan's Ministry of Economy, Trade and Industry (METI). See 'Newly Selected Startups Announced under the Public-Private Initiative for Supporting Startups Called the J-Startup Program' (13 March 2025).
8 George J. Stigler, 'The Theory of Economic Regulation', The Bell Journal of Economics and Management Science 2, no. 1 (1971), pp. 3–21.
9 OECD–Eurostat high-growth-firm definition (ten or more employees, ≥20 per cent average annualised growth over three years). The ~44,600 figure is the government's own, from DBT's Backing Your Business evidence annex, based on ONS high-growth business statistics.
10 On the returns to sector targeting, see note 6 above.
11 Saudi Arabia's Regional Headquarters (RHQ) Programme, administered by the Ministry of Investment (MISA); under controls in force from 1 January 2024, government agencies may not contract with foreign companies lacking an RHQ in the Kingdom (or their related parties), save in narrowly defined exceptions. See DLA Piper, 'Saudi Regional Headquarters Program: an overview of tax and non-tax incentives' (2024).
12 See the National Audit Office's investigation, Supporting investment into the UK (2023). The findings that the Department does not survey investors who decide against the UK, and that the Office for Investment had worked on only around fifty projects, are from the House of Commons Committee of Public Accounts, 'Supporting investment into the UK', HC 996 (Session 2022–23).
13 On the operation and evaluation of Israel's Office of the Chief Scientist (now the Israel Innovation Authority), see Manuel Trajtenberg, 'R&D Policy in Israel: An Overview and Reassessment', NBER Working Paper 7930 (2000). The matched-firm evaluation is Saul Lach, 'Do R&D Subsidies Stimulate or Displace Private R&D? Evidence from Israel', NBER Working Paper 7943 (2000).
14 British Business Bank, Small Business Equity Tracker 2024; the comparison of round-progression rates and round sizes for UK and US companies from the first to the sixth funding round is set out in its fourth chapter.
15 Mistral AI, 'Mistral AI raises €1.7B to accelerate technological progress with AI' (September 2025) and ASML, 'ASML, Mistral AI enter strategic partnership' (9 September 2025).
16 Cumulative totals are PitchBook's and are approximate: as of mid-2026 PitchBook records Anthropic at around $161bn and OpenAI on a comparable order of magnitude; other trackers differ materially. For Mistral, its €1.7bn Series C valued it at €11.7bn (~$14bn) — see Mistral AI; counting all equity and debt it has raised under $4bn, and the ~$23bn figure reflects the roughly €20bn valuation Bloomberg reported (June 2026) for a round then in talks.
17 See New Financial, 'Comparing the Asset Allocation of Global Pension Systems', and also the Pensions and Lifetime Savings Association on the long decline in domestic-equity holdings.
18 EY, 'London stock market ends 2024 on a high after Q4 IPO boost' (January 2025); the underlying data is in EY's IPO Eye Q4 2024.
19 Centre for British Progress, ISA and Pension Reform: Why Forced Investment Is Not Real Investment (2025), by Pedro Serôdio and David Lawrence.
20 Ibid.