Read the business data published across July and early August in isolation and the UK looks in reasonable shape. Advertising investment grew 9.3 per cent. Venture capital doubled. A major bank posted £4.3 billion of half-year profit. Manufacturing output grew at its fastest rate in almost two years. Buyers paid premiums above 70 per cent to acquire listed companies.
Read the small business data from the same weeks and the country looks quite different. Confidence reached the weakest level on record, 55 per cent of firms reported falling revenues, and more expected to shrink, sell up or close than expected to grow.
Neither set of figures is wrong. What they have in common is more interesting than the contradiction, because in every single case the strength appears at the large end of the distribution.
Five Datasets, One Pattern
Start with advertising. UK investment reached £11.7 billion in the first quarter, up 9.3 per cent. But the growth was concentrated in retail media, up 17.9 per cent, social media at 17.7 per cent and search, which alone took £4.6 billion of the total. Those are formats bought continuously at scale. Total television, the format most accessible to a mid-sized advertiser buying a campaign, grew 0.8 per cent.
Venture capital next. UK startups raised €14.8 billion in the first half, a 102 per cent increase. Yet late-stage companies took 68 per cent of it, against 42 per cent a year earlier, and the number of completed rounds fell. Four AI companies accounted for around €5.5 billion between them. The doubling was real and narrowly distributed.
Then manufacturing. July output grew at its fastest pace in almost two years, across all three product categories. The same release records a mild downturn in production volumes at small-scale manufacturers, contrasting with growth at medium and large firms. The sector accelerated and its smallest members did not.
The Same Shape in Finance and Ownership
Banking shows it too. Lloyds reported a banking net interest margin of 3.19 per cent, widened by 15 basis points, driven substantially by structural hedge income. A structural hedge is available to an institution with a large balance sheet and a treasury function. It is a mechanism by which size itself generates return, independent of what the Bank of England does.
Ownership completes the set. Four UK listed companies agreed takeovers inside a week, at premiums of 73, 49 and 41 per cent. Peel Hunt counts 154 bids for UK companies worth more than £100 million since the start of 2023, totalling £165 billion. The buyers were overseas trade acquirers and private equity. The sellers were mid-sized British companies whose listed valuations had fallen far enough to make the gap worth paying for.
Why the Aggregate Misleads
None of this is a claim that the data is being presented dishonestly. Every figure quoted above is accurately reported by its source, and several sources note the distributional point themselves.
The problem is structural. Aggregate measures are sums, and a sum is dominated by its largest components. When four companies raise €5.5 billion of a €14.8 billion national total, the national total is substantially a description of those four companies. When search takes £4.6 billion of £11.7 billion of advertising, the advertising market’s growth rate is substantially a description of search.
A business reading those headlines to calibrate its own expectations is comparing itself to a number generated by firms operating at a completely different scale. The reasonable conclusion, that conditions are improving and one’s own performance is therefore disappointing, does not follow from the data.
What Scale Is Actually Buying
The mechanisms differ by sector but they rhyme.
In advertising, scale buys continuous presence in auction-based channels, where sustained bidding is cheaper per unit than intermittent campaigns. In venture capital, scale means demonstrated revenue, which is what late-stage investors are paying for when they take 68 per cent of the money. In banking, scale means a balance sheet large enough to run a hedge. In manufacturing, scale means the working capital to hold inventory through volatile lead times, which is precisely what smaller producers could not do in a month when the whole sector ran stocks down.
In each case the advantage is not cleverness. It is a structural feature of how the market works that only becomes available above a certain size.
Where the Small-Firm Numbers Sit
Against that, the Federation of Small Businesses recorded a net confidence balance of minus 14 per cent, the weakest in the history of its index. Eighteen per cent of small firms expected to grow. Thirty-two per cent expected to shrink, sell up or close. Eighty-nine per cent reported higher running costs than a year earlier, and taxation was the most cited driver at 58 per cent.
Those figures describe the same country and the same quarter as the ones above. They are also, in population terms, a description of far more businesses. The FSB survey covers the segment that makes up the overwhelming majority of UK firms by number.
The record low is therefore the more representative statistic, and the growth figures are the more economically weighted ones. Both are true measures of different things.
One Piece of Evidence Cuts Across It
There is a genuine counterweight worth stating, because a pattern this tidy invites over-reading.
Credit is available to smaller firms and is expanding. Funding Circle extended £1.7 billion of credit to SMEs in the first half, up from £1.1 billion, while nearly quadrupling its own profit. Lloyds grew its loan book and its risk-weighted assets. Two lenders of different types both expanded supply.
That matters because the most damaging version of a two-speed economy is one where smaller firms cannot access finance at all. On this evidence that is not the constraint. The constraint is demand, cost and the price of that finance rather than its availability.
What Would Have to Change
A pattern this consistent across five unrelated datasets is unlikely to reverse because sentiment improves. The mechanisms producing it are structural, so the things that would narrow the gap are structural too.
The first is the cost of capital. Concentration is partly a rational response to a higher risk-free rate: when money costs more, it flows to the most certain places, which are large and proven. Bank Rate at 3.75 per cent with three of nine committee members preferring an increase is not an environment in which capital spreads out. A sustained fall would change the calculation, and neither the Bank nor the market is currently signalling one.
The second is fixed costs. Business rates, employer costs and energy fall on a small firm as a proportion of a much smaller turnover. The 20 per cent rates relief arriving for 32,000 pubs, clubs and small venues in April 2027 addresses a fraction of one sector, and the FSB’s argument has been that the underlying structure rather than the exemption list is the problem.
The third is demand, which 64 per cent of small firms named as their leading constraint. That one is not amenable to a targeted intervention at all, and it is the reason a business waiting for policy to fix its order book is likely to wait longer than its cash allows.
Anyone forecasting on the assumption that the gap closes soon should be able to say which of those three they expect to move, and why. On the evidence published this summer, none of them has.
How to Use This
For a business owner the practical instruction is narrow: benchmark against firms of your own size, in your own sector, and treat national aggregates as context rather than as comparison.
A manufacturer whose output fell in July was not underperforming a sector that grew, it was experiencing what the survey records as normal for small-scale producers that month. An advertiser whose costs rose was not buying badly, it was bidding in auctions where spend grew 17 per cent. A company that raised nothing this year was not rejected by a market that doubled, it was outside the late-stage category that took two-thirds of the money.
The aggregate is not the peer group. This summer’s data makes that unusually easy to demonstrate.
About The Author
More Stories
Independent Brewers Cannot Get Into 62% of Their Own Local Pubs
Aviva Beat on Every Measure and Its Shares Still Trail the FTSE 100
Late Payments Now Cost £11bn and One in Five Firms Just Writes Them Off