In July, the headline was hard to miss. UK startups and scaleups raised $17bn in the first half of 2026 — up 102% on the same period last year, and the strongest opening to a year since 2022.

‍

If you are a founder three months into a seed raise with a diary full of polite maybes, you probably read that and thought: then why does this feel like the worst market I have ever raised in?

‍

Because that number is not describing your market. It is describing a handful of companies you will never compete with for capital.

‍

The headline and the reality are two different data sets

‍

Of that $17bn, AI companies raised $12.6bn — nearly three quarters of all venture capital deployed in the UK. They took 19 of the country's 28 megarounds above $100m, and all four of the rounds above $1bn. Late-stage rounds accounted for 68% of every pound raised, up from 42% a year earlier.

‍

That is the top of the market. Now look at the bottom of it.

‍

The British Business Bank's Small Business Equity Tracker, published in July, measures equity going into UK smaller businesses. It fell 4% to £12.3bn in 2025. The top ten fundraisings alone made up 23% of all investment — the highest concentration since 2020.

‍

And the figure that should actually change your plan: seed-stage deals were 27% lower in 2025. Venture-stage deals were down 13%.

‍

Record capital going in. Fewer early-stage cheques coming out. Both true, at the same time.

‍

Why founders keep planning against the wrong number

‍

Market averages are a comfort blanket. They let you write a plan that says we will raise in month nine, because that is roughly what the market seems to be doing, and everyone in the room nods.

‍

But funding markets are not averages. They are queues. And in 2026 the queue for a non-AI seed round is longer, slower and stricter than the aggregate suggests. When capital concentrates into fewer, larger deals, the money does not trickle down — it just goes somewhere else.

‍

The founders who get burned by this are rarely the ones with a bad business. They are the ones whose model assumed a market that only exists in a press release.

‍

Three numbers worth rebuilding this month

‍

1. Your runway floor, not your runway target. Most models show runway assuming the raise lands. Build the version where it does not. What is the last date you can still make a controlled decision rather than a panicked one? That date, not your target close date, is the one that should be on the board pack.

‍

2. Your default-alive date. At current growth and current burn, when do you cover your own costs? If the honest answer is never, at this cost base, that is not a failure — it is information, and it is far cheaper to act on now than in month eight of a raise.

‍

3. Your evidence, not your narrative. In a concentrated market, investors have more optionality and more time. That means more diligence, and more of it aimed at your numbers. Cohort retention that reconciles to the ledger. Gross margin that survives a question about hosting costs. A revenue recognition policy you can explain in one sentence.

‍

The regional footnote nobody mentioned

‍

Buried in the same Equity Tracker: London's share of UK equity investment fell from 60% to 57%, while the North West rose 82%, the South West 104% and Scotland 74%. Concentrated at the top, yes — but also more geographically spread than it was.

‍

If you are outside the M25 and have been told the money is all in London, that is now measurably less true than it was two years ago. Worth knowing before you spend six months building a London-only investor list.

‍

The uncomfortable part

‍

A 27% fall in seed deals does not mean seed is closed. It means the bar moved, and it moved quietly, while the headlines were busy celebrating.

‍

The businesses still getting funded at seed are not the ones with the best deck. They are the ones who can answer a hard question about their unit economics in under a minute, because someone competent has already asked them that question internally.

‍

That is the whole job. Not fundraising theatre — knowing your numbers well enough that a difficult market cannot surprise you with them.

‍

Not sure whether your model survives a market like this one? Take a free hour with us and we will pressure-test your runway and your assumptions — no pitch, no boring numbers. Finance leadership, without the headcount.

‍

Sources: UKTN, "Record AI investment drives strongest UK startup funding since 2022" (analysis by HSBC Innovation Banking UK and Dealroom, July 2026), uktech.news ; British Business Bank, "AI dominates UK smaller business equity market with record investment share as overall funding falls slightly" (Small Business Equity Tracker, 2 July 2026), british-business-bank.co.uk