There is more capital available for technology today than the market picture from two or three years ago would suggest. The problem lies elsewhere: money is flowing at a much slower rate. Investors are funding growth, but they are much more selective in determining which growth deserves a high valuation.
In the United States, VC funds invested $320 billion in 2025, 51 per cent more than the previous year. However, artificial intelligence accounted for 65.4 per cent of the value of all transactions. Five companies — OpenAI, CoreWeave, xAI, Anthropic and Databricks — raised nearly $60 billion. Funding rounds exceeding $100 million accounted for just 3.2 per cent of the total number of transactions, yet absorbed 67 per cent of the total capital.
The first half of 2026 only served to reinforce this imbalance. US start-ups raised over $400 billion, more than in the whole of 2025, but the NVCA and PitchBook emphasise that the record is driven primarily by AI and mega-deals.
Europe presents a similar picture, albeit on a smaller scale. In the second quarter of 2026, VC investment totalled $25.6 billion across 1,636 deals. Among the largest funding rounds were $2.1 billion for Isomorphic Labs, $1.3 billion for Wayve and $1.1 billion for Ineffable Intelligence. All are linked to AI or technologies requiring very large capital outlays.
The market has therefore not returned to the dynamics seen in 2021. Rather, two capital markets have emerged: one for a small group of companies operating in the most sought-after technology segments, and another for all the rest.
Valuations must stand up to the test of cash
In the era of zero interest rates, it was possible to finance expansion for a long time, treating profitability as an issue for a later stage of development. Today, the opportunity cost of capital is a reality. In July 2026, the European Central Bank kept the deposit rate at 2.25 per cent, having raised it in June.
This change is reflected in how software is valued. For private B2B SaaS companies, the most important valuation metrics currently include ARR growth rate and net revenue retention – that is, the ability to maintain and grow revenue from existing customers. The median growth rate in the survey of private SaaS companies stood at 22 per cent for 2025, compared with 25 per cent the previous year.
The popular ‘Rule of 40’ is not a market law, but it accurately reflects the underlying mathematics: the sum of growth and free cash flow margin should hover around 40 per cent. McKinsey’s analyses show that as a company matures, the importance of growth rate alone diminishes, whilst an increasing proportion of value is generated by the FCF margin, customer retention and sales efficiency. For mature software producers, a healthy free cash flow margin is usually at least 10–15 per cent.
Airtable: a $11.7 billion valuation was not the same as a $11.7 billion valuation
One of the most telling examples of this shift is Airtable. At the height of the tech boom in 2021, the company was valued at $11.7 billion. In August 2026 , Bending Spoons agreed to acquire it at an enterprise value of approximately $1.29 billion. Airtable generates around $480 million in annual recurring revenue, which equates to a multiple of approximately 2.7 times ARR.
This is a more significant signal than yet another record-breaking AI funding round. It shows what happens to a valuation when a company, funded according to 2021 standards, is faced with the price that a real buyer is prepared to pay a few years later. A high ARR was not enough to sustain the historical valuation.
The problem affects a much larger part of the market. According to the NVCA, there are 859 private unicorns operating in the US with a combined valuation of $4.34 trillion. At the same time, the organisation estimates that only around 5 per cent meet the criteria for sufficient revenue scale and efficiency to meet current public market standards.
AI cannot tolerate capital discipline
The paradox of the current market is that companies are spending record amounts on technology at precisely the time when it is becoming increasingly difficult to accept a project without a measurable impact. Gartner forecasts global IT spending in 2026 at $6.31 trillion, up 13.5 per cent year-on-year. AI-related spending alone is set to rise by 47 per cent, to $2.59 trillion.
This increase in budgets does not signal a return to blanket investment in ‘transformation’. In a Gartner survey, 31 per cent of sales directors cited the difficulty in demonstrating a return on investment from AI tools as one of the main barriers to achieving their 2026 targets.
It is therefore not the appetite for technology that has changed, but the standard of proof required. Capital continues to reward speed, scale and risk. However, it is far less likely to pay for the mere promise that business viability will follow later. Following a period of cheap growth, the value of technology is increasingly being measured where the investor pitch ends: in retention, margins, cash flow and the price that someone is ultimately actually willing to pay.
