Why everything feels a bit flatter in 2026, and what that means for risk-taking and investing in innovation

Innovation
Written By Sharp Insight

By Dr Chloe Sharp, Sharp Insights

Something about the last few years feels flatter. If I had to describe life in 2026, I'd say we're starting to see a kind of cultural convergence. AI-generated language repeats the same phrases and structures. Brands and websites increasingly look and sound alike. Recommendation algorithms push us towards more of what already performs, like certain designs and phrases.

Zoom out and look at UK culture, which feels like the recent past is being recycled back at us, or feels like cost-cutting exercises with little inspiration. For example, new houses and schools being built in my local area are boxy and minimalist. The cost of living and the cost of running a business both keep climbing. And the digital products we depend on seem to get more expensive and less good at the same time.

Every problem arrives already broken into individual pieces, automated and put behind a subscription. Sometimes creating more work to unpick it, get to a human to speak to or make sense of it. This makes me feel more like a machine or data point than a human trying to experience life.

It is tempting to call all of this decline. But that flattens several different things into one. They have different causes, different quality of evidence behind them and quite different implications for anyone trying to build something new in the UK.

Three declines: Cultural, Economics and Enshittification

The first is cultural repetition. Mark Fisher got there well before it became a common complaint, with his idea of hauntology: culture haunted by futures it can no longer imagine, so it loops the past instead. Kyle Chayka's more recent version is about mechanism rather than mood. Algorithmic recommendation pushes taste towards whatever performs, helping explain why the coffee shops, Airbnbs, interfaces and fonts increasingly converge whether you are in Nairobi or Northampton. Whether this is actually decline, rather than simply a flat stretch that will eventually acquire its own nostalgia, is genuinely open. Pop culture has had weak periods before and recovered.

The second is economic, and it has much harder data behind it. UK real wages stagnated for the longest period since consistent records began, with no growth on average between 2008 and 2023. Against OECD comparators, the UK fell from fifth or sixth place for wage growth in the pre-crisis years to sixteenth for 2008 to 2022. Underneath that sits an investment problem that predates the financial crisis. Fixed investment has averaged around three percentage points below the G7 average for four decades. IPPR's most recent analysis puts UK business investment second lowest in the G7 at 11.1 per cent of GDP, with British workers having roughly 38 per cent less capital to work with than their peers, rising to 47 per cent in manufacturing.

The third is what Cory Doctorow named enshittification. A platform starts by being good to users to acquire them. It then degrades that experience to serve business customers, before eventually degrading the experience for both in order to extract more value for itself. It is what can happen to a two-sided market once lock-in substantially reduces the cost of treating either side badly. The word has now been named word of the year on three continents, which tells you something about how widely the experience resonates.

These are different phenomena, but they produce a similar effect: the space for alternatives gets narrower. Systems become better at optimising and extracting value from what already exists than creating the conditions for something different to emerge.

What this means for innovation

When capital is scarce, and costs are rising, experimentation becomes harder to justify. Everyone making a decision has a reason to choose the legible, safe, risk-averse option. In culture, that might be the reboot of a recognisable franchise. In innovation, it can be the proposal that resembles last year's winning proposal, the technology with a familiar business model or the company that already looks enough like previous successes to feel safe.

And alongside this, the numbers show a substantial decline in business innovation activity. The UK Innovation Survey published in June 2026 found that 34 per cent of UK businesses were innovation active in 2022 to 2024, down from 36 per cent in the previous period and from 53 per cent a decade earlier. That headline nineteen-percentage-point fall deserves handling carefully, because the series has not moved in a straight line.

The proportion dropped from 38 per cent in 2016 to 2018, rose to 45 per cent in 2018 to 2020, and has fallen in each of the two periods since. The rise is worth pausing on, because that period covers the first year of the pandemic, when a great many businesses changed how they operated at speed, and rapid process change is the kind of thing the survey counts. Some of the subsequent fall is therefore that unusual period unwinding. But not all of it, because the current figure is four points below the pre-pandemic low. The definitional change runs in the same direction shown by UKIS 2023, which widened what counts as business process innovation, which the department notes makes it not directly comparable with earlier surveys, and a wider definition should push recent figures up rather than down.

That creates an interesting contradiction. The appetite to start businesses is extremely high. The proportion of businesses actively innovating has fallen substantially.

There is also a more immediate context. The global economy in 2026 is being pulled in two different directions. The IMF describes an energy shock from conflict in the Middle East on one side and a technology-driven investment boom, particularly around AI, on the other. Its July forecast puts global growth at 3 per cent this year, while the World Bank's June forecast is lower at 2.5 per cent. Both point towards a weaker and unusually uneven environment rather than a conventional global recession.

What matters for innovation is where that growth and investment are occurring. The IMF finds that the benefits of the current technology cycle are concentrated among economies plugged into AI hardware and technology supply chains, while energy-importing economies outside that boom experience more of the downside. The UK faces an awkward version of this problem: NIESR forecasts growth of just 1.1 per cent in 2026 and expects business investment to fall, while energy costs remain elevated and Bank Rate sits at 3.75 per cent. That is not an easy environment in which to finance expensive experimentation.

Where the money went

Taken together, these create a particular tension with the UK's Industrial Strategy. Many of its priority sectors: advanced manufacturing, clean energy, defence, life sciences and frontier technologies, require precisely the things that have become more difficult: patient capital, expensive R&D, specialist infrastructure and room for technical uncertainty. These businesses cannot always iterate cheaply towards product-market fit. In hard tech and deep tech, the experiment might be a manufacturing line, a clinical study, a new material or a physical pilot.

Three specific constraints explain at least part of the environment in which that is happening, and they fall particularly heavily on the deep tech and manufacturing sectors the UK says it wants to grow.

Energy is the most obvious: In 2023 the UK had the highest industrial electricity prices of the twenty-four IEA member countries, around 50 per cent above France and Germany and four times those of the US. Prices have eased from the 2022 peak but remain roughly double pre-crisis levels. IPPR points out the trap directly: modern manufacturing competitiveness depends partly on greater automation and digitisation. Both require electricity. High power prices therefore act as a brake on some of the investment that could help close the productivity gap.

R&D tax relief moved against small claimants: From April 2023, the SME enhancement fell from 130 per cent to 86 per cent and the payable credit rate from 14.5 per cent to 10 per cent. For firms outside the R&D-intensive carve-out, support worth as much as 33.4p per pound of qualifying spend previously fell below 19p. The subsequent statistics are striking. Total relief claimed for 2023 to 2024 fell only 2 per cent, but the number of claims fell 26 per cent, driven overwhelmingly by smaller claims disappearing. Claims worth £1 million or more went from accounting for 43 per cent of all relief to 54 per cent in two years. Whatever the intention of the reforms, the system has become significantly less attractive to smaller claimants.

Grant funding shows a related pattern: The final Innovate UK Smart Grants round before the scheme was paused in January 2025 received 2,134 applications for a pot cut from £25 million to £15 million. The resulting success rate was around 2 per cent, with applicants needing scores above 87 per cent to receive funding. Smart Grants had been the major broad, open, sector-agnostic competition available to SMEs. Meanwhile, a Freedom of Information request showed that the Innovation Loans programme, towards which many businesses were redirected, held a £100 million budget across four 2025 competitions but awarded £28.3 million. That leaves £71.7 million unallocated.

The process grows as the investment shrinks

There is another consequence of making innovation capital scarce: competing for it becomes an activity in its own right. Grant writing becomes a profession. Accelerators multiply. Advisory firms emerge on both sides of the R&D tax relief system, some helping businesses make claims and others helping them survive the resulting scrutiny. None of those services is inherently unnecessary. Many exist because navigating these systems genuinely requires expertise. But something strange happens when the transaction cost of accessing innovation support keeps rising. A Smart Grants success rate of around 2 per cent means roughly 2,090 teams in a single round spent time developing applications that were not funded.

Some of that is unavoidable in any competitive allocation system.

But collectively it represents an enormous amount of founder, scientist, engineer and adviser time spent competing for permission to do the work rather than doing the work itself. And the burden is not evenly distributed. Organisations with experienced teams, professional advisers and enough cash to absorb unsuccessful applications are better equipped to keep playing. There is a related problem in the kinds of innovation we fund.

Collective problems can end up being addressed through individual consumer products because those are easier to package into a recognisable business model. We get an app that helps someone cope with a broken system rather than an intervention that repairs the system itself. And if British firms increasingly build applications on infrastructure owned elsewhere, much of the long-term value accrues to the layer beneath them.

That begins to look less like owning an innovation economy and more like renting one, however busy the ecosystem above it appears.

So is Britain actually in decline?

Declinism is itself a very British genre, roughly a century old and reliably bad at predicting timing. Correlli Barnett was writing about British industrial failure in the 1980s. People were making similar arguments long before him.

There is also substantial counter-evidence.

ARIA exists to fund work considered too speculative for conventional mechanisms and held £514.1 million in active R&D funding agreements at March 2026. Its lower-cost seed rounds are experimenting with forms of funding that traditional programmes struggle to provide. The UK retains major strengths in research, life sciences, fintech and other technology sectors.

And HMRC's compliance work reduced estimated error and fraud in R&D relief from 17.6 per cent to around 6 per cent. That is a real achievement, even if some of the cost of achieving it has fallen on legitimate small claimants. So I don't think the interesting argument is simply that Britain is declining.

It is narrower than that.

Several pressures are compounding at once: stagnant investment, higher operating costs and systems increasingly designed to manage scarcity. The response has been an innovation environment that is more selective, more concentrated and, for many smaller businesses, more expensive to enter. The nineteen-percentage-point fall in innovation-active firms is one indication of what that environment looks like in practice.

We haven't run out of ideas. We've created economic, technological and institutional systems that increasingly reward legibility (how easily a decision-maker can recognise something as a familiar, fundable, investable proposition), optimisation and low-risk repetition over experimentation. Britain says it wants to build an economy around difficult innovation, while making difficult innovation increasingly expensive to attempt.

So that leaves us with the question: If the UK's problem is not a shortage of ambition, but a shortage of room to take risks, what would an innovation system designed to make experimentation easier actually look like?

References‍ ‍

Sharp Insight https://www.sharp-insight.co.uk
Previous

How to Commercialise Deep Tech: From Technology to Market
Next

What evidence and when: Commercialising complex innovation across software, hardware and hybrid