In today’s venture landscape, the noise is deafening. We are witnessing a capital formation cycle that rivals the dot-com era, with hyperscalers committing hundreds of billions to infrastructure and startups achieving extraordinary valuations on promise alone.

But a clear distinction is now emerging: between AI as a novelty, and AI as an indispensable enterprise tool.

Goldman Sachs CEO David Solomon has recently cautioned that while the macro foundations for AI are real, the market itself is becoming frothy. His core concern is not the technology, but the pace of enterprise adoption. Integrating AI meaningfully into large organisations is proving harder and slower than many forecasts suggest. History tells us that when expectations outrun reality, valuation resets tend to follow.

For investors, this marks an important shift. The opportunity is no longer in backing the loudest narratives, but in backing companies with early, demonstrable commercial traction. These are the businesses most likely not only to withstand a correction, but to emerge as highly attractive acquisition targets during the consolidation phase that typically follows periods of excess.

From Floodplain to Workflow

The reality inside enterprises today is messy. Allen Fazio, CIO of Houlihan Lokey, has described an environment where his firm went from tracking 60 AI products to nearly 200 in just a few months. The result is an overwhelming floodplain of disconnected tools.

Most of these products are being experimented with rather than deployed. They are kicked around by curious employees, but they are not embedded, not mission-critical, and not sticky.

The long-term winners are those moving decisively beyond experimentation and into core workflows.

OpenAI’s own usage data reinforces this divide. While casual usage continues to grow, the real economic value is showing up in what it describes as frontier organisations. In these companies, reasoning-token consumption has increased more than 300-fold year-on-year. That is a signal of structured, repeatable, production-grade usage, not ad-hoc prompting.

This is where investors should focus. The most compelling startups are automating what Allen Fazio calls 100-level tasks: high-volume, repetitive work carried out by analysts and associates. When AI removes hours from these workflows, the ROI is immediate, measurable, and easy to justify internally. Adoption accelerates not because the technology is impressive, but because the economics are undeniable.

Why Time-to-Exit Is Compressing

Commercial traction matters not just for resilience, but for liquidity.

At Haatch, we are seeing clear evidence that speed to revenue for UK software companies is compressing. AI enables smaller teams to build, ship, and monetise products faster than in previous cycles. As a result, the traditional gap between startup and scale is narrowing.

At the same time, enterprises are approaching a consolidation cliff. CIOs are increasingly vocal about their desire for orchestration layers that reduce complexity rather than add to it. They do not want employees trained on dozens of tools, nor do they want sensitive data scattered across fragmented systems.

This creates a highly specific M&A environment:

1. Enterprises will cut anything that is ‘nice to have’ but not deeply integrated.

2. Incumbents and well-capitalised platforms will acquire point solutions that already own critical workflows, using them as building blocks for broader orchestration strategies.

In this environment, companies with early revenue and embedded usage are disproportionately valuable. They offer acquirers immediate utility, proven demand, and defensible customer relationships, rather than speculative technology alone.

Why the Haatch EIS Fund Is Well Positioned

This is precisely where Haatch’s EIS strategy is advantaged.

We back B2B software companies at a point where commercial reality already exists. Many businesses in our most recent fund were generating significant revenue at the time of investment, with several operating at multi-million-pound ARR. That is not incidental; it is central to our approach.

By focusing on companies that are already embedded in real workflows, we increase the probability that our portfolio becomes part of the enterprise toolset of the future. As consolidation accelerates, these companies are far more likely to be acquired, and often on shorter timelines than historically expected.

In other words, we are not underwriting adoption risk. We are backing businesses that enterprises are already paying for, relying on, and integrating.

Origination Is the New Alpha

In markets defined by hype cycles, trying to pick winners from a crowded field is statistically difficult. The real advantage lies in origination: consistently accessing companies with genuine traction before they become obvious.

As noted in The Odin Times, published by Odin, one of Haatch’s portfolio companies, improving the quality of the deal pool by identifying businesses with genuine commercial traction delivers far higher returns than attempting to outsmart the market on speculative bets.

The discipline for this cycle is clear. Ignore vanity metrics such as sign-ups and demos. Focus instead on workflow integration, revenue velocity, and customer dependency.

The companies doing real work inside enterprises today are the ones that will be acquired tomorrow. The rest, however impressive they sound, are just froth.