International expansion is often presented as a structured, research-led process. Founders are encouraged to analyse markets, validate demand, and build carefully phased expansion plans before committing capital.

But the reality uncovered in Breaking New Ground: 2026, developed in partnership with HSBC Innovation Banking, was very different. Across 50 interviews with founders, CROs, operators and expansion specialists, one finding stood out: 68% of founders conducted minimal or no structured market research before entering a new market.

At first glance, that sounds surprising. Yet the more conversations we had, the clearer the rationale became. Founders rarely avoid research because they do not care. More often, they were balancing speed, investor pressure, customer pull, and confidence in their existing GTM model.

One of the clearest themes in the research was the pressure founders feel to act decisively once international demand begins to appear.

For earlier-stage technology companies, expansion is often tied directly to growth expectations. Investor pressure, competitive positioning, and limited runway all create urgency. In that environment, extensive market research can feel less like prudent planning and more like delay.

Some founders simply believed the cost of researching exceeded the cost of learning in-market. Others were already responding to inbound customer demand or partnership opportunities that made expansion feel inevitable.

This explains why “sell first, formalise later” emerged as one of the dominant behaviours across the report. Many founders validated demand remotely before investing in local infrastructure, hiring, or operational setup.

In some cases, this worked well. Founders gained early commercial signals without committing significant capital too early. But the research also showed the risks of entering a market with assumptions that had never been properly tested.

A recurring pattern throughout the interviews was confidence bias.

Founders frequently assumed that if their product solved a problem at home, it would solve the same problem internationally in the same way. They assumed their ICP would transfer. They assumed their sales motion would transfer. They assumed buyers would behave similarly.

In practice, that was rarely the case.

The report found that home-market playbooks often need to be rebuilt rather than translated. Messaging that resonated in the US or Australasia frequently failed to connect with UK buyers. Pricing models that worked elsewhere created friction in Europe. Enterprise sales cycles took significantly longer than founders expected.

Importantly, these were rarely market failures. Companies generally chose sensible markets. The problem was expectation failure.

Founders underestimated how long it takes to build credibility from zero, particularly in enterprise environments where local references, buyer trust, and visible commitment matter heavily. Enterprise sales cycles in new markets consistently stretched to 12–24 months, often double or triple founder expectations.

When those timelines slipped, pressure intensified. Boards became impatient, resources tightened, and companies started making reactive decisions simply to demonstrate momentum.

Another emerging theme from the research was the growing role of AI-generated research in shaping expansion decisions.

Founders increasingly use tools such as ChatGPT, Claude, and Perplexity to assemble competitor analysis, ICP frameworks, and market summaries within hours rather than weeks.

Used well, these tools can provide a valuable starting point. The challenge is when AI-generated outputs are treated as complete answers rather than inputs into a wider validation process.

Several experienced operators interviewed for the report expressed concern that founders were gaining confidence faster than they were gaining real market understanding. AI can summarise information quickly, but it cannot independently validate whether buyers will actually move, how procurement works in practice, or what local credibility signals matter most in-market.

The founders who performed best combined AI-enabled research with direct market exposure: customer conversations, local operators, experienced advisors, and structured feedback loops that challenged assumptions rather than reinforcing them.

The issue is not the use of AI itself. It is relying on AI without enough human judgement, local context, and commercial validation around it.

Despite the challenges, the report also revealed consistent patterns among companies that gained traction internationally.

Successful founders treated expansion as an execution challenge rather than simply a market-selection exercise. They validated demand before committing significant infrastructure. They prioritised securing local lighthouse customers early. They invested heavily in credibility-building and founder presence within the market.

Most importantly, they aligned their organisations around realistic timelines.
The companies that performed best did not assume rapid revenue conversion. They planned for slow sales cycles, budgeted for iteration, and accepted that expansion requires rebuilding trust, relationships, and reputation from the ground up.

That mindset shift proved critical.

Because ultimately, one of the clearest findings from Breaking New Ground was this: international expansion failures are rarely caused by choosing the wrong market. More often, they happen because companies underestimate what it takes to win in the right one.

The evidence suggests founders do not need perfect market research before expanding, but they do need enough validation to challenge assumptions before significant capital is committed.