Wonga’s beta system processed its first loan in late 2007.

By the time we opened the doors to the market the following summer, Lehman had fallen. The autumn of 2008 was, by every external measure, the worst possible time to be running a consumer-finance company that almost no one had heard of. Banks that were too big to fail did exactly that. Capital evaporated. The European venture ecosystem virtually shut down. Trust in traditional finance collapsed in the span of a single news cycle. Most of the founders I knew at the time spent those months not building, but defending — cutting headcount, pulling marketing, calling investors who were not picking up, and asking whether they would still be alive in March.

We did the opposite. We built faster.

Wonga was founded on a single insight: that by 2005 there was enough data and alternative information available that machines could make less prejudiced, more consistent, and more appropriate credit decisions than humans for many financial purposes. Credit decisioning was still dominated by bank managers whose intuition, however well-meaning, was implicitly biased and structurally slow. The thesis was unfashionable. The product was unproven. The crisis arrived a year into the build.

What followed shaped the company for the next decade. It also shaped how I think about backing founders today, fifteen years and seventy-five investments later. The conditions of 2026 are not the conditions of 2008, but they rhyme. Capital is harder to raise than it was in 2021. Trust in several categories of incumbent has thinned. The labour market has tilted, in places, towards the patient operator who is willing to wait out the noise. The founders who recognise the pattern are the ones who will compound through the next decade.

There were three lessons from those eighteen months. None of them were theoretical at the time. Each of them was the difference between a company that survived 2008 and a company that did not.

The first lesson was about discipline

Capital scarcity is, in retrospect, the best operating constraint a young company can have.

When venture money is plentiful, the temptation is to outspend the problem. You hire ahead of demand because the next round will absorb the run-rate. You buy growth because growth is what your investors are pattern-matching to. You build optionality on optionality on optionality, because the cost of an extra hire feels small relative to the cost of being caught flat-footed when the market turns. The discipline is theoretical. The capital is abundant. The two cancel each other out.

When venture money disappears, the calculation inverts. You are forced to build the unit economics that actually work, and to build them now. There is no later. There is no Series B to fix what the seed round let slip. The runway is the runway, and the company either makes the maths work inside it or it does not.

Wonga grew profitably through the GFC because it had to. By 2012, four years after market launch, the company was originating loans, decisioning them in seconds, collecting them at scale, and producing revenue of around £309 million and net profit in the region of £62 million. Those numbers came from a cost structure set in 2008 by a team that had no alternative but to make it efficient. The figures later moved in both directions, but the shape was permanent.

The pattern is not specific to fintech. Most of the great companies of the last thirty years were founded inside, or just after, a crisis. Microsoft, Cisco, Amazon, Google, Salesforce, Atlassian, Stripe, Airbnb — every one of them learned to be a business before they were allowed to be expensive. The founders who built them did not have the luxury of imitating the spending patterns of the previous cohort. Their cohort was on fire. Their constraints became their architecture.

The companies that did not learn that lesson — that scaled their cost base on borrowed conviction during the easy years — are the same companies whose names appear in this decade’s down-round and recapitalisation announcements. Discipline learned late is not the same as discipline learned early. The first kind is a survival reflex. The second kind is a competitive advantage.

The second lesson was about trust

In good times, customer trust accrues to incumbents. The bank is the bank. The carrier is the carrier. The insurer is the insurer. The brand does the work because the consequences of switching feel uncertain and the alternatives feel unproven.

In a crisis, customer trust transfers — and it transfers fastest in the categories where the incumbents are most exposed. The banks were the most exposed institutions on the planet in late 2008. Customers who had spent a decade trusting that their deposits were safe and their borrowing decisions were considered watched both assumptions break inside a single quarter. The willingness to try an alternative — any alternative — rose with the front-page headlines.

Wonga in that moment offered something the banks did not. A decision in seconds, on data the customer could see. An interest rate calculated up front and stated in pounds and pence rather than buried in an APR that no one could parse. A repayment schedule that could be modelled before the loan was taken, not discovered after the first surprise charge. Customers responded with their wallets and their referrals. The product had been built for a generation of people who had been quietly excluded from the formal financial system by institutions that did not trust them. That generation came back, and then they told their friends.

The lesson generalises. When trust in incumbents thins, the customer is not looking for a marginally better version of the incumbent. They are looking for relief from it. Companies that win in that moment are not selling a product. They are offering a defection. The product is the vehicle, but the trust transfer is the event.

I watched the same pattern play out at Tide, where I was an early backer. Tide launched in January 2017 as a digital current account for UK small businesses, a year before PSD2 came fully into force. The traditional high-street banks had not lost public trust in the way that consumer banks had in 2008, but they had eroded it among small-business owners who were tired of waiting six weeks to open an account and being charged for the privilege. Tide did not need to be perfect. It needed to be a relief. Hundreds of thousands of small businesses chose it on those terms.

The third lesson was about talent

The labour market for engineers and data scientists in late 2008 was a buyer’s market for the first time in a decade.

People we would never have been able to recruit eighteen months earlier — credit-risk specialists from the institutions that were unwinding, data scientists from quant funds that had blown up, platform engineers from incumbent banks whose tech budgets had been frozen — were available, sometimes for the first time in their careers, to consider a company whose offices fit on a single floor.

The people we hired in those eighteen months were the foundation of one of the first fully-automated real-time consumer-credit decision engines in operation anywhere. They were available because the firms that would normally have hired them had stopped hiring. They were motivated because they wanted to work on something that solved a problem rather than something that defended a balance sheet. They were patient because the alternative was a market in which patience was the only currency that still had value.

The same is happening now, quietly, in 2026. The senior engineers laid off in 2022 and 2023 are no longer the same engineers who were available in 2024. Many have spent the intervening years either consulting, founding, or sitting on the bench long enough to be selective. The compensation arbitrage has narrowed at the top, but the willingness arbitrage has widened. The senior operator who would not have taken your meeting in 2021 will take it now, and will ask sharper questions when they do.

Talent is the most durable of the three advantages, because the hires you make in a downturn compound for years. You inherit the institutional memory of the firms that did not survive. You hire people who watched what broke. You build a team that has lived through what most of your competitors have only read about.

The composite advantage

Discipline, trust, talent. Each of these on its own would be a meaningful structural advantage. Together, they are the reason a crisis is not a problem to be survived. It is a window to be walked through.

The architecture of a company built through scarcity is different from the architecture of a company built through abundance. The unit economics are tighter. The customer relationships are deeper. The team is heavier on senior operators and lighter on the kind of growth-hacking specialists who flourish when capital is cheap. The decision-making is colder. The product is more deliberate.

I think about this often when I write cheques today. The conditions for building have rarely felt easy in the last fifteen years. There has been a financial crisis, a sovereign-debt crisis, a pandemic, a war in Europe, a return of inflation, a collapse in venture funding, and a reordering of the global trade system within the same career. Founders who wait for conditions to be benign tend to wait too long. The window closes while they are still preparing for it to open.

The dust road is most useful when it is the only road open. Builders who know how to read the road in bad weather are the ones who arrive first when the weather clears. They have done the work the others were saving for later. The others have spent the same period accumulating overhead that the road will not support when conditions worsen again.

Building when the world is burning is not about being brave. It is about recognising that the structural advantages — capital discipline, trust transfer, talent availability — are precisely the advantages the incumbents lose first when conditions turn. The advantage is not despite the crisis. It is because of it.

The companies I have backed since Wonga, almost without exception, were started in conditions that the market called wrong. I wrote one of the first cheques into Wise in 2011, when it was a pre-revenue currency-transfer experiment. I backed Cazoo at the deck stage. Tide had not yet been founded. Bitcoin I have publicly defended as a censor-resistant store of value through a decade in which the rest of the financial system was treating it first as a curiosity, then as a threat. Each of those moments looked, to most of the people I respected, like the wrong moment to commit. Each of them was the only moment that mattered.

The road only looks obvious in the rear-view mirror. The work is reading it in real time, while the dust is still up and the consensus has not yet arrived.

Questions

Why did Wonga survive the 2008 financial crisis when so many lenders did not?
Three reasons, in order of importance: discipline, trust, and talent. Capital scarcity forced the company to build unit economics that worked from day one rather than outspending the problem. Customer trust in banks collapsed at the moment a transparent, data-driven alternative could be offered. And the labour market for credit-risk specialists and data scientists became a buyer's market for the first time in a decade. Each of those was a structural advantage the crisis itself created.
Why is capital scarcity an advantage rather than a constraint?
Capital scarcity forces the founder to make the unit economics work before the company grows. The companies that survive downturns learned to be profitable while small; the companies that did not learned to be expensive while small. That habit, once formed, persists through the upturn and becomes a durable competitive advantage. Most of the great companies of the last thirty years were founded inside or just after a crisis.
What should founders raising capital in 2026 take from this?
The same conditions are present now. Venture funding has contracted, several categories of incumbent have lost the trust of their customers, and the labour market has tilted in favour of the patient operator. Founders waiting for conditions to be benign tend to wait too long. The ones who read the road in bad weather are the ones who arrive first when the weather clears.
Where does the dust-road thesis come in?
The dust road is most useful when it is the only road open. When the paved alternatives are blocked — when incumbents have lost the trust of their customers, when capital has dried up for the easy bets, when the talent that used to be unaffordable becomes available — the unpaved route becomes the fastest. A crisis does not invent dust roads. It reveals which ones are passable.

Read next

Share LinkedIn → X / Twitter → RSS →