Rebuilding the Ship at Sea

Marcus Wolter

5 minutes ago

5 min read

Why established firms struggle to redesign themselves for AI, and why we started in the dry dock

M&A done the old way, but quicker?

Picture a mid-market sale. The seller hires a bank to run the process and a law firm to paper it. Each adviser has its own team, its own systems and its own way of working, and they meet mainly in each other's inboxes.

Over six months, hundreds of emails pass between them. The data room is built by hand. The disclosure letter starts from a precedent someone remembers using three years ago, eventually found in a colleague's mailbox. The banker's buyer list and the lawyers' due diligence findings sit in different files, on different servers, in different firms.

The deal closes and everyone moves on. Very little of what was learned is kept in a form anyone can use again. The next team starts from a folder of old documents and whatever the partner happens to remember.

Then the firm buys an AI tool. The associate drafts the disclosure letter faster, and the banker summarises the data room in an afternoon. Everyone feels more productive, and the economics of the deal for all participants do not change.

AI should be transformational rather than peripheral

That last sentence is not a guess. McKinsey's global survey on the state of AI, published in August 2026 and based on 1,719 respondents in 97 countries, describes exactly this gap.

Eight in ten respondents say AI has improved their own productivity, and half say it helps them make better decisions. Yet only 37% say AI has contributed anything to their organisation's earnings, the same as a year earlier. Around 6% attribute 5% or more of their EBIT to AI and describe the value as significant, and that share has not moved either.

What sets that small group apart is not simply budget. According to Mckinsey, they are 3.3 times more likely than others to intend to use AI to fundamentally transform their business within three years. Nearly three quarters of them have fundamentally redesigned their workflows because of AI, up from 55% a year earlier, against a quarter of everyone else.

McKinsey's conclusion is direct. The organisations that turn individual productivity into lasting financial performance "are likely to be those that transform their businesses, not just adopt AI tools."

There is no dry dock at sea

Large organisations are not standing still. In the same survey, 40% of respondents from companies with more than USD 1 billion in revenue report scaling AI agents, up from 27% a year earlier. They are adopting faster than anyone, yet across the survey the share of organisations seeing any effect on earnings has not changed.

The reason is structural. Organisations, not unlike the individuals that make up such organisations, are like sailors who must rebuild their ship on the open sea.

An established firm is in exactly that position. It can replace one plank at a time: a new tool, a pilot, a better template. It cannot haul the ship out and redesign the hull, because every partner, client and live matter is still on board, and a hull taken apart at sea sinks. In professional services, four things hold that hull in place.

The first is economics: Law firms are largely paid by the hour, and in most advisory firms the year's profits are paid out to partners or as bonuses. AI that halves the time a task takes also halves what an hourly firm can bill for it, and an investment that pays back over five years is hard to fund from a profit pool emptied every twelve months.

The second is data: Decades of work product sit in inboxes, drives and document systems that differ from office to office. None of it was captured with reuse in mind, and it is close to impossible to put structure on twenty years of email after the fact.

The third is modern education: Legal advice, deal advisory and asset management usually sit in separate firms, each with its own systems, conflicts process and incentives recruiting strictly along the lines of what somebody has studied (more or less by accident). The client gets several advisers and the job of coordinating them, and often no single adviser sees the whole transaction, or understands it as a real world problem, but focuses on a faction of it that fits into his or her university degree.

The fourth is people. Thousands of professionals with established habits do not change how they work because a new tool arrives. McKinsey makes the same point: the limiting factor is increasingly an organisation's ability to absorb change, not the technology.

So the large firm does what it can. It buys tools, runs pilots and layers AI onto the workflows it already has. Its people become more productive, and the ship keeps sailing exactly as before.

Starting in the dry dock with the new tools available

When we started Novara, we had one advantage no established firm can buy: established businesses coming together in a new firm we could built from scratch with full awareness of the new AI enabled capabilities. Our ship was still in dry dock, when the new tech became available. We could design the hull before anyone came aboard.

Starting in dry dock is an advantage that lasts only a few years, because every firm builds up its own history and ours will be no exception. We have therefore designed Novara so that it can return to dock whenever the water changes, one section at a time, while the rest of the firm keeps working. That depends first on how we hold our data. Our client, matter and precedent records are designed to sit in one structure the firm owns, independent of any single tool, so that when a better model or system appears we can connect it to what we already have instead of moving years of work across by hand.

It also depends on how we work. We write our processes down and build them into our systems rather than leaving them in individual habits, which means that changing how we run a matter is a change made once and adopted by everyone, not a programme to retrain a workforce.

Finally, it depends on how we are paid. We agree fees engagement by engagement, as fixed fees, caps, retainers or success fees as well as by the hour, so when our tools make us faster the benefit is shared with our clients instead of eroding our revenue.