The biggest winners of AI need not be the companies with the most engineers, the largest data teams, or the highest software budgets. They may be the companies with the lowest margins. The manufacturers, trucking carriers, distributors, staffing agencies, and field-service operators that have run on thin, single-digit margins for decades – the businesses nobody would ever call AI companies. AI transformation creates value through three levers: revenue, cost, and risk. Most of the attention has gone to revenue through better products, faster sales, and more productive employees. But for low-margin businesses, the biggest lever to pull is cost – when profit margins are already thin, even small reductions in operating expense can create an outsized increase in earnings. A software company running at 30% margins can use AI to become more efficient, but that efficiency gain usually does not change the trajectory of the business. A business running at 3% margins is different. Even a <1% cost reduction can lead to >25% profit increase. Low-margin industries have historically been trapped in structural, low-margin environments. Properly implemented AI changes that equation. It gives low-margin businesses a way to attack costs that were previously treated as permanent – and the companies that move first are able to capture that gain as margin before competitors force it back into lower prices. Efficiency spreads across a commoditised market eventually, but the early movers are the ones who bank the earnings uplift and reset their cost position ahead of the field. By the end of this article, you should understand how the lowest-margin businesses can finally attack the coordination costs that have kept them structurally low-margin for decades – and why the companies that move first will pull away from the rest of their industry. The providers who solve this will build billion-dollar companies – and the businesses they transform will be the ones that escape the margin trap first. The structural barriers low-margin companies face For most low-margin businesses, there were structural barriers that have kept them locked into this position. They usually compete in commoditised markets, have limited pricing power, and carry large operating cost bases that were previously impossible to reduce without hurting service quality. Because they cannot move the market price – the market sets it, not the individual company – cost is effectively the only lever they control. A meaningful share of that cost base is labour – and beyond the physical work itself, these companies also carry the cost of coordinating it. There is a long list of coordination work that erodes the margin of these companies over time. For instance, scheduling, dispatching, approvals, exception handling, and countless administrative loops are incurred by labour-intensive companies, and thus, eat away at a company's bottom line. That coordination work is where AI has the clearest opportunity to move the needle for labour-intensive, low-margin businesses. In these types of companies, labour costs typically account for nearly 25% of revenue. Roughly a quarter of this labour spend is tied to managing, coordinating, and administering the work, equating to ~6% of revenue. For a company operating at a 3% margin, easing the coordination burden by 10% can improve earnings by ~20%, changing the entire earnings profile of the business. As a result, AI does not just make them slightly more efficient. It gives the companies that adopt it early a chance to open a structural cost advantage over their competitors – and to run as a genuinely higher-margin business, perhaps for the first time. The problem is that the companies with the most to gain from AI are often the least able to adopt it Most solutions sold in enterprise AI today have the assumption that employees will adopt a new tool, use it correctly, and slowly turn usage into value that gradually gets realised in the P&L. Considering that this assumption doesn’t hold even inside tech-forward companies, it is only worse inside a manufacturing company, logistics business, or any other labour-heavy company where the workforce is not used to adopting a new software product. These businesses are often the least susceptible to change management. The real question is how to get AI-driven margin expansion without relying on employee adoption – or at least without enforcing new interaction surfaces. That is the challenge, and its solution might be the most addressable trillion-dollar opportunity in AI right now. Three steps to solve the trillion-dollar low-margin challenge 1) Find the hidden coordination cost Most people think about AI cost savings too narrowly. They imagine replacing a task, reducing headcount, or making an employee faster. That can matter and will likely happen in the future, but where AI capabilities are today, a significant piece of the opportunity is the work behind the work: the overhead required to keep mes