The McKinsey Report That Is Terrifying Every Mid-Size Company in America Right Now

Someone is sitting with a printout of the McKinsey numbers in a conference room somewhere in a mid-size company, such as a logistics company that has been quietly profitable for twenty years, a professional services company in Charlotte, or a regional manufacturing firm in Ohio. They are trying to figure out exactly what to do with them. The coffee is becoming chilled. The laptop has the consultants’ slide deck open. The fact that US small and mid-size businesses are only 47% as productive as large companies is another statistic that keeps coming up and keeps the room from relaxing. Not eighty percent. Not sixty-five percent. Forty-seven. The McKinsey Global Institute’s analysis of over 12,000 companies revealed this disparity, which is not the result of rounding errors. It’s a structural issue that has been developing for years while everyone’s attention was diverted.

This is the essence of what McKinsey has been publishing in a number of reports: a collection of research on mid-size and small business productivity that, when combined, presents a genuinely unsettling picture for the businesses that don’t make headlines but, taken as a whole, employ almost six out of ten American workers and contribute roughly 39% of US business value added. These businesses are not marginal. The suppliers, local banks, professional service companies, and specialty manufacturers are the backbone of the economy. According to McKinsey, the productivity difference between them and their biggest rivals is equal to 5.4% of the US GDP. That is not a topic of discussion for consultants. That is a huge amount of unproduced economic output.

The AI component adds complexity and urgency to the situation. According to McKinsey’s 2025 AI report, 88% of businesses say they use AI for at least one business function. That seems like progress at first glance. However, the narrative swiftly shifts when the number is removed. Two-thirds of those businesses are still in the pilot or experimentation stage; they continue to conduct controlled experiments, inquire as to whether this will be successful, pay consultant hours and licensing fees, and see no results on a profit-and-loss statement. Just 39% of all businesses polled were able to link their AI investments to any real impact on earnings. Additionally, roughly 6% of respondents fall into the category that McKinsey defines as true AI high performers, which are businesses where AI contributes more than 5% of earnings before interest and taxes. Six percent. That’s six people raising their hands in a room full of one hundred companies.

It turns out that there is a fairly obvious explanation for the difference between the six percent and everyone else, and it has nothing to do with a company’s choice of AI platform or its budget. Instead of just adding AI tools to their current processes, high performers are three times more likely to have completely redesigned their workflows. More important than nearly anything else in the data is that distinction. The majority of businesses view AI as an efficiency overlay; they identify a manual step, automate it, and refer to it as progress. The businesses that were actually making money posed a different query: what would this process look like if we were to rebuild it from the ground up, knowing what AI is capable of? Only 55% of high performers asked that question, compared to 20% of everyone else, which is likely because it is more difficult, slower, and disruptive. The disparity in ambition leads directly to the disparity in results.

This puts a certain amount of pressure on mid-size businesses in particular. They typically have less institutional experience with significant transformations, lack the financial reserves and talent infrastructure of large firms, and face competition from both nimble startups and the giants above them. In semiconductor materials, for example, top performers generate twice the EBITDA margins of the weakest competitors, according to McKinsey’s analysis of the financials of 20,000 midcap companies. This variation implies that the category isn’t experiencing consistent difficulties. A few mid-sized businesses are handling the changeover pretty well. However, the median company isn’t, and in industries where AI adoption is accelerating, the gap between the median and best-in-class is growing.

It’s difficult to ignore the irony that permeates everything. The company that published these results, McKinsey, laid off 3,000 workers in 2024, the most in its 100-year existence. The company that has spent decades offering transformation advice to businesses was forced to restructure. That doesn’t negate the analysis, but it does imply that even the organizations conducting the research are finding it extremely challenging to deal with the current situation. Advisors are not exempt from the forces reshaping business productivity.

A fairly simple set of requirements for survival is what McKinsey’s accumulated work on mid-size businesses consistently returns to: set an ambitious goal that genuinely frightens the organization, invest in talent before the need becomes apparent, and resist the temptation to treat technology as something that can be bolted onto a company’s existing structure without changing the structure.

The Napa wine cluster, which McKinsey cites as an example of collaborative scale, and the Indian specialty chemicals company, which grew tenfold over a ten-year period by making the journey feel personal to its workforce, are two examples of companies that have successfully accomplished this by approaching transformation as a sincere commitment rather than a project. The question in that conference room with the cold coffee and the open laptop is whether today’s mid-size businesses can accomplish the same, fast enough, in the face of growing expenses, AI-enabled rivals, and a productivity gap that keeps getting wider.