Every Growth Wave Gets Mistaken for a Cost-Cutting Tool.

Should you point AI at growth or at cost cutting? A growth wave rewards growth, and AI is a growth wave. Six times in the last two centuries a general-purpose technology reset the rules of business. Each time, most companies reached for the same prize, which was to do what they already did for less money. And each time, the largest returns went to the companies that used the new capability to build something they could not build before. For a small or midsize business, that pattern is the opportunity, because the growth posture is still open and the companies best suited to take it are the ones small enough to change their own model.

What do the last technology waves have in common?

In every one, the payoff arrived through redesigning the business around the new capability, not through doing the old work more cheaply. Steam, electricity, oil, the transistor, the personal computer, the internet: the list is familiar, and the pattern that runs through it is consistent enough to trace wave by wave.

The steam engine is where it first shows up clearly. In 1865, William Stanley Jevons pointed out something that read like a contradiction: as steam engines became more efficient and burned less coal for each unit of work, Britain's total coal consumption did not fall, it climbed, and it would keep climbing for the rest of the century.¹ Cheaper power meant power got put to more uses, in more places, by more firms. Efficiency did not conserve the resource. It financed an expansion. The companies that understood this were not trying to spend less on coal. They were asking what they could now build, move, and make that had been uneconomical the year before.

Electricity carried the lesson further, and it took its time. The Stanford economist Paul David studied why electric power delivered so little measured productivity for decades after it arrived, and then a great deal at once.² The first factories to electrify simply dropped an electric motor where the steam engine had been and kept the same overhead shafts and belts driving every machine from a single central drive. The gains were small. The leap came around the 1920s, when plant designers rebuilt the factory around the motor itself, one motor per machine, machines arranged by the flow of the work rather than by their distance from the drive. That was a different factory, and it was far more productive. The technology had been available for forty years. The business model to use it had not.

Computing produced the most quoted version of the same delay. In 1987 the economist Robert Solow observed that you could see the computer age everywhere except in the productivity statistics.³ The transistor, the semiconductor, and the personal computer had already remade daily work, yet the payoff stayed hidden until the 1990s, when it showed up alongside real changes in how companies organized around the machines. Firms that used PCs to do the old paperwork faster got faster paperwork. Firms that rebuilt how information and decisions moved got something much larger.

The internet is the wave I watched from the inside, and I watched it twice. The first time, I had the cost-cutting instinct myself. I started my career at Air Products marking up engineering drawings by hand and mailing them to the plants, and observing process charts by standing at the equipment for a week with a pen, noting the level as the paper rolled past. When the network arrived, the first thing it did was the old job faster. The markup became an electronic redline. The weeklong trip became a session over the office network, and then a modem at my kitchen table. The postage and the airfare were real savings, and they were the smallest part of what changed. A shared electronic redline let several people mark the same drawing at once, with everyone downstream holding it instantly, which paper could not do. Once watching a process no longer meant standing beside it, presence stopped being the limit on how far the work could reach. Those advances ran alongside a company in growth mode, adding plants across the United States and then Mexico, Brazil, Korea, and a joint venture in Germany. The savings were the trips we stopped taking. The growth was a footprint that presence alone could never have covered.

The second time, I was inside a company built for the wave. In the late 1990s I joined FreeMarkets, a Pittsburgh company that ran online reverse auctions for industrial purchasing. The point of FreeMarkets was never cheaper paperwork. The internet made a market possible that had not existed before, a global, real-time competition among suppliers for a buyer's business. That was a new business model riding a new technology, and it created a category. The companies that treated the internet as a cheaper brochure got a cheaper brochure. The ones that asked what new thing it made possible built the decade.

Why does a cost-cutting mindset underperform in a growth wave?

Because efficiency gains are capped and easy to copy, while growth gains compound and set you apart. There is a floor under how much you can save, which is the cost of the thing you are trying to do. There is no matching ceiling on what a new capability lets you sell, reach, or offer. A competitor can buy the same tool you bought and erase your savings advantage by the next budget cycle. A competitor cannot as easily copy a business you built around the tool while they were still trimming theirs. Cost work is worth doing. It is simply the smaller half of the prize, and in an innovation wave it is the half most companies reach for first.

Isn't AI mostly an efficiency tool?

The current evidence points the other way: the companies getting the most from AI are the ones that refused to stop at efficiency. In McKinsey's 2025 State of AI research, the firms reporting the strongest results are those that set growth and innovation as objectives alongside cost, and they report better outcomes on differentiation, profitability, and revenue than the firms chasing savings alone.⁴ BCG's 2025 work put a number on the spread: the small group it calls future-built, about 5 percent of companies, expect roughly twice the revenue increase and 40 percent greater cost reductions than the laggards in the areas where they apply AI.⁵ Read those findings together and the instruction is plain. The leaders are refusing to leave the growth on the table.

What a growth use case looks like for a smaller company is concrete. It looks like reaching customers you could not serve before, offering something you could not staff for, or moving faster than a larger competitor can. We worked with a regional operator whose first AI initiative expanded its hiring reach so sharply that a manager took more than fifty calls on a single job posting, the first time in a twenty year career. The constraint that surfaced next was a growth problem, how to handle all the interest, which is a problem a cost-cutting project rarely produces.

Why are small and midsize businesses better positioned to take the growth posture?

Because taking it means changing how the business works, and a smaller company can change itself in weeks where a large one needs a year. Fewer layers sit between the person who spots the opportunity and the person who can approve it. There is less legacy process to unwind and less internal politics guarding the old way of doing things. The entry tools are now cheap enough that size is less of an advantage than it was in earlier waves. The first-mover window for a genuinely new idea is open to the companies fast enough to walk through it.

How do you adopt AI as a growth strategy?

Start by naming a growth constraint instead of a cost line. Pick the thing you would do more of if you had the people, the reach, or the speed, and ask whether AI removes that limit. Redesign the work around the new capability rather than bolting it onto the old process, which is the lesson every wave in this piece teaches. Then put what you free up back into the next constraint rather than banking it as a cut. That is where the compounding starts, and it is the same discipline whether you are pricing the program so it pays for itself or setting the goals and guardrails before you begin.

Proxigee Services is an AI advisory firm in the Pittsburgh area that helps small and midsize companies take their first strategic steps into AI, choosing the growth use cases worth attacking and building the capability to run them. The waves before this one rewarded the companies that asked what the new capability made possible and then built for it. This one is asking the same question.

Efficiency is the smaller half of the prize, and in wave after wave it is the half most companies reach for first.

The wave is here. The question that decides your return is which half of it you reach for.