Skip to main content

Who Captured the AI Productivity Gain?

on

Who Captured the AI Productivity Gain?

Suppose your AI program works. Really works.

Employees are more productive. Costs fall. Output rises. The initiative produces measurable economic value.

Excellent.

Now ask the question most AI ROI conversations skip: Who captured the gain?

Because creating value and keeping value are not the same thing.

That distinction sits at the center of the second stage of the Create → Capture → Defend framework.

The productivity gain has to go somewhere

A recent Harvard Business Review article, AI and the Looming Competition for Margin, argues that widely diffused technologies can produce a surprising economic outcome.

Productivity rises. Competition intensifies. Prices fall. Margins come under pressure.

In other words, customers may ultimately capture a meaningful portion of the productivity gain. That is good for consumers. It is less obviously good for the margins of the firms making the investment.

This creates a problem for the standard AI story.

Companies often imagine the sequence as:

AI → Productivity → Lower Cost → Higher Profit

But competitive markets can insert another step:

AI → Productivity → Lower Cost → Competitive Response → Lower Prices

The productivity is real. The margin expansion may not be.

Value creation ≠ value capture

That distinction is worth making explicit:

Value creation ≠ value capture

Suppose AI reduces the cost of producing a service by 20%.

Initially, the company may enjoy higher margins. Then competitors deploy similar systems. One lowers prices. Another bundles additional services. A third increases output.

Customers begin expecting more for less. Eventually, much of the original productivity improvement may be reflected in market prices rather than exceptional profits.

The company created value. The market redistributed it.

So who wins?

The answer can vary.

  • Shareholders: The company may retain the gain through higher profits, margins, cash flow, or enterprise value.
  • Customers: Competition may transfer the gain through lower prices, faster service, better quality, or greater choice.
  • Employees: Workers may gain additional capacity, better tools, higher compensation, or less repetitive work.
  • Suppliers: Changes in bargaining power or demand may shift some benefit upstream.
  • Competitors: Rivals may reproduce the efficiency quickly enough to neutralize the original advantage.

There is no reason to assume that the company paying for the AI investment automatically captures most of the economic benefit.

That is what makes this a management problem rather than merely a productivity problem.

HBR's uncomfortable argument

The HBR authors go further.

They argue that the industries most successful at converting AI into productivity may eventually face some of the fiercest price competition and margin pressure. The bigger the productivity impact, the larger the potential deflationary force.

That turns the conventional AI narrative upside down:

AI success itself may intensify the battle over who keeps the value.

This does not mean AI investment is pointless. Quite the opposite.

It means companies may need AI simply to remain competitive. The authors describe that possibility directly: what appears to be investment for advantage may eventually become investment required for survival.

That is a very different ROI conversation.

From ROI to value capture

Traditional ROI asks: Did this investment create more value than it cost?

Still important. But AI may require a second question: How much of the value did we actually retain?

Imagine two firms each generate a 15% productivity gain.

  • Firm A converts most of that into higher margins and reinvests the proceeds.
  • Firm B passes nearly all of it through to customers because of price competition.

Both created productivity. Their economics are very different.

Now imagine Firm C improves only 10% but possesses strong customer relationships, proprietary data, pricing power, or switching costs that allow it to retain more of the gain.

The largest productivity improvement does not necessarily produce the strongest strategic outcome.

That suggests companies should monitor not only: How much value did AI create? but: Where did that value land?

The questions get more strategic

For each major AI initiative, leaders may eventually need to ask:

  • Did cost savings improve margins?
  • Did lower costs simply lead to lower prices?
  • Did increased capacity create additional revenue?
  • Did customers capture most of the benefit?
  • Did competitors match the gain?
  • Did the initiative strengthen our market position?

That is much harder than measuring usage. But it is much closer to the real economic question.

AI adoption is not the race

As AI tools become increasingly available, possessing AI may become less important as a differentiator.

The HBR authors argue that broad accessibility makes many AI advantages difficult to defend because competitors can use similar underlying models and tools.

Which leads to a larger idea:

AI adoption is not the race. AI value capture is the race.

Creating the gain is only the beginning. The company still has to retain enough of that gain to justify the investment.

And even that may not be enough.

Because once the company successfully captures value, another question appears: How long can it keep the advantage before competitors catch up?