Do data-driven companies perform better?

Anyone who advocates for evidence-based decisions has an annoying obligation: apply the same rule to themselves.

If the question this site keeps asking is "how do you know?", it also applies to Data-Driven Culture itself. Do companies that make decisions with data perform better? Or is that just a nice line for a keynote?

We went to the research. Not to one study, but to several. Including meta-analyses, which are studies that pool dozens of other studies.

The study that opened the discussion

In 2011, Erik Brynjolfsson, Lorin Hitt and Heekyung Kim analyzed 179 large publicly traded companies. Those that adopted data-based decision-making had "output and productivity 5 to 6% higher" than expected, even considering their other technology investments.

The important detail is in "even considering". It was not who bought more software. It was who brought data to the decision.

The authors also tested whether the story ran the other way around, with good companies simply using more data. They found signs that this is not the whole explanation.

What 125 studies say together

In 2022, Thuy Duong Oesterreich, Eduard Anton and Frank Teuteberg brought together 125 studies, from 26 countries, covering ten years of research.

Result: analytics is associated with better operational, financial and market performance.

But the most interesting finding comes after. "People, managerial capacity and organizational culture weighed more in value creation than technology."

A second meta-analysis from the same group, with 107 studies, arrived at the same place: technical factors matter, but the human and organizational structure factors predominate.

"Technology helps. Who transforms capacity into results are people, management and culture."

Culture appears measured

Two studies measured data-driven culture directly.

• In Turkey, 432 specialists from 132 companies showed that data-driven culture bridges the gap between having analytical capacity and having operational and financial performance (Karaboga et al., 2023).
• In 149 micro, small and medium enterprises, the ability to spot change and the ability to analyze it improved operational performance. And "data-driven culture amplified that effect" (Wong and Ngai, 2023).

The second study matters for a simple reason: the logic is not the exclusive domain of tech giants.

Better analysis alone is not enough

A 2025 meta-analysis, with 34 studies, found a positive effect of big data analytics on performance. Except that part of this effect passes through something else: "organizational agility" (Liu, Singh and Al Shibli).

In business language: there is no point producing a better analysis if the organization cannot react to what it reveals.

This is exactly the idea of consequence. Data that changes nothing is just a record.

The number that circulates incorrectly

Perhaps you have already seen it: "data-driven companies are 19 times more profitable".

The origin is a 2013 McKinsey survey of 400 executives, on a specific type of analysis: customer analytics. The most intensive users were 19 times more likely to report above-average profitability than the least intensive.

It is a strong finding. But it is about one type of analysis, it is self-reported, it compares extremes and does not prove causation. Turning this into "data multiplies profit by 19" is exactly the kind of leap that Data-Driven Culture exists to prevent.

What the research does not say

Intellectual honesty also means saying where the evidence stops.

• They do not say that using data automatically causes profit. Only one of the studies attempts to isolate causation, and in large companies.
• They do not say that buying technology improves performance. They say almost the opposite.
• Most studies use questionnaires and perceived performance, and studies without effects tend to be less published.
• None of the studies we reviewed has a Brazilian sample.

The pattern

Putting it all together, a chain appears:

perceive and analyze better → decide better → react better → perform better

And, regulating how much of technology reaches the result: people, management and culture.

This is not proof that data generates profit. It is something more useful: a consistent clue about where value is born and where it is lost.

This is an open line of research. We will keep adding studies. Including the ones that find no effect.