It has never been easier to know how to grow a company. Why is it still so hard to do it?

Knowing what to do has become cheap. What remains scarce is what a company can actually do with what it knows.

Key ideas

• Productivity differences between companies in the same industry are large and persistent, and in some sectors they widened in the middle of the internet era.
• Between information and results there is a chain: absorb, choose, adapt, execute and coordinate, learn and accumulate. Any link can break.
• Two companies don’t get the same value from the same information, because their ability to absorb it depends on what they already know.
• Widely adopted best practices produce competitive parity, not advantage: advantage still requires choices and combinations that are hard to reproduce.
• Capabilities accumulate over time and can’t be compressed, but what accumulates can also turn into rigidity.
• AI lowers the cost of access to certain capabilities, but it may raise the value of the assets around it. Its net effect on the gap between companies is still unknown.

The gap that information didn’t close

Before explaining the paradox, it is worth confirming that it exists. Economics has a fairly emphatic answer. Chad Syverson, of the University of Chicago, reviewed the evidence on productivity differences between firms and highlighted a figure that surprises anyone seeing it for the first time: within narrowly defined U.S. manufacturing industries, the plant at the 90th percentile of productivity produces almost twice as much as the plant at the 10th percentile with the same measured inputs.¹ In China and India, according to the work by Chang-Tai Hsieh and Peter Klenow that Syverson cites, the ratio exceeds five to one.² And these aren’t passing differences: a firm’s productivity tends to persist from one year to the next.

The recent trend is even more striking. An OECD team (Dan Andrews, Chiara Criscuolo and Peter Gal) compared the most productive firms in each sector with the rest, across dozens of countries. Between 2001 and 2013, frontier firms in manufacturing increased their productivity by 2.8% a year, compared with 0.6% for the rest; in services, the difference was 3.6% versus 0.4%.³ The authors attribute part of this divergence to winner-takes-all dynamics, to the growing importance of tacit knowledge and to a slower diffusion of leading firms’ practices to everyone else. In other words, in the same period in which the internet made access to information cheaper than ever, the distance between the best companies and the rest widened.

These figures describe the gap; on their own, they don’t prove what causes it. Competition, regulation, access to capital and many other factors play a role. But they rule out the simplest explanation: if differences between companies depended mainly on who has access to knowledge, they should be shrinking. The question, then, is where the value of knowledge gets lost on the way to results.

From information to results: a six-link chain

Saying that some companies “execute better” explains nothing unless we spell out what executing better means. To structure the problem, we propose The Information-to-Results Gap by Advisent®, a conceptual framework (not a validated causal model) that separates what is usually treated as a single thing. Between available information and a company’s results there are six links, and organizations differ at every one of them.

Available information

1. 01AbsorbRecognize what is relevant and understand itBreaks when the company can’t tell what applies to its case
2. 02ChooseJudgment and decisions, with trade-offsBreaks when everything is a priority
3. 03AdaptTranslate the general into one’s own contextBreaks when a practice is copied as is
4. 04Execute and coordinateTurn the decision into collective behaviorBreaks when every area delivers and the whole doesn’t work
5. 05LearnUnderstand what worked, what didn’t and whyBreaks when experience changes nothing
6. 06AccumulateTurn what was learned into capabilities and assetsBreaks when knowledge leaves with people

Results

↺ Renew: what has accumulated is revisited when the environment changes

Conceptual framework by Advisent®, not a causal model or validated scale. The links don’t always occur in this order and they influence one another. Source: Advisent®.

The first link, absorbing, is the most underestimated, and we develop it in the next section. The second, choosing, is a reminder that abundant information doesn’t remove the need to decide; it may even increase it, because the more reasonable alternatives there are, the more important the judgment needed to discard most of them. The third, adapting, explains why a best practice isn’t necessarily the best decision for every company: it has to be translated to a specific market, customers, resources, regulation and moment. The fourth combines execution and coordination because, beyond a certain size, executing is mostly coordinating: individual capability doesn’t guarantee collective capability.

The last two links are what turn a good year into an advantage. Learning means turning experience into concrete change, and not all experience produces learning: a company can operate for ten years repeating the same mistakes. Accumulating means that what has been learned stays in the organization, as a capability or an asset, and not only in a few people’s heads. Any link can be the bottleneck: good information with poor judgment produces bad decisions; a good decision with low capability doesn’t get executed; good execution without coordination produces inconsistent results; execution without learning produces temporary performance; and learning without accumulation produces knowledge that gets lost.

Why the same information is worth more to some companies

In 1990, Wesley Cohen and Daniel Levinthal proposed a concept that is more relevant today than when they formulated it: absorptive capacity. They defined it as a firm’s ability to recognize the value of new external information, assimilate it and apply it to commercial ends, and argued that it depends largely on the prior related knowledge the firm already has.⁷ In other words, to make use of an idea you need to know quite a lot about the subject. Two companies can read the same study, attend the same conference or receive the same answer from an AI, and only one of them will recognize which part applies to its case, how to combine it with what it already does and what to change in order to use it.

Cohen and Levinthal’s work predates the internet as we know it and, of course, generative AI. It studies neither. But its logic supports a reasonable inference: if information becomes abundant, the ability to filter, interpret and apply it becomes the constraint. Democratizing access doesn’t democratize absorptive capacity.

There is empirical evidence consistent with this idea. Nicholas Bloom, Raffaella Sadun and John Van Reenen studied why U.S. productivity accelerated more than Europe’s as information technology spread. They found that, within Europe, subsidiaries of U.S. multinationals got more productivity out of the same technology investment than those of multinationals from elsewhere, and that establishments taken over by U.S. firms increased the productivity of their technology. The main explanation lay in their people management practices.⁴ The technology was available to everyone; what made it productive was not.

Intellectual honesty requires an important caveat. In an experiment with large textile firms in India, Bloom and his co-authors offered free consulting on management practices to randomly chosen plants. Productivity rose 17% in the first year, and within three years the treated firms opened more plants. When analyzing why those firms hadn’t adopted such profitable practices earlier, the authors concluded that the main reason was informational barriers.⁵ Sometimes, then, the problem really is not knowing. The thesis of this article isn’t that information no longer matters, but that as it becomes accessible it explains a shrinking share of the differences, and that even when it is missing, its value depends on whether the organization can absorb it.

Best practices produce parity, not advantage

There is a second reason why accessible knowledge doesn’t guarantee growth, and it has to do with competition. In 1996, Michael Porter distinguished between operational effectiveness and strategy.⁸ Operational effectiveness means performing activities similar to rivals’, but better. It is essential, and best practices help achieve it. The problem is that they spread: the more companies adopt them, the more alike they become and the less difference they make. Strategy, by contrast, means choosing a different set of activities, accepting trade-offs and configuring those activities so they reinforce one another.

Porter illustrated the point with a well-known case from U.S. aviation. Continental Airlines tried to imitate Southwest’s low-cost model on some routes, through an operation it called Continental Lite, while keeping its full-service model everywhere else. It copied visible elements but not the system of choices that made them work, and according to Porter’s account the initiative ended with losses of hundreds of millions of dollars. The case doesn’t prove that imitation always fails, but it shows clearly that visible practices detached from the system that supports them rarely reproduce the results.

Jay Barney formalized a complementary idea: the resources that can sustain a competitive advantage are valuable, rare, hard to imitate and hard to substitute.⁹ It isn’t a recipe, because a resource’s value depends on the market and the moment. But it has a direct implication for our question: if a piece of information is available to every competitor, possessing it can’t, on its own, be a source of sustainable advantage. The difference, if there is one, lies in what is built around it.

The distinction has an immediate practical consequence. Many investments that companies present as strategic are really investments to avoid falling behind. Both can be rational, but they are different, and confusing them leads companies to expect differentiation from something that only produces parity.

Same tools, different systems

A hypothetical example shows the difference in a concrete area: customer acquisition. Two consulting firms use LinkedIn, a prospecting tool, the same CRM, generative AI, email campaigns, content and paid advertising. They have read the same books and have access to the same frameworks. The first, however, better understands which segment gets the most value from its service, formulates a clearer offer, tests messages, learns from every call, records objections, adjusts its pricing, builds case studies, improves its delivery and generates referrals. The second implements the same tools without developing that learning system. Both have the same tech stack. They don’t have the same system, the same capability, or the same results ahead of them.

The mistake the example illustrates is common: copying the tool you can see. “They use HubSpot,” “they do outbound,” “they post on LinkedIn,” “they use ChatGPT.” The tool is only the visible part of a deeper system, and superficial benchmarking copies visible outputs without understanding what produces them. We look at how that full commercial system works in our article on the profitability of the customer relationship.

Visible easy to copy

Tools · Tactics · Channels · Content · Campaigns

What can be seen from the outside

Operating less visible

Data · Processes · Workflows · Decisions · Customer knowledge

Deep hard to reproduce

Learning · Relationships · Accumulated experience · Coordination · Organizational design

Conceptual representation by Advisent®. Copying the visible layer rarely reproduces the results the full system produces. Source: Advisent®.

What time accumulates and what time hardens

If capabilities can be learned, why aren’t they copied quickly? Ingemar Dierickx and Karel Cool gave one of the most elegant answers in 1989. A company’s strategic assets (reputation, customer knowledge, relationships, experience) are stocks that build up over time through flows of investment, like water in a bathtub, rather than being bought ready-made in the market.¹⁰ Flows can be changed overnight; stocks can’t. The authors also identified two decisive properties. Time compression diseconomies: doubling the effort for half the time doesn’t produce the same result, which they illustrated with the anecdote of the English gardener who, asked how to achieve a perfect lawn, replies that you simply keep it up for five centuries. And asset mass efficiencies: whoever already has a large stock grows it faster, because success breeds success.

On that basis, we propose the concept of capability compounding, an Advisent® editorial label rather than an established academic term: the process by which accumulated experience, learning, assets, relationships, data and routines progressively increase an organization’s ability to produce certain results. A company that began building a capability three years before its competitor has accumulated, in that time, experiments, mistakes, customers, data, case studies, referrals, processes and reputation. When the competitor copies the visible practice, it can’t copy the path that produced it. That is why two companies making the same decision today can get very different results: they don’t start from the same place. Nor should this be read as a sentence for latecomers: followers can learn from others’ mistakes, and first-mover advantage isn’t universal.

But time doesn’t only accumulate; it also hardens. Dorothy Leonard-Barton showed that the core capabilities that give a company its advantage can become core rigidities when the environment changes: the same knowledge, systems, values and routines that enabled it to win make it harder to adapt.¹¹ The case of Polaroid, studied by Mary Tripsas and Giovanni Gavetti, shows this precisely.⁶ The company invested early and seriously in digital imaging (by 1989, more than 40% of its research and development spending went to that area) and achieved significant technical advances. But its leadership kept reasoning with the logic that had made its business successful, in which the money was in the film rather than the camera, and that mindset shaped which products it developed and which it didn’t. It had the technical capability; what it lacked was the renewal of the beliefs that guided how to use it. Experience, therefore, isn’t a permanent advantage. Advantage requires accumulating and renewing at the same time, which is what the literature calls dynamic capabilities, a topic we discuss in our article on internationalization and competitiveness.

A competitor can copy what your company does. Copying the path that led it there is far harder.

AI makes knowledge cheaper, not necessarily capability

Generative artificial intelligence takes the paradox to a new level. It helps to be precise about what it is making cheaper: the cost of searching, summarizing, translating, drafting, standard analysis and coding. That turns much generic content and basic execution of codifiable tasks into a commodity. But not all knowledge is public, not all data is accessible, not everything a model produces is correct, not everything can be codified, and not every organization has the same ability to evaluate what it receives. AI may democratize access to certain capabilities. That isn’t the same as equalizing companies.

Decades ago, David Teece offered the concept that best explains why. Studying who captures the value of an innovation, he showed that it is often not the inventor but whoever controls the complementary assets needed to bring it to market: distribution, brand, customer relationships, production capacity, related services.¹² Applied to AI, this reasoning changes the question. It is no longer only which model a company uses, but what it has around that model that lets it turn it into value: proprietary data with usage rights, integrated workflows, expert knowledge to evaluate outputs, customer relationships, distribution and feedback loops. Two companies can use the same foundation model; the organizational capability they build around it can be completely different. We develop this with evidence in our article on artificial intelligence and competitive advantage.

What we don’t yet know is whether AI’s net effect will be to narrow or widen the gap between companies. Some studies show that, within an organization, AI can shrink the differences between more and less experienced workers. Between companies, however, the growing value of complementary assets could widen the distance between those who have them and those who don’t. It is an open question, and any categorical answer is, for now, a bet. What does change is the value of a certain kind of expertise. When generating answers is almost free, framing the problem well, prioritizing, critically evaluating what you receive, integrating and implementing all become more valuable. The easier it is to produce answers, the more important it may be to know which ones are useful.

Widely available practice

CRM · automation · AI models · digital marketing

↓ adoption

Competitive parity or necessity

Hard-to-reproduce combination

Capabilities + complementary assets + integration + learning, accumulated over time

↓ building

Potential differentiation

Conceptual representation by Advisent®. A complex combination doesn’t guarantee advantage: it only creates one if it delivers value to customers and is hard to imitate or substitute. Source: Advisent®.

What this means for a company that wants to grow

In service companies, knowledge and expertise are often a large part of what is sold, which makes the paradox more visible. Growth runs into dependence on founders and a few specialists, concentrated knowledge, delivery capacity, quality consistency and leadership’s limited time. The central question is how to turn individual expertise into organizational capability, a topic we cover in our article on talent and capabilities. In product companies, advantage usually depends on combinations of product, manufacturing, supply chain, distribution, brand, data and capital. The best practices in each of those areas may be public; the organizational configuration needed to combine them isn’t.

In Latin America the argument has its own nuances, without treating the region as a single block. Capital constraints, currency volatility, informality, fragmented markets, regulatory differences and direct competition with global companies shape which capabilities make sense to build and how fast. At the same time, access to global technology has become practically the same as in any other region, which makes it even clearer that the difference doesn’t lie in access. Bloom’s evidence from India, and similar evidence from Mexico that we discuss in the talent article, also suggests that management practices can improve measurably when they are worked on deliberately.

None of this means everything depends on capabilities. Resources matter: a company with more capital can hire faster, experiment more, buy distribution and withstand periods without profits. Some markets are shaped by scale, regulation or network effects, where capabilities explain little. Capabilities can also be copied or bought, although integrating and coordinating them is still hard. And excellent execution of the wrong strategy only accelerates failure. But capital isn’t capability, and misallocated capital destroys value. Companies with fewer resources can compete through specialization, focus, faster decisions and closeness to customers, without that being a reason to romanticize scarcity. To avoid the bias of looking only at the companies that won, the useful question isn’t just what successful companies do, but what resources and conditions allowed them to do it.

The paradox also changes the role of those who advise companies. If an executive can ask an AI how to do something, an advisor’s value can’t rest on knowing something the client doesn’t. It has to lie in diagnosing what the real problem is, prioritizing, contextualizing, designing, implementing, coordinating, measuring and adapting. Information describes possibilities; the hard work is deciding what they mean for a specific organization and how to make them real. The same logic runs through the topics we analyze at Advisent® What’s Next: expansion and internationalization, the revenue system, brand, sales, operations, artificial intelligence and talent. They aren’t eight independent problems or a rigid sequence, but eight places where the chain between knowing and achieving can break.

From knowing more to being able to do more

For a long time, knowing something others didn’t was a source of advantage. First the internet and then artificial intelligence drastically reduced the cost of accessing much of that knowledge. They didn’t make it irrelevant, but they did shift the question of what differentiates a company. Productivity differences remain large because the distance between knowing what to do and being able to do it isn’t covered with information, but with absorptive capacity, with choices that involve trade-offs, with coordinated execution, with learning and with assets that accumulate over time and must be renewed.

That is why this article doesn’t end with a conclusion, but with three questions. If tomorrow all your competitors had access to the same information, the same tools and the same AI models as your company, what would still be different? Is that difference truly hard to reproduce? And the most uncomfortable one:

Are you strengthening it, or simply adopting the same best practices as everyone else?

Information can show what other companies did. Strategy begins when you have to decide what yours should do, and which capabilities it needs to build to get there.

Frequently asked questions about competitive advantage and growth

Why do some companies grow and others don’t, if they have access to the same information?

Because between information and results there are several links where companies differ: the ability to absorb what is relevant, to choose and accept trade-offs, to adapt general knowledge to their own context, to execute in a coordinated way, to learn and to accumulate capabilities and assets. The evidence shows large and persistent productivity differences between companies in the same industry, which access to information has not closed.

Is information still a competitive advantage?

Less and less on its own. Information available to every competitor can hardly sustain an advantage, because it is neither scarce nor hard to imitate. It still matters: there is evidence that a lack of information holds back the adoption of good practices. But as it becomes accessible, the difference shifts toward what each company can do with it.

What is a company’s absorptive capacity?

It is the ability to recognize the value of new external information, assimilate it and apply it to commercial ends. The concept was proposed by Wesley Cohen and Daniel Levinthal in 1990, who argued that it depends largely on the prior related knowledge the company already has. That is why two companies can get very different value from the same information.

Why don’t best practices work the same way in every company?

Because a best practice tends to work in many contexts, but it has to be adapted to each company’s market, customers, resources, regulation and capabilities. And once a practice spreads, it produces competitive parity: it helps a company avoid falling behind, but it stops differentiating. Michael Porter called this the difference between operational effectiveness and strategy.

What is the difference between competitive advantage and competitive parity?

Competitive parity appears when a company adopts practices or technologies its competitors also have: not having them would be a disadvantage, but having them doesn’t set it apart. Sustained competitive advantage comes from combinations of resources and capabilities that create value for customers and are hard to imitate or substitute. Many investments presented as strategic are really investments to avoid falling behind.

Why doesn’t copying a competitor’s tools produce the same results?

Because the tool is only the visible part of a system. Beneath it are data, processes, decisions and customer knowledge, and further down learning, relationships, accumulated experience and organizational design. Those assets accumulate over time and can’t be bought ready-made: a competitor can copy the visible practice, but not the path that produced it.

If every company uses AI, how can one stand out?

Through what it builds around the models: proprietary data with usage rights, integrated workflows, expert knowledge to evaluate outputs, customer relationships, distribution and feedback loops. Access to similar models tends to produce parity; the difference lies in complementary assets and in the organizational capability to turn technology into value.

What are complementary assets?

They are the resources needed to capture the value of a technology or innovation, such as distribution, brand, customer relationships, production capacity or related services. David Teece showed in 1986 that the value of an innovation is often captured not by whoever creates it, but by whoever controls those assets. The concept helps explain why the same technology produces different results in different companies.

Sources and further reading

Empirical evidence

1. Syverson, C. (2011). What determines productivity? Journal of Economic Literature, 49(2), 326–365. Working paper: NBER
2. Hsieh, C.-T., & Klenow, P. J. (2009). Misallocation and manufacturing TFP in China and India. The Quarterly Journal of Economics, 124(4), 1403–1448.
3. Andrews, D., Criscuolo, C., & Gal, P. N. (2016). The best versus the rest: The global productivity slowdown, divergence across firms and the role of public policy (OECD Productivity Working Papers, No. 5). OECD. Summary: Brookings
4. Bloom, N., Sadun, R., & Van Reenen, J. (2012). Americans do IT better: US multinationals and the productivity miracle. American Economic Review, 102(1), 167–201. Summary: CEP, LSE
5. Bloom, N., Eifert, B., Mahajan, A., McKenzie, D., & Roberts, J. (2013). Does management matter? Evidence from India. The Quarterly Journal of Economics, 128(1), 1–51. Summary: CEP, LSE
6. Tripsas, M., & Gavetti, G. (2000). Capabilities, cognition, and inertia: Evidence from digital imaging. Strategic Management Journal, 21(10–11), 1147–1161.

Strategy theory

7. Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35(1), 128–152.
8. Porter, M. E. (1996). What is strategy? Harvard Business Review, 74(6), 61–78.
9. Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1), 99–120.
10. Dierickx, I., & Cool, K. (1989). Asset stock accumulation and sustainability of competitive advantage. Management Science, 35(12), 1504–1511.
11. Leonard-Barton, D. (1992). Core capabilities and core rigidities: A paradox in managing new product development. Strategic Management Journal, 13(S1), 111–125.
12. Teece, D. J. (1986). Profiting from technological innovation: Implications for integration, collaboration, licensing and public policy. Research Policy, 15(6), 285–305.

The Information-to-Results Gap by Advisent®, the parity-versus-advantage representation and the capability iceberg are proprietary conceptual developments, not validated models. Capability compounding is an Advisent® editorial label based on the work of Dierickx and Cool (1989). The concepts of absorptive capacity (Cohen and Levinthal), operational effectiveness and strategy (Porter), strategic resources (Barney), asset stock accumulation and time compression diseconomies (Dierickx and Cool), core rigidities (Leonard-Barton) and complementary assets (Teece) belong to their authors. The productivity data describe differences between firms; on their own, they don’t establish their cause. The Polaroid figures come from Tripsas and Gavetti’s study; the Continental Lite case, from Porter’s article. The two consulting firms are a hypothetical example.

Last updated: October 5, 2026