Assembly required: 8 myths about knowledge management debunked

The Hamas-led attack on Israel late last year has been called the most significant failure in military intelligence since 9/11. In both situations, intelligence agencies did not piece together critical information that would have warned of the impending threats. Although intelligence operatives, a type of knowledge worker, had many clues suggesting imminent attacks, they did not integrate this information quickly enough to accurately foresee the tragic outcomes. Data was plentiful yet deriving meaning through open dialogue remained elusive.

In response to those shortcomings, after 9/11 the United States formed the Department of Homeland Security, along with many other reforms to break down silos and coordinate intelligence collection, sharing, and analysis. These reforms have been effective, as there have been no subsequent foreign terrorist attacks on American soil. Nevertheless, the ongoing challenge of adapting to the strategies of terrorists and rogue states persists, especially as the volume of data flooding military intelligence capabilities has surged.

General Stanley McChrystal, who once led Joint Special Operations Command, shared this key insight, “The military’s historic struggle with a lack of intelligence has shifted towards a new challenge: managing an overwhelming influx of information. This requires swiftly identifying and assembling relevant data to form an accurate understanding of the situation, akin to completing a 1,000-piece puzzle of an unknown image while facing an advancing enemy.”

With the dawn of generative AI, management gurus continue pitching concepts of turnkey “learning organizations,” and “knowledge management.” While knowledge management is about business process improvement, the bravado and hype that surrounds it often buries what we believe to be the most valuable derivative: how leaders and employees assemble knowledge in pursuit of their mission.

Knowledge assembly defined

Knowledge assembly is a dynamic, creative process that involves gathering diverse, often disparate pieces of information from various digitized and nondigitized data sources, both internal and external to the organization. It entails actively seeking, combining, and synthesizing information to generate new insights and perspectives. Knowledge assembly acknowledges the unpredictable and sometimes accidental nature of knowledge creation, and it requires a culture that incentivizes open information sharing, while ensuring managers have the requisite knowledge to make the most informed decisions.

Knowledge assembly in action

To better understand why organizations fall short when assembling knowledge, we must first understand how knowledge assembly unfolds, starting with some basic concepts:

• Data are raw, unorganized facts, such as numbers, text, and images, that lack context and meaning on their own.
• Information results from organizing, processing, and contextualizing data to provide meaning and relevance, enabling people to answer specific questions such as “who,” “what,” “when,” and “where.”
• Knowledge is created by the systematic process of capturing, synthesizing, analyzing, and interpreting information to achieve a higher level of understanding of some problem or course of action.

The scientific method is the gold standard for creating knowledge. From observations (data), scientists formulate a clear and specific hypothesis that provides a possible explanation for the observed phenomenon.Scientists then conduct experiments to test the hypothesis by collecting and analyzing information to draw conclusions that may either support or refute the hypothesis.The scientific method promotes a systematic and transparent approach to creating knowledge.

Here is an example from the business world: A large global producer of cocoa and coffee grows, dries, and grinds its beans. Since grinding machines differ in their yields, senior leaders at the firm recognized that valuable knowledge could be generated by investigating these differences. Monthly reports aggregate data from all the machines to identify high- and low-yielding ones. Tiger teams of engineers are then dispatched to investigate the causes of these differences. In one case, an engineer spent weeks observing a particularly high-yielding machine until one day, the engineer and the machine’s operator happened to arrive at the same time. The engineer noticed the operator switched the machine on before getting his morning coffee, after which he began operating the grinder 15 minutes later. The engineer surmised the higher yield resulted from allowing the machine to warm up. He tested this hypothesis by having some machines at the facility warm up and others not. The warmed-up grinders produced the highest yields. This knowledge, generated through observation, reflection, study, and social interaction, led to a new companywide policy: “Let the grinder warm up for 15 minutes,” resulting in millions of dollars of extra profit at no additional cost.

Far from the world of coffee, another business organization that relies on knowledge assembly to improve outcomes is the mafia. In our book, Relentless: The Forensics of Organized Crime Business Practices, we describe the daily routine of an undercover FBI agent who infiltrated a New York City crime family. The mobsters hung out at their private social club, discussing possible crimes to commit that night. They would discuss how to fence stolen jewelry, furs, or other products. This routine assembled knowledge, whereby each criminal provided his personal information of goods available for theft, others input their personal information of law enforcement presence, and others input their personal information about possible ways to fence their burgled goods. The knowledge assembled during the day informed the criminal decision-makers about their activities that night.

Some means of knowledge acquisition are more serendipitous. Penn Medicine researchers and recent Nobel prize winners, Dr. Katalin Karikó and Dr. Drew Weissman, for example, revealed that their unexpected collaboration leading to the invention of mRNA breakthroughs began during random conversations while fighting over the use of an office copier.

Serendipity is the unintentional discovery of novel insights, ideas, or solutions that lead to breakthroughs. Serendipitous knowledge creation often involves the ability to recognize the value of unexpected findings. Senior leaders must be willing to explore uncharted territories and embrace accidental discoveries to create new knowledge.

8 myths of knowledge management

Having described some fundamental concepts underlying knowledge assembly, we now discuss common myths and misconceptions that prevent many firms from fully benefiting from a robust knowledge assembly process.

Myth 1: Knowledge management is the same as knowledge assembly

Knowledge management is a business process involving the systematic capture, organization, storage, and dissemination of an organization’s knowledge. It focuses on maintaining and utilizing existing knowledge resources to enhance learning, decision-making, and innovation. This process is akin to managing tangible assets, where the emphasis is on efficient utilization and distribution of resources within the organization.

The term “knowledge assembly” is not as widely recognized a term as “knowledge management.” Knowledge assembly encompasses knowledge management and more. Knowledge assembly is the process of gathering disparate pieces of information from a variety of sources to provide insight. “Knowledge management” suggests managing a known quantity, such as inventory, without taking into account the assembly process — or serendipitous nature — of knowledge creation.

Medtronic is one company that understands this difference. The company, which develops and sells a variety of healthcare products, including insulin pumps, implantable cardiovert defibrillators, and spinal implants and devices, expends enormous resources interacting with healthcare practitioners and then feeds this information back to the relevant Medtronic decision-makers. Medtronic engineers, senior leaders, and scientists combine this customer-provided information with other information about government regulations, manufacturing processes, and new technology to revise long-term strategies, R&D budgets, and capital investment plans. Hundreds of pieces of information, held by numerous Medtronic employees, are “assembled,” not managed, to create new knowledge that is federated to the appropriate decision-makers.

Knowledge assembly requires senior leaders to understand the variety of sources generating new information and to design internal processes that cause those individuals inside (and outside) the organization to honestly disclose their privately held information. For example, off-site retreats for new product planning and quarterly business reviews can be effective means of assembling knowledge.

Myth 2: Employees freely share their information leading to knowledge assembly

Corporate cultures that prioritize individual achievements over collective success cause information to become a personal asset rather than a resource to be shared. Knowledge workers will often use private information for personal advantage. By failing to provide clear financial and nonfinancial incentives for information sharing, such organizations inadvertently encourage employees to hoard information for job security, career advancement, or maintaining a competitive edge over colleagues. Fostering a culture of trust, collaboration, and clear incentives for information sharing are essential to create an environment where individuals feel empowered to contribute their information freely.

Prior to Satya Nadella becoming CEO in 2014, Microsoft had a toxic, non-innovative culture known for information and product silos, cutthroat competition through forced ranking of employees, and office politics. Nadella recognized that to enable innovation he had to re-ground the culture in a new set of values that encouraged learning, sharing, and collaboration. He and the leadership team promoted new “growth mindsets” and unleased curiosity to drive cross-functional sharing of information and problem-solving.

One tangible practice to accelerate knowledge assembly were hackathons directed across Microsoft that have attracted tens of thousands of employees to think together, share knowledge, and spur innovation. Microsoft’s market value has exploded in the years since Nadella took over — due in no small part to the way he has helped them reimagine knowledge assembly.

Myth 3: Generative AI can solve the knowledge assembly problem

AI is certainly a disruptive technology, but it is only one part of knowledge assembly. AI, and especially large language models (LLMs) such as ChatGPT, can use only digitized data from text, images, and voice. There is an enormous and growing amount of digitized data; yet it still represents a small amount of digitized and nondigitized information combined. Nondigitized data are in everyone’s brain — trillions of bits of everything we have ever experienced. Our biological neural networks sift and process all the inputs we get.

A Medtronic manager observing a leading neurosurgeon implanting a Medtronic implant creates information in that manager’s brain. This information is not on the web and thus is not accessible to an LLM. So, while AI can be useful for finding patterns in digitized data, it falls short when much of the information is held by individuals in the form of past conversations, experiences, and impressions. Until brain-computer interfaces exist that can transfer data and information in human brain neurons to the cloud, humans will remain your best bet for assembling knowledge. In the meantime, generative AI can greatly enhance the process, making it more efficient and effective.

Myth 4: Remote working promotes knowledge assembly

Remote work, a transformative trend accelerated through COVID-19 lockdowns, offers flexibility, convenience, and the promise of a better work-life balance. It also hinders knowledge assembly within organizations.

Economists refer to “knowledge spillovers” that occur by chance within the workplace and in social settings. These spillovers result when people cluster to exchange ideas. One Tesla engineer described the value of meeting in person: “[T]he people on our [Tesla] Autopilot team are always sitting together, and the ideas flow real fast, and what we do as a team is better than what any one of us could do.” He also noted that was why Elon Musk favors in-person rather than remote work (Walter Isaacson’s Elon Musk, p519). Researchers have also found that inventors living near other inventors generate more patents.

Serendipitous interactions are important for creative, innovative, or nonformulaic activities. Harvard economist Edward Glaeser wrote, “Many of the finest achievements of human civilization occurred because smart people learned from one another in cities. … [T]he chance encounters facilitated by cities are the stuff of human progress.” Steve Jobs understood the importance of knowledge spillovers from clustering when he designed Apple Park with separate pods for individual work, team meetings, and group socializing. Jobs envisioned a porous structure where ideas would be more freely shared across common spaces.

Remote work impedes serendipitous interactions both within the workplace and in public spaces. Scheduled virtual meetings replace the spontaneity of in-person interactions, leading to a more structured, formal communication environment that compromises the dynamic nature of collaboration and information sharing. In a recent survey, the preponderance of respondents said that remote work makes employees feel disconnected from the organization and weakens the organization’s culture.

Another study used smartphone geolocation data to measure face-to-face interactions among workers at various Silicon Valley firms. The study documents “substantial returns to face-to-face meetings … (and) returns to serendipity.”

Not all knowledge assembly results from serendipitous interactions. Nonetheless, ample anecdotal evidence documents that serendipity plays a major role in creating new knowledge, novel products, and unexpected combinations. Hybrid remote models requiring, say, three days a week in person present more opportunities for serendipitous knowledge creation than fully remote work. But with fewer random encounters at work and beyond, hybrid work models still result in less knowledge spillovers and hence less knowledge assembly.

Paradoxically, senior leaders must now incentivize new ways to facilitate moments of serendipity between employees that once helped generate valuable new ideas and innovations.

Myth 5: Budgeting has nothing to do with knowledge assembly

Budgeting is a time-consuming process that serves two main functions: planning and control. In planning, it aligns activities with the organization’s goals, sets objectives, and allocates resources based on knowledge assembled from hundreds of pieces of information from across the organization. In its planning role, budgeting is an annual forcing function with usually quarterly updates of rolling forecasts that assembles the latest knowledge.

As a means of control, budgets measure performance against planned targets, influencing employee behavior. Tying compensation and promotions to meeting and surpassing budgeted performance incentivizes managers to exert additional effort to achieve the budget and control spending.

But a tension exists between these functions. When rewards are tied to budget targets, it can lead to distorted information (“sandbagging”), where managers understate sales forecasts or overstate costs to set lower, more achievable targets. Basing managers’ pay on beating budgets impedes knowledge assembly.

In dynamic industries where knowledge decays rapidly, pay should not be tied to achieving budget targets. This eliminates managers incentive to bias their personal information assembled during the budgeting process, thereby improving knowledge assembly and decision-making. Instead, companies should use metrics other than budget targets for rewards. For example, multinational healthcare giant Johnson & Johnson rewards its subsidiary senior managers, not for meeting budget targets, but rather on their ability to develop new markets, solve problems, add value to their organization, and manage and motivate their subordinates.

Myth 6: Learning organizations routinely assemble knowledge

Learning organizations promote training and continuous learning at all levels to continually transform themselves and adapt to change. They value employee development and innovation through off-site and online courses. Many firms operate in-house “universities” that offer courses in everything from AI to leadership. But training and knowledge assembly are not the same. While valuable, these programs do not directly lead to generation of new knowledge but train employees in how to use existing knowledge.

In contrast, knowledge assembly involves converting information into new knowledge. Nvidia offers extensive educational opportunities for professional development to its employees that emphasizes continuous learning and skill enhancement. But beyond providing employee learning, Nvidia has invested about $1.55 billion in more than 13 startups offering Nvidia a window into developing tech trends. These strategic investments give Nvidia leaders information about what features are needed in their chips to improve the next iteration. Knowing what emerging tech startups want in microprocessors enables Nvidia senior leaders to assemble knowledge that drives its R&D program.

Successful firms must provide continuous learning and assemble knowledge. Both are very important, but very different. Senior leaders should not conflate the two.

Myth 7: All firms must aggressively pursue knowledge assemblyTop of Form

Firms in dynamic, highly competitive, or life-and-death settings must routinely assemble knowledge to survive. Such organizations include innovation-driven companies, R&D-intensive firms, healthcare services, global enterprises with diverse markets, professional services firms, customer-centric organizations, and highly regulated industries with complex compliance requirements. The pertinence of knowledge in these industries erodes quickly as new technologies disrupt markets and new entrants arise, so new knowledge must replace obsolete knowledge.

Amazon, being in several dynamic, highly competitive markets, relies heavily on a robust knowledge creation process. Jeff Bezos’ founding principle was a relentless focus on the customer. He built Amazon’s culture to prioritize customer feedback and preferences. Amazon continuously learns and adapts its strategies to better serve its audience, notably fast delivery times. It leverages vast amounts of data about customer interactions, purchases, and browsing behaviors to derive actionable insights. Its knowledge assembly culture reflects a commitment to constant innovation and learning via experimentation and risk-taking that has led to a diverse range of businesses, from e-commerce to cloud computing. It also creates knowledge by leveraging a vast network of partners, sellers, and developers.

Companies in mature, stable industries can rely on their existing knowledge for long periods of time and hence have less demand for robust knowledge assembly systems. These industries are characterized by slow growth rates, well-established products or services, and a relatively stable market structure. Firms in these industries can best use their budgeting systems to build performance measures based on beating the budget (i.e., the control role of budgets).

Myth 8: Knowledge management software can assemble knowledge

The chief knowledge officer for NASA’s Johnson Space Center describes NASA’s Knowledge Online as a knowledge-sharing clearinghouse containing massive searchable databases with everything from past spacewalks to archived lectures. But he goes on to lament that NASA retirees leave with “so much data … in their heads that is not written down somewhere.” The fact that NASA is not capturing retirees’ “data” demonstrates NASA Online is not a knowledge management system but an information clearinghouse. Calling it a “knowledge management system” doesn’t make it so.

Numerous knowledge management systems — ranging from content management systems to portals to CRM systems and project management tools — promise better ways of managing knowledge. Characterizing these as “knowledge management systems” constitutes false or at best misleading advertising. These systems are collaboration and information sharing tools. They handle only digital data generated through customers, policies, or projects, not information and knowledge sitting in human brains that can only be surfaced through dialogue, serendipity, or more reflective processes that force leaders to take a step back and think broadly.

Conclusion

Leaders in rapidly changing industries must move beyond traditional knowledge management and adopt a more dynamic and creative process of knowledge assembly. They must understand that knowledge is not static but a dynamic resource.

When it comes to knowledge assembly, one size does not fit all. Each firm is unique, operates in different environments, and faces varying rates of technological change and government regulation. Managing the firm’s knowledge assembly processes requires high-level executive understanding of the organization’s core competencies for creating customer value. CEOs and COOs are often best positioned to know exactly where knowledge assembly must happen, yet no member of the C-suite is immune from addressing the eight myths. Discovering what knowledge is important to assemble, who has the key inputs of information, how to create the culture and incentives to encourage free-flowing information sharing, and what knowledge various decision-makers require is essential. CIOs and chief digital officers can be force multipliers by linking their strategic and operational plans to the key moments in the enterprise when knowledge assembly matters the most.

The journey toward effective knowledge assembly isn’t merely about leveraging technology or assigning responsibilities. It’s about fostering a culture where incentives are aligned and help facilitate open information flows, where serendipity routinely sparks innovation, and where learning becomes an everyday pursuit. As we challenge these myths, we encourage leaders to embrace a future in which information isn’t just managed or assigned to a machine for processing, but knowledge is meticulously and intentionally assembled to propel organizations forward, one shared insight at a time.