What is Knowledge Work?

In my last article I state "knowledge work transforms information into refined information". In this article I will break down:

• definitions for information and knowledge work
• a data driven method for measuring how much knowledge work is performed per role

There are No Concrete Definitions for Knowledge Work

Knowledge work, as introduced by Peter Drucker in 1959, involves primarily intellectual tasks such as problem-solving, critical thinking, and the application of specialized knowledge acquired through formal training. Drucker recognized knowledge workers as the most vital resource in modern organizations due to their role in developing products and services. He further identified knowledge worker productivity as the next major management challenge, underscoring the importance of strategically leveraging these professionals’ expertise to drive institutional success.

Nonaka (1991) emphasizes that knowledge should be viewed as a dynamic, evolving resource rather than a static input. He asserts that organizations, more akin to living organisms than mechanical systems, thrive by actively renewing and transforming knowledge. Central to this approach is the role of knowledge workers, who act as catalysts for innovation and continuous improvement.

Savage (1995) positions the knowledge-centric era as the third major socio-economic wave, following agriculture and industry. In this era, wealth is tied to owning and applying knowledge to create or enhance products and services, encompassing factors like cost, durability, suitability, timeliness, and security. Projections indicate only 2% of the workforce remains in agriculture and 10% in industry, leaving the majority as knowledge workers.

All of these definitions talk about what type of work is done in knowledge work, but none specifically dilineate who does knowledge work and who does not. To build this definition, we need to first define the input ("information") and how it is different from raw data. Then we can frame other variables in roles that measure how much knowledge work is being performed.

Data

Data can be in the form of numbers, text, images, or any other type of input. It refers to raw, unorganized facts or figures that are collected and stored in some type of information platform. Some examples of data include:

• A spreadsheet containing sales figures
• A database with customer records
• Temperature readings from a weather station
• GPS coordinates from a mobile device

Data can be messy, scattered, and independent, making it less immediately useful. By itself, data lacks context and meaning. Most people in an organization don't work with raw data, at least not in non-trivial ways. Data is the most basic form of representation and requires further processing to become useful. Most of the time data first needs to be joined with other data, aggregated from multiple external parties, or multiple internal systems into a logical information source.

By structuring and organizing data (such as transforming a series of sales figures into a statement about sales performance) organizations create information that is both interpretable and actionable. This transformation is vital in industries such as the insurance industry, where meaningful insights drawn from data can guide strategic decisions and ultimately lead to improved risk management, pricing, and customer relations.

Information

Information is the processed and organized form of data. It is data that has been analyzed, structured, and given context. Information provides meaning and can be used to answer questions or make decisions. It is the result of data being transformed into a more meaningful and useful state.

• A sales report highlighting top-selling products
• An article summarizing research findings
• A graph showing the correlation between two variables

Data and information are closely related but serve distinct functions in any organizational context. Data typically consists of raw and unrefined numbers, symbols, or characters that lack a clear purpose on their own. For example, figures such as “10,” “blue,” or “A” are merely data points unless they are given context. Information, on the other hand, emerges when these data points are organized and contextualized. In this sense, it has inherent value because it supports decision-making, provides insight, and can be used to form knowledge.

Workers who create value in the core line of business use information, not raw data, as their input. Processing raw data into information is a function of IT (data engineering, data warehouse teams) and is not logically part of the role responsibilities in the medical field, insurance industry, or the banking industry.

Building a Definition for Knowledge Work

So if we broaden our definition for knowledge work, we can state our role classification process such as:

• a worker takes information (visual, written, digital) in as input to consider
• the worker assesses the input against their internal state and accumulated knowledge (e.g., "information processing", "cognitive labor")
• the worker then produces output based on instructions in the form of
  • a physical action or physical labor (not knowledge work)
  • outputs new information that is based on the input and their existing internal knowledge (is knowledge work)
• In the case of outputting new information, this may either be input information for someone else, or input to their own next cognitive task

To summarize, if someone is not actively producing physical labor but they are processing information and outputing information to be used in other tasks or by other people, then that is our minimal definition for knowledge work.

More simply: "knowledge work is cognitive labor that takes information as input and outputs more information".

Roles that are Not Knowledge Work

Examples of work that are not knowledge work:

• Forklift operator
• Concrete mixing operator
• Bricklayer

Note that all 3 economic roles above, while valuable in their own right, use information (e.g, "sight", "written instructions") to produce physical labor outputs as their way to create value.

Roles that are Knowledge Work

Examples of knowledge work (e.g., "cognitive labor")

• claims analyst in an insurance company
• family practice doctor
• accountant

These 3 examples don't perform cognitive labor in the same amounts as one another, but none of them have "physical labor output" as their primary way to be productive and create value.

Baselining with Economic Data

The Bureau of Labor Statistics publishes many wonderful datasets on the United States economy. They also write many reports analyzing this data, such as bls.gov: Minds at Work - whats required according to the occupdational requirements survey where they gather economic data and analyze the "cognitive and mental requirements" of every role in the U.S. economy.

These measurements are meant to quantify things such as:

• "How much problem-solving time goes into an architect’s job?"
• "What types of mental capabilities are required of employees who work in a telework environment?"
• "What types of interpersonal skills are required for customer service jobs?"

While they aren't specifically using the term "knowledge work" in this study, it is a data-based economic survey that measures "cognitive and mental requirements" of each role in the U.S. economy. I consider this to be the best proxy variables available for a data-driven analysis on the subject.

Measuring Knowledge Work per Role and Industry

I calculated a knowledge work factor per economic role (0.0 to 1.0, based on the variables listed below), to get a relative measure on how much knowledge work is in a given role. (Note: not perfect, but still "better than arbitrary definitions" and it gives us a relative measure to compare any role in the economy.)

1. Problem solving: Percent of workers, problem solving is required more than once per day
2. Education, training, and experience: Percent of workers, minimum education level is a bachelor's degree
3. Public interaction and people skills: Percent of workers, with more than basic people skills
4. Able to step away from work: Percent of workers, with the ability to pause work
5. Telework: Percent of workers, with the ability to work remotely

Once we have a knowledge work score for every role, we then create a industry sector by weighting each role's impact on the industry score by what percent of total employment for the given role for a given industry sector.

In the table below we extend our "structural transformation" table to show our calculation for knowledge work measure per NAICS industry sector.

As stated above by Mosco and McKercher (2007), under a broader definition of knowledge work we might consider all jobs to have some component of information processing to them. Under that lens we then want to compare every sector to one another with respect to how much knowledge work is performed. This "knowledge work"-score per industry sector allows us to see a relative score for "how many people have roles that score high in the 'knowledge work' attributes".

So What is Knowledge Work?

In this article I've framed knowledge work as something that is a relative component of most roles in today's services heavy economy. I then showed a data-driven analysis of the relative measure of knowledge work per industry sector so the reader can better understand their own market segment.

I go on to quantify that Non-knowledge work economic roles are focused on producing labor as their value creation output; Knowledge work economic roles are focused on producing information as their output (information to be used in labor, or information to coordinate with other workers).

For the purposes of understanding knowledge work in practical application in the US labor force, I define knowledge work as any activity that involves the above criteria and does not result in labor as the output. If it outputs more information to be processed further downstream in a workflow, then the activity is considered knowledge work.

Knowledge work is simply the transformation of information into more information.